A Review of Hybrid Renewable and Sustainable Power Supply System: Unit Sizing, Optimization, Control, and Management

Abstract

Since rising worldwide energy consumption is anticipated with increasing rapid industrialization and urbanization, green energy sources have become the ineluctable choice among energy engineers, power engineers, and researchers for carbon-free and sustainable electric power generation. By integrating several energy sources, a hybrid renewable and sustainable power supply system (HRSPSS) has been created to solve the global warming problem. HRSPSS aims to develop contemporary electricity grids that benefit society, the environment, and the economy. However, there is a need for thorough assessment of these complex HRSPSSs for making the most use of renewable energy potential and carefully crafting suitable solutions. This paper provides a thorough investigation of the most effective methods for sizing, optimizing, controlling, and managing energy, as well as how to combine different renewable energy sources to create a hybrid sustainable power supply system. Information on several software simulation tools and optimization methods that have been used to support HRSPSS development, research, and planning is presented in this study. Additionally, this study covers energy management and control strategies that have been used to ensure efficient and optimal operation of HRSPSS. Furthermore, this article presents an extensive comparison among various strategies utilized in each area (sizing, optimizing, controlling, and managing energy) to provide conclusive remarks on the suitable strategies for respective applications. The outcome of this study will help various stakeholders in the energy sector to make appropriate decisions during the design, development, and implementation phases of a hybrid sustainable power supply system.

1. Introduction

In the near future, industries will continue to exhaust all available resources to fulfill the world’s growing energy needs [1]. Due to the spread of the Coronavirus and the total lockdown, the world’s energy use dropped by 5.9% in 2020 compared to 2019. A recent study indicates that global economic growth led to a 3.8% reduction in power consumption during the first quarter of 2020 compared to the same period in 2019 [2]. Traditional sources of energy cause more CO₂ to be released into the air, which may be one of the primary causes for global warming [3]. Also, most of the world’s remote communities (like those on islands and in rural areas) get their energy from traditional sources that must be shipped in, which makes fuel costs go up a lot with distance. Renewable or “green” energy sources, like solar, wave, wind, tide, water, etc., have been pushed to solve the problems above [4]. Even though green sources are thought to be the most potentially environmentally friendly energy sources, their main problems are that they are hard to predict and rely on weather and climate conditions for high power output. Renewable energy sources do not always make energy, so they might not always be able to meet the load’s needs. Green energy sources like those listed above can be used to make a good mixed power supply design that can always meet high load demand, while also making the system more efficient and reliable [5].

The global energy landscape has seen considerable transformation in recent decades, influenced by rapid industrialization, increasing urbanization, and demographic growth. By 2020, global energy consumption had been rising steadily, underscored by a shift toward renewable energy to address climate change challenges [6]. In 2019, global total final energy consumption reached approximately 418 exajoules, predominantly fueled by fossil fuels, which constituted about 80% of total consumption. Despite ongoing diversification efforts, coal, natural gas, and oil continued to dominate the energy mix. This heavy reliance led to elevated CO₂ emissions, with global energy-related CO₂ emissions peaking at a historic 33 gigatons in 2019, according to the International Energy Agency (IEA). This figure experienced a subsequent decline, largely attributed to the widespread effects of the COVID-19 pandemic [7].

In parallel, developing nations face significant challenges in ensuring reliable electricity access, with an estimated 759 million individuals lacking basic energy services in 2019, predominantly in Sub-Saharan Africa and parts of Asia. Notably, there have been significant improvements in electrification rates, spurred by the adoption of cost-effective renewable energy technologies such as solar and wind [8]. These advancements contribute to the broader energy transition, characterized by an increased integration of renewables into the global energy grid. By 2021, renewables like solar, wind, and biomass comprised nearly 29% of global electricity generation. In response, numerous countries have established ambitious goals to diminish fossil fuel reliance and enhance renewable integration, with the European Union targeting climate neutrality by 2050 and China aiming for carbon neutrality by 2060 [9].

Focusing on Bangladesh, the government has pledged to generate 15% of its total energy from renewable sources by 2030. Although considerable progress has been made in deploying solar, wind, biomass, and hydroelectric power, significant hurdles persist, particularly in electrifying rural regions. The potential of hybrid renewable energy (HRE) systems emerges as a promising solution to provide electricity in off-grid areas, where extending the national grid remains economically unviable. The total grid-based generation capacity of Bangladesh has increased to 25,951 MW at present, marking a sixfold increase compared to the 4942 MW capacity in 2009. These hybrid solutions could play a crucial role in bridging the electricity access gap, especially in isolated communities [10,11,12].

A single centralized controller and many local controllers manage the various energy storage, as well as power source, mechanisms in a hybrid power supply system. The provided illustration, denoted as Figure 1, displays a generalized block diagram of an HRSPSS that utilizes renewable sources, such as solar photovoltaic, wave, and wind energy. The centralized controller manages the energy for all of the generators, batteries, and local controllers. It shifts between the main and secondary sources depending on the state of the system, the availability of power, and a set of specified control logic [13]. At each sampling instant, the local controller sends command signals to the master controller to adjust the voltage and current of various power sources and storage devices. In addition, the local controller monitors system power management by controlling the DC-DC bidirectional converters using pulse width modulation signals. Power is shown flowing from sources both main and secondary to the DC connection, and then from the DC link to the loads, represented by the solid arrow lines. Control actions and data flows between controllers and their sources are represented by dashed arrow lines. The HRSPSS provides reliable and constant power to the loads.

The parameters that make up an applicable HRSPSS optimization method are shown in Figure 2. Regardless of the complexity of the project, conducting a thorough theoretical feasibility analysis and rigorously evaluating the reliability of results through specialized simulation software and relevant case studies are critical for effectively understanding, designing, analyzing, and planning an HRSPSS project [15]. Consequently, accurate project design and efficient RES consumption are closely linked to thorough HRSPSS studies [16]. To use renewable energy sources correctly and economically, an optimal sizing approach is required. The optimal sizing strategy may help secure the lowest investment by maximizing the use of PV, wind, and storage, enabling the hybrid system to operate at its most dependable and economical levels. You may obtain the optimal balance between cost and dependability by setting cost targets and monitoring the system’s performance over time [17]. It has been discovered that a variety of commercial software packages, including HOMER Pro v3.11.5, PVsyst v6.35, PVSOL v2023 (R7), iHOGA v2.2, RETScreen v9.0, INSEL v8.3.1.0b, TRNSYS v18.06.0002 and others, are helpful in sizing and optimizing HRSPSS.

Recent scholarly reviews have extensively analyzed the application of optimization techniques, control structures, and sizing methodologies within hybrid renewable and sustainable power supply systems (HRSPSSs). Thirunavukkarasu et al. (2023) highlight the potential of AI-based techniques for rapid global optimization in HRSPSS, yet they suggest that hybrid approaches combining multiple algorithms could better address the complexities of sizing and design [19]. Similarly, Dawoud et al. (2023) discuss optimization strategies in microgrids, emphasizing the essential need for appropriate storage solutions and component placement to enhance both the economic and operational efficiency of these systems [20]. The intermittency of renewable sources, as noted by Ammari et al. (2022), poses significant challenges that necessitate robust control strategies to ensure a consistent energy supply [2].

Further contributions to the field are presented by Saharia et al. (2018), who detail the effectiveness of evolutionary algorithms such as genetic algorithms and Particle Swarm Optimization in controlling and sizing HRSPSS. They identify a critical gap in the comprehensive strategies required to dynamically integrate multiple energy sources and storage units [21]. Additionally, Ishaq et al. (2022) and Khan et al. (2022) stress the need for advanced optimization and control methods to handle high penetration levels of distributed energy resources within microgrids, pointing out that the current control infrastructures are insufficient for ensuring stability and efficient operation [22,23].

In terms of energy management strategies, Modu et al. (2024) and Tyagi and Singhal (2024) advocate for hydrogen storage and advanced uncertainty modeling methodologies, respectively, as essential components to mitigate the variability of renewable energy sources [24,25]. While hydrogen provides a viable solution to the storage limitations posed by conventional batteries, its adoption faces hurdles due to high costs and infrastructural challenges. A total of 38 publications were reviewed, compared, and classed to offer an overview of the present status of optimization and simulation projects for hybrid clean energy structures, highlighting concisely and accurately the key trends and findings. An overview of the existing reviews considering different aspects of HRSPSS is presented in

Table 1. A summary of the literature review.

Author(s)	Year	Contribution	Limitations	Ref.
Thirunavukkarasu et al.	2023	AI and hybrid algorithms for optimizing HRSPSS.	Lacks comprehensive analysis of unit sizing, control strategies, and EMS.	[19]
Dawoud et al.	2023	Optimization techniques for microgrids with emphasis on storage and component placement.	Comprehensive comparative analysis of control systems and detailed evaluation of energy management strategies are not included.	[20]
Saharia et al.	2018	Use of evolutionary algorithms for HRSPSS control and sizing	No critical examination of advanced optimization tools and a comprehensive scrutiny of control strategies within hybrid systems.	[21]
Ammari et al.	2022	Review of methods for HRSPSS sizing, optimization, and management	Use of cutting-edge computational tools and techniques for optimization and control within HRSPSS frameworks is not discussed.	[2]
Ishaq et al.	2022	Review of control strategies and optimization methods for microgrids	Severely lacks a systematic comparative study on the efficacy of control and management system across various microgrid configurations.	[22]
Khan et al.	2022	Optimization methods and management strategies for HRSPSS	Critical analysis of different EMS characteristics and integration of modernized management systems are not covered.	[23]
Modu et al.	2024	Hydrogen storage in hybrid renewable and sustainable power supply systems	Exploration of sophisticated control techniques and critical examination of management systems for hybrid configurations are not explored.	[24]
Tyagi and Singhal	2024	Sizing and uncertainty modeling methodologies for hybrid energy systems	Critical evaluation of the efficacy of diverse EMS and control strategies under conditions of uncertainty is not performed.	[25]

.

From Table 1, it can be summarized that most prior studies primarily concentrated on the control and management aspects of hybrid renewable and sustainable power supply systems (HRSPSSs) without offering an in-depth critical analysis of the prevailing methodologies. Additionally, while several reviews have addressed sizing and optimization strategies for HRSPSS, there is a noticeable lack of detailed investigation into the software tools and modern approaches employed for these purposes. Furthermore, earlier research did not adequately compare and present different control systems or thoroughly evaluate the characteristics of various energy management approaches. Consequently, there is a substantial gap in the literature regarding a comprehensive discussion and conclusive analysis concerning which types of controllers or management systems are most effective for differing configurations of HRSPSS.

In contrast, this paper provides an extensive overview of the latest research concerning four critical dimensions of HRSPSS: sizing, optimization, control, and management, respectively. This includes discussing conventional and commercial tools for sizing, as well as hybrid, traditional, and artificial methods for optimization. The paper also explores various methods for controlling integrated renewable energy systems—ranging from centralized to multilevel approaches—and managing their electrical infrastructure with strategies focused on economic, techno-economic, and technical aspects. Additionally, this work addresses the optimal sizing of hybrid systems that are either grid-connected or operate independently, drawing on the limitations and challenges outlined in previous research. The contributions of this study are articulated as follows:

• Review and analysis of tools and methods for sizing, optimizing, controlling, and managing HRSPSS, integrating both traditional and advanced AI-based techniques.
• Evaluation of various optimization and control techniques, detailing their effectiveness, limitations, and suitability for diverse applications, and providing insights into different control paradigms and energy management systems for informed decision-making.
• Highlights the significance of hybrid strategies that combine classical and AI approaches, which optimize system performance and reduce processing times.
• Provides specific guidance on selecting the most appropriate sizing, optimization, and control methods tailored to the unique requirements of different HRSPSS configurations.

The HRSPSS framework, as illustrated in Figure 3, provides a comprehensive structure for integrating renewable energy resources, storage systems, control strategies, and key processes to address the complexities of modern energy systems. The fundamental components—renewable energy sources, energy storage systems, and controllers—provide the essential inputs for the system’s functionality. These inputs feed into the key processes, such as system sizing, optimization, and energy management, which ensure efficient operation and stability. For instance, renewable energy sources like solar and wind require storage systems (batteries or hydrogen) to address intermittency, while controllers (centralized, distributed, or hybrid) regulate energy flow. The optimization processes, supported by classical, hybrid, and AI-based methods, improve system performance, and energy management techniques leverage these processes to forecast demand, balance loads, and integrate seamlessly with the grid.

The framework also addresses challenges inherent to renewable energy systems, such as the intermittency of resources, cost management, and grid reliability. To overcome these limitations, it suggests integrating advanced AI solutions, developing improved storage systems, and aligning with supportive policies and regulations. These future directions are designed to enhance system performance, reliability, and scalability.

The following is the outline for the review article: Section 2 provides an overview of the present state of size techniques based on commercial applications or conventional techniques, achieving the benefits and limitations of each approach. The optimization techniques utilizing artificial, hybrid, and classical approaches are covered in the next section. Section 4 of the document delves into the various techniques employed to regulate, as well as the characteristics, geographical placement, and installed potential of HRSPSS. Section 5 illustrates the management system employed by HSPSS. A summary is provided in Section 6. Furthermore, Section 7 of the document addresses the challenges encountered and outlines the proposed approach for future research. The rest of the sections offer a conclusive overview that encompasses a thorough analysis and concluding reflections.

2. Methods for Sizing a HRSPSS

The size of the hybrid system is a key stage in identifying the capacity of the generators. There is a risk of minimizing or oversizing the system if proper sizing is not used. The most difficult obstacles are evaluating real load and step time to appropriately account for changes. Most studies, however, use average months, days, or hours as data samples. The main issues and conclusions that provide a wealth of information for further study on hybrid system capacity configuration are emphasized [26].

2.1. Procedure Scaling Methodologies

Figure 4 shows a graphical representation of various methodologies used for quantifying units based on meteorological data. Conventional methods categorize weather data based on energy balance and reliability, but this may not be feasible in remote areas. Researchers are exploring the use of artificial intelligence (AI) methodologies like artificial neural networks (ANNs), wavelet transforms, Particle Swarm Optimization (PSO), genetic algorithms (GAs), and fuzzy systems to address this issue, especially in remote or secluded regions where meteorological data are unavailable.

Equation (1) represents the Energy Balance Equation, which is a foundational component in system sizing. This equation ensures that the total energy generated by the system is sufficient to satisfy the anticipated demand.

$$\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{B}\mathrm{a}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{e}=\sum \left(\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{I}\mathrm{n}-\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{O}\mathrm{u}\mathrm{t}\right)$$

(1)

Reliability is a key factor in system sizing, especially in remote areas. Equation (2) represents the generic reliability formula:

$$\mathrm{Reliability}\left(\mathrm{R}\right)={\mathrm{e}}^{-\mathsf{\lambda }\mathrm{t}}$$

(2)

where λ represents the failure rate, and t represents the operational time.

2.2. Software-Assisted Hybrid Renewable Energy-Based System Sizing

There are several commercial programs accessible for calculating hybrid system sizes, including the Hybrid Optimization Model for Electric Renewable (HOMER), RET Screen, Integrated Simulation Environment Language (INSEL), and Hybrid Optimization by Genetic Algorithm (iHOGA) [29]. Windows and the Visual C++ programming language are the backbones of the majority of these applications. Both modern software and more traditional techniques of sizing are available (see Table 1).

A hybrid energy charging station that uses wind and solar power was built and optimized using HOMER software, as reported by Orhan Akren et al. [30]. Across the world, the sizing process may be applied where 44.4% wind energy and 55.6% solar energy make up the ideal hybrid system configuration, which produces 843,150 kWh of power per year for 0.064 USD/kWh. In [31], simulation tools HOMER and RET Screen are used to assess the performance of every power system design in terms of economic evaluation and optimization. To determine the suggested system’s minimal net present cost (NPC), two ideal systems were simulated in [32] using HOMER software. According to the sizing and modeling, the PV, wind turbine, and bio-generator capacity factors and contribution percentages for the optimum-1 system were determined.

Table 2. Comparison between different sizing-method software [19,33].

Software	Utilization	Features	Advantages	Drawbacks
HOMER Pro v3.11.5	Microgrid design and optimization Off-grid and remote power system design	Optimization and sizing Scalability and flexibility	Comprehensive modeling capabilities User-friendly interface	Lack of dynamic modeling Limited customization options
iHOGA v2.2	Research and development Multi-objective optimization	Life-cycle assessment Optimization algorithms User-friendly interface	Advanced optimization algorithms Detailed component modeling	Limited integration with other tools Limited database of components
Hybrid2 v1.3b	Education and training Techno-economic assessment	Extensive documentation Open-source and flexible Optimization capabilities	Detailed time-series Simulation Open-source and flexible	Limited database of components Limited commercial support
RET Screen v9.0	Project feasibility assessment Cost–benefit analysis	Collaboration and sharing Multilingual and globally applicable Feasibility assessment	Equipment performance modeling Multilingual and globally applicable	Potential data limitations Specialized training required
DER-CAM v5.9	Scenario and sensitivity analysis Integrated energy system modeling	Grid impact analysis Flexibility and customization	Integrated load and generation modeling Grid impact assessment	Complexity and steep learning curve Limited out-of-the-box functionality

outlines the utilization, features, advantages, and limitations of various software programs for sizing mixed renewable energy systems. The best software should be chosen based on the system’s usage and optimization. HOMER and RETScreen are the best tools for this, as they can size any system but only use simple optimization equations. PVsyst, a specialized software, cannot analyze renewable energy systems using other sources. Hybrid2 has internal issues, and DER-CAM may have less community support.

2.3. Traditional Approaches for Sizing Hybrid Renewable Energy-Based Systems

Traditional scaling methods include the following:

• Analytical method,
• Probabilistic method,
• Artificial intelligence methods.

2.3.1. Analytical Method

This technique perceives the combined system as a numerical model and specifies the size of the combination system as an indicator of viability. Nassar et al.’s work [34] offers a simple and practical approach to calculate the size of a pumped hydroelectric storage (PHS) combined hybrid PV–wind power system needed to power a densely populated region like Libya’s Brak. The study [35] simulates a system of hybrid energy storage that includes a supercapacitor and a hydrogen fuel cell. The goal is to determine the ideal size for a commercial load powered by solar panels. In [21], the authors primarily examine how hybrid renewable energy-based systems (both off-grid and connected to the grid) are currently classified, how they are evaluated, and how their sizes are determined. The paper categorized HRSPSSs based on their demand and development status in different locations of the world, dividing them into the more conventional PV–wind systems and the hybrid systems with hydro/PHS.

2.3.2. Probabilistic Method

The probabilistic techniques for sizing an integrated framework consider the influence of wind speed isolation and system design modifications. This is among the most basic size approaches, but the findings demonstrate that it might not be the greatest option for finding the optimal answer [18].

Table 3. Probabilistic methods synthesis.

Setup	Parameters of Input	Objectives	Advantages	Disadvantage	Ref.
PV/wind/ Storage	Wind speed and solar radiation	Specific load profiles, comparing cost versus EIR	Enhancing the forecasting precision and providing quantified uncertainty information	Lack of discussion on practical considerations	[39]
PV/wind/ DG	Wind speed and solar radiation	Assessing the long-term performance and reliability	Provides a direct way to quantify the long-term energy performance of the HSWPS	Focusing only on the wind as an imbalance driver	[40]
Wind/diesel generator	Wind speed	Calculating fast-response reserves	Reduction of the fast-response reserves required	Laplace distribution and symmetric fluctuations are not accurately	[41]
PV/wind/ DG/ battery	Wind speed/solar radiation/inverter efficiency	Presenting a Monte Carlo simulation	Efficiently model the performance of a hybrid RE system under uncertainty	Complexity of models and computationally intensive	[42]
PV/wind/ battery	Wind speed and solar irradiation	Demonstrating the effectiveness of a community	Providing a practical application and validation	The potential uncertainties or variations are not considered	[43]

presents a summary of probabilistic approaches. The annual cost of a hybrid microgrid covering an island that uses solar photovoltaic, tidal currents, wind turbines, batteries, and diesel generators are estimated in [36], using an upgraded multi-objective Grey Wolf Optimizer (GWO) technique, to keep the system’s yearly expenses as low as possible and avoid the risk of a power distribution deficit. Han et al. in 2019 [37] proposed an energy-balanced PV–fuel cell–battery system with a DC microgrid. Hassanzadeh Fard et al. [38] proposed a solution to minimize the cost of operating a mixed microgrid system with variable energy demands that utilizes solar panels, wind blades, batteries, a fuel cell, and a reformer.

2.3.3. Artificial Intelligence Methods

Artificial intelligence is defined as “the capacity of a machine or artifact to carry out similar types of functions that Characterize thinking as human” in its widest definition. Garud et al. [44] and Mellit et al. [45] discussed how to size mixed system elements using AI approaches like GA, PSO, ANN, FL, or a combination. Based on energy balancing, Kaabeche and Bakelli [46] suggested sizing an independent wind and solar output system. The nature-inspired Ant Lion Optimizer (ALO), Grey Wolf Optimizer (GWO), Krill Herd (KH), and JAYA algorithms minimize unit energy cost (UEC), while meeting customer stability needs.

Table 4. AI methods used in hybrid-system sizing.

System Configuration	AI Technique	Contributions	Limitations	Ref.
PV/WT/battery	Generic algorithm	Offers cost-effective and reliable power solutions for standalone environments	Environmental impact of fossil fuels	[47]
PV/wind/battery/ converter	Artificial neural network (ANN)	Offers high performance, efficiency, and precision in power system control	Solar and wind energy have limitations related to energy instability	[48]
PV/battery	Combination of GA and ANN	Enhances sizing of photovoltaic systems in remote areas.	Regression models can affect the accuracy of predictions	[49]
Wind farms based on WIPSO	Combination of neural network wavelet transform	Reduces prediction errors significantly, improving wind farm operational efficiency	Algorithms significantly impact the accuracy of predictions	[50]
PV/wind/battery/ converter	Particle Swarm Optimization (PSO)	Designing independent microgrids based on sustainable resources with EV integration.	Uncertainties of wind turbine and PV output power not modeled	[51]
PV/WG	Generic algorithm	Simulation results show hybrid PV/WG systems have lower costs	Battery size not included in the optimization process	[52]
PV/battery/diesel	Artificial neural network (ANN) and fuzzy inference system	Cost-effectiveness comparison of hybrid system with standalone PV and DG	Require proper maintenance of the system	[53]
PV/wind/FC/battery/electrolyzer	Artificial neural network (ANN)	Application of supplemental algorithms to enhance ANN performance	Optimization difficulties due to low-quality renewable energy data	[54]
PV/storage/DG	Genetic algorithm and Teaching/Learning-Based Optimization Algorithm	Network training for levelized COE, annual energy, and capacity factor	Lack of experts in energy sources and advanced technology knowledge	[55]
PV/wind/battery	Particle Swarm Optimization (PSO)	Evaluation of system reliability using loss of power supply probability	Battery life span is a limitation due to the least life span	[56]

displays the artificial intelligence techniques that were employed in the sizing of hybrid systems, and

Table 5. Overview of several studies on ideal size.

Techniques	System Elements	Purposeful Operations	Limitations	Ref.
Generic algorithm (GA)	CSP/PV	Reduce total initial investment and LCOE, and maximize capacity factor	Storage options for PV systems are limited	[57]
Particle Swarm Optimization (PSO)	PV/fuel cell /battery	Comparison and investigation of ICA and PSO algorithms	Energy stored decreases	[58]
Multi-objective self-adaptive differential evolution	PV/DG/WT/ battery	Reduce LPSP and COE in two microgrid system designs.	Limitations in technical and economic optimization	[59]
Iterative approach	PV/wind/ storage	Cut down on life cycle costs (LCCs)	Higher initial cost	[60]
Discrete Harmony Search (DHS)	PV/biodiesel/ wind/battery	Cut down on life cycle costs (LCCs)	Higher pollution emissions	[61]
Simulated Annealing–Chaotic Search	PV/wind/ Storage	Reduce the whole life cycle’s cost (TLCC)	Systems require a higher initial investment	[62]
Artificial neural network (ANN)	PV/WT/ hydrogen	Reduce the whole life cycle’s cost (TLCC)	Hydrogen tank capacity is limited	[63]
Particle Swarm Optimization (PSO)	PV/WT/ storage	Reduce LCC	Optimization methods are computationally intensive and time-consuming	[64]
Fuzzy logic	PV/wind/ storage	Cut down on the system’s annualized cost (ACS)	Resources can cause problems in system reliability	[65]
Artificial Bee Swarm Algorithm	Hybrid solar/ HT/FC/battery	Determine the component sizes	Lack of adaptive mechanisms	[66]
Cuckoo Search (CS)	PV/wind/ battery	Reduce the TC	Algorithm’s performance is challenging	[67]
Mine Blast Algorithm (MBA)	PV/wind/FC /DG/HT	Reduce the yearly total expense (TAC)	Low efficiency of the fuel cells	[68]
Multi-Objective Line-Up Competition Algorithm (MLUCA)	PV/wind/DG /battery	Reduce the amount of TAC and greenhouse gas (GHG) emissions	Unreliable power supply	[69]

presents an overview of numerous research studies on the appropriate size.

2.4. Comparison Between Sizing Methods

The performance analysis of traditional approaches is constrained due to their simplicity and speed. Artificial intelligence utilizes multi-objective variables to solve complex problems, and this limitation may be ignored when utilizing this methodology. It is possible to solve this problem through an iterative approach based on a simple algorithm that uses a recursive process to attain the optimal system size. The drawback of this approach is that it ignores certain significant parameters. Utilizing a basic numerical model, the analytical method sizes hybrid systems rapidly with low adaptability. Artificial intelligence, in contrast to earlier techniques (analytical and iterative), is the most effective way to handle complicated procedures. It can operate without constraints and achieve superior outcomes compared to other methods. The complexity codes that this algorithm employs pose the greatest challenge.

Table 6. Comparison of features and sizing techniques employed in the hybrid system [18,70].

Methods	Input Variables	Advantages	Drawbacks
Graphical method	Solar radiation and wind velocity	Enabling more adaptability	Challenging when there are more than two variables
Load analysis method	Solar and wind energy	Enhanced system reliability	Uncertainty in load estimation
Analytical method	Sun radiation and wind speed on a statistical basis	Cost-effective system design	Limited system dynamics representation
Energy balance	Values for solar and wind statistics	Simplicity	Limited optimization capabilities
Artificial intelligence (AI) methods	Statistical solar average	Self-adaptation, forecasting, and prediction	Complexity and computational requirements
Probabilistic method	Wind speed and sun radiation statistics	Aids in network’s complete variety	Unsuitable for dynamic performance

illustrates a comparative analysis of diverse sizing techniques employed in hybrid systems.

3. Optimization

A well-planned simulation program helps determine the optimal dimensions of battery banks, PV arrays, wind turbines, hydro generation capacity, and other generation systems for a grid-integrated or autonomous HRSPSS depending on the load and the intended chance of power supply failure criteria. Optimization methods are divided into three main categories: hybrid, classical, and artificial. The artificial methods are illustrated in Figure 5.

3.1. Classical Methods

Differential calculus is used by classical optimization methods to provide the most effective methods for differentiable and continuous functions. In instances where the objective variables are not distinguishable and/or continuous, the classical approaches are not as effective. Nonetheless, there are several traditional optimization techniques, such as linear programming model (LPM) [27,72,73,74], Multi-Choice Goal Programming [75,76], Multi-Objective Evolutionary Algorithms (MOEAs) [77,78], mixed-integer linear programming (MILP) [79,80,81], dynamic programming (DP) [82], and non-linear programming (NLP) [83,84].

3.2. Hybrid Methods

Any approach that combines two or more algorithms and enables utilizing their benefits to go beyond the limitations of a single algorithm is considered hybrid [85]. Fathy et al. [86] developed a Social Spider Optimizer (SSO)-based technique to determine the optimal size of a hybrid RESs Integrated Microgrid (MG) with solar (PV) cells, wind turbine (WT) generators, batteries, diesel generators (DGs), and inverters. Akpan et al. [87] designed, sized, and optimized a solar–wind hybrid power system using Hybrid Optimized Model for Electric Renewable (HOMER) software to determine its financial viability. An overview of current optimization techniques is presented in

Table 7. A summary of modern optimization methods.

Methods	System Components	Objective of Study	Advantages	Limitations	Ref.
Genetic algorithm	PV/DG/WT/battery	Reduced LCE and CO2 emissions	Considers both component sizing and energy management parameters	Error in heat demand estimation	[92]
Fuzzy logic	PV/battery/WT	Control the hybrid system’s production.	Integrated energy storage and hydrogen production	Systems require the integration of storage	[93]
Artificial neural network (ANN)	PV/battery/WT	Resolve the technical issue (lower DC-DC switching loss)	Efficient DC-DC converter for integrating hybrid renewable energy sources into a microgrid system	AGONN control schemes are complex and costly	[94]
Fuzzy analytic network process	RE/fuel oil/gas oil and nuclear	Creating a new framework to evaluate the combination of energy sources	The FANP method allows for a comprehensive and nuanced assessment	Energy mix has negative environmental effects	[95]
Fuzzy clustering	PV/FC/WT	Reduce electricity outages, voltage, expense, and emissions	Evaluates the feasibility and effectiveness of the MHBMO algorithm	Complexity and computational overhead	[96]
Particle Swarm Optimization (PSO)	PV/thermal storage/CSP	Determine the best size and reduce the levelized cost of energy (LCOE)	Exploring complex dispatch strategies	Middle power output points may decrease the energy stored	[97]
Multi-Objective Crow Search Algorithm (MOCSA)	PV/FC/DG	Minimize the LPSP and Total NPC	Integration of hydrogen energy technology	Operating reserve significantly increases the system cost	[98]

. Moreover, there are hybrid methods such as the Group Method of Data Handling Neural Network and Modified Fruit Fly Optimization Algorithm (GMDHMFOA) [88], Genetic Algorithm and Particle Swarm Optimization (GAPSO) [89], Multi-Objective Crow Search Algorithm (MOCSA) [90], and Monte Carlo simulation combined with Multi-Energy-Balance/Financial Equations [91].

3.3. Artificial Intelligence Techniques

Artificial intelligence techniques are extensively utilized to enhance the financial benefits of a hybrid system through optimization. Examples of these techniques include artificial neural networks, genetic algorithms, fuzzy logic, PSO (Particle Swarm Optimization), and ACO (Ant Colony Optimization). An outline of AI techniques for solving optimization issues is displayed in Figure 6. The multi-objective methods, some of which are based on EA, are the primary strategies employed.

3.3.1. Genetic Algorithm

Table 8. Optimization of HRSPSS using genetic algorithms.

Setup of the System	Utilizing Algorithms	Noteworthy Involvement	Advantages	Limitations	Ref.
PV panel	PV panel parametric optimization	To maximize the parameter values that were found for the PV panel’s diode equivalent model	Higher efficient optimization of the PV parameters	The identification method increases the computational complexity	[104]
Solar system	Techno-economic optimization of PV system	Optimized system sizing on an hourly basis utilizing categorized hourly solar radiation data	Provide a robust methodology for determining the optimal standalone PV system design	The algorithm may not be suitable for all types of data	[105]
PV/wind/ HC/ electrolyzer/FC	LPSP optimization, LPSP, and CO2 emissions	This project investigates the implementation of an HRSPSS in structures to develop zero-energy buildings	Improving reliability and reducing grid electricity dependency	Increased frictional losses and thermal problems	[106]
PV/DG/ WTG/BESS/ FC/FW/UC and AE	Optimizing the controller gain for the hybrid system	In the time domain, the GA-optimized controller surpasses the typical classical controller in terms of setup time, overshoot, and oscillations.	Optimization GA techniques which can lead to improved performance and stability	Reliability issues when using magnetic bearing technology to overcome frictional losses	[107]
PV/wind/DG/ battery	Minimizing the total/net present cost	Minimization of total cost, greenhouse gas emissions, and system probability	Enhances renewable energy fraction, reduces COE, and maintains reliability	Complex and difficult to solve using classical mathematical methods	[108]
PV/battery/ BG/wind/ dump load	Optimization of technology, economics, and LCE	To look at a hybrid energy system that includes a solar module, biogas generator, wind turbine, and vanadium redox flow battery to provide steady electricity	NSGA-II and IDEA techniques are superior to the HOMER software tool in terms of cost and environmental performance	The multi-objective optimization technique results in lower life cycle emissions compared to the single-objective optimization technique.	[109]
PEMFC for telecom	Technique vs. the Ziegler–Nichols control approach	Simple control logic, adaptability, robustness, tracking improvement, and flexibility to satisfy shifting load needs	Reducing operation and maintenance costs for telecom companies	Backup supply increased the operational expenditure, unreliability of power supply, and environmental concerns.	[110]

depicts the use of genetic algorithms to optimize hybrid renewable and sustainable power supply systems. Due to its low algorithm complexity and strong global search efficiency, the genetic algorithm (GA) in evolutionary computing has emerged as one of the most important algorithms in the current optimization landscape [100]. Big-data life prediction, which is crucial to fuel-cell life prediction and extension, can also benefit from the use of GA [101,102]. It is among the key technologies in computing with artificial intelligence. The formulas for genetic algorithms (GAs) [103] are outlined in Equation (3) through Equation (7).

$$Fitness(x)={\displaystyle \frac{1}{1+f\left(x\right)}}$$

(3)

$$P\left(x_i\right)={\displaystyle \frac{Fitness\left(x_i\right)}{{\sum }_{j=1}^{n}Fitness\left(x_j\right)}}$$

(4)

$${\mathrm{Child}}_1={\mathrm{parent}}_1\left[:c\right]+{\mathrm{parent}}_2\left[c:\right]$$

(5)

$${Child}_2={parent}_2\left[:c\right]+{parent}_1\left[c:\right]$$

(6)

$$x_i^{\prime }=x_i+\left(1-2\cdot \mathrm{rand}\left(\right)\right)$$

(7)

Equation (3), the Fitness Function, evaluates solution efficacy inversely to the objective function to prioritize optimal solutions. Equation (4) details the Selection Process, which probabilistically selects solutions for reproduction based on their relative fitness, promoting the survival of the fittest. Equations (5) and (6) describe the Crossover Operation, where genetic information from two parents is mixed at a designated point to generate diverse offspring. Lastly, Equation (7) outlines the Mutation Operation, introducing random changes to solutions to explore new areas of the solution space and prevent convergence on local optima.

3.3.2. Fuzzy Hybrid Machine Learning

The use of technology research and artificial intelligence (AI) technologies has been greatly enhanced to model and compute attempts in solving actual building challenges. Derrouazin et al. managed the energy flux of a hybrid system comprising solar photovoltaic, battery, and wind turbine using fuzzy logic multi-input/output. The acquired results demonstrate that the electronic switch signals successfully and instantaneously monitor the hybrid power system-imposed input energy states [93].

The usage of a fuzzy logic-based algorithm in HRSPSS is presented in

Table 9. The use of a fuzzy logic-based algorithm in hybrid sustainable energy-based systems.

Setup of the System	Utilizing Algorithms	Noteworthy Involvement	Limitations	Ref.
PV/WT/DG/FC/AE/FW/UC and BESS	PSO-tuned a fractional-order fuzzy logic controller (FO-FLC)	Traditional PID and integer order fuzzy PID are outperformed by the chaotic PSO-tailored FO fuzzy controller	Parameters using the PSO algorithm may add complexity to the control system	[123]
PV/wind turbine hybrid system	The fuzzy PSO algorithm	On the inverter, the PI and fuzzy PI-based technique is used to regulate grid-injected current	A fuzzy control scheme may be more complicated compared to the PID control	[124]
PV/fuel cell	Fuzzy logic controller	The PV system’s MPPT is used to regulate the DC-DC boost converter	Single-phase grid may limit scalability and ability to handle larger loads	[125]
PV/WT/battery	The Pareto-fuzzy (IPF) algorithm is iterative	The excess electricity is also considered when determining how well the system operates	Wind speed and sun irradiation might create power variations	[126]
PV/SOFC/BESS	FLC-based voltage–frequency control	Comparing FLC vs. traditional PI controllers	Modeling of the SOFC may not fully capture real-world conditions	[127]
PV and battery	FLC with PSO optimization	FLC and low-pass filters are being explored	Battery peak current reduction is insufficient	[128]

. Fuzzy logic has been combined with machine learning, a branch of artificial intelligence. Machine learning mainly focuses on creating computational techniques and models that can learn from data and address various problems. Researchers have taken advantage of this combination by using fuzzy logic with machine-learning models. Construction difficulties such as data categorization [111,112], predictive modeling [113,114], and pattern recognition [115] are some uses of fuzzy hybrid machine-learning algorithms. Fuzzy hybridizations using ANNs [116,117,118,119] and fuzzy clustering approaches [120,121,122] are examples of frequent fuzzy hybrid machine-learning approaches. The formulas for fuzzy hybrid machine learning [93,111,112,113,114,115,116,117,118,119,120,121,122] are outlined in Equation (8) through Equation (10):

$$R:\mathrm{If}x\mathrm{is}A\mathrm{and}y\mathrm{is}B\mathrm{then}z=f\left(x,y\right)$$

(8)

$$u=\sum _{i=1}^n\left(w_i\cdot \mathrm{COG}\left(A_i\right)\right)$$

(9)

$$w_i^{\left(t+1\right)}=w_i^{\left(t\right)}+\eta \left(\widehat{y}-y\right)x_i$$

(10)

Equation (8) illustrates a fundamental fuzzy logic rule, defining how inputs are mapped to outputs through fuzzy associations. Here, x and y are input variables associated with fuzzy sets A and B, respectively; and z is the output determined by the function, f, which maps the inputs to an output based on the fuzzy logic.

Equation (9) describes the defuzzification process, where the control output, u, is computed as a weighted sum of the centers of gravity of the fuzzy sets, translating fuzzy inferences into actionable outputs; w_i represents the weights assigned to each fuzzy rule; and COG(A_i) is the center of gravity for the fuzzy set A_i, which helps in converting fuzzy values back to a crisp output, facilitating actionable decisions.

Equation (10) focuses on the dynamic adaptation of fuzzy rules, updating weights via a learning process that minimizes prediction errors by adjusting the degree of change based on the difference between predicted and actual outputs. Here, w_i^(t) is the weight of the i-th rule at time t; η is the learning rate, adjusting the degree of weight modification; and $\widehat{y}$the predicted output, $y$ is the actual output, and $x_i$is the input feature, driving the learning process to minimize prediction errors.

3.3.3. Particle Swarm Optimization (PSO)

This method takes advantage of the movement of fish or birds while concurrently using their constantly shifting locations and speeds in a space with three dimensions where every single particle (or animal) stands for a potential solution [129]. Liu et al. [97] employed a technique in conjunction with a genetic algorithm to optimize a hybrid sustainable system based on solar and concentrator solar photovoltaics with batteries. The results show optimization in both technical (10.92% stationarity outcomes and 305.94 GWh power output) and economic (civilized cost with 16.33 ¢/kWh) parameters when compared with other systems and strategies.

Table 10. Application of Particle Swarm Optimization (PSO) algorithms for HRSPSS.

Setup of the System	Utilizing Algorithms	Noteworthy Involvement	Limitations	Ref.
PV/wind	Fuzzy MPPT and PSO fuzzy MPPT control	MPPT algorithm development involving both WT and PV generator output	The efficiency of the PV system remains poor	[129]
PV/wind/DG/battery	Dispatch-coupled scaling	The dispatch-coupled sizing method integrates the battery to improve the economic dispatch.	The PSO algorithm may not be able to optimize the economic dispatch	[68]
PV/WT/fuel cells	Mine Blast Algorithm (MBA)	The MBA algorithm is compared to the PSO, CS, and ABC algorithms	Using FC is costly for maintenance and availability of hydrogen fuel	[131]
Self-contained PV system	Optimization and power management approach	To optimize the combination system, PSO and HOMER are compared	Higher NPC and CO2 emissions compared to systems without desalination	[132]
PV/wind/MT/battery	Power dispatch algorithm	To satisfy the equality requirements for power balancing, a roulette-wheel allocation technique is presented	The DoD effects and deep discharging can reduce the life cycle of the battery and increase replacement costs	[133]

illustrates the use of Particle Swarm Optimization techniques for the HRSPSS. The formulas for Particle Swarm Optimization (PSO) [130] are outlined in Equation (11) through Equation (12).

$$v_i^{\left(t+1\right)}=\omega v_i^{\left(t\right)}+{\varphi }_p\cdot {\mathrm{rand}}_1\cdot \left(p_i-x_i^{\left(t\right)}\right)+{\varphi }_g\cdot {\mathrm{rand}}_2\cdot \left(g-x_i^{\left(t\right)}\right)$$

(11)

$$x_i^{\left(t+1\right)}=x_i^{\left(t\right)}+v_i^{\left(t+1\right)}$$

(12)

In Equation (11), ω is the inertia weight, which moderates the influence of a particle’s previous velocity, helping to balance the global and local exploration. The cognitive (ϕ_p) and social (ϕ_g) scaling coefficients guide the particle toward its personal best (P_i) and the global best (g), respectively, ensuring that each particle learns from its own experiences, as well as the successes of others in the swarm. Additionally, random numbers, rand₁ and rand₂, introduce stochastic elements to the velocity update, fostering diverse exploration of the search space and enhancing the algorithm’s ability to avoid local optima and discover more optimal solutions.

Equation (12) adjusts the position of particle i for the next iteration, directly adding the newly calculated velocity to the current position. This step physically moves the particle within the search space toward potentially more optimal solutions.

3.4. Optimization Software

Simulation tools are widely used to evaluate the performance of hybrid systems, assessing energy generation costs and efficiency. Popular software tools include HOMER Pro v3.11.5, iHOGA v2.2, HYBRID2 v1.2, and HYBRIDS v2.0 Using hourly environmental data, as well as simulations, the National Renewable Energy Laboratory’s Hybrid Optimization Model is used to assess hybrid renewable energy before making improvements to the system online. Numerous investigations on the optimal design of HRSPSS involving ESS have been conducted using HOMER [134]. A diesel generator–photovoltaic–wind–battery hybrid [135], a photovoltaic–wind hybrid [87], a mini-hydro–wind combination [136], a solar–biomass combination [137], and a hydro–solar–wind hybrid [138] were all optimized using HOMER. Ref. [139] describes how a researcher uses a WT-PV-DG to improve the design of a generator of biogas for a mixed remote area electricity system in a far-off community. The intermittent nature of wind speed and solar irradiation meant that this hybrid system could not provide a consistent supply for the associated demands. Energy storage technology should be included in HRSPSS to solve the reliability issue.

A multitude of scholars are investigating the optimal methods for building hybrid renewable energy-based power stations using HOMER software, connecting the national grid system, and incorporating energy storage equipment with remote area electrification systems.

Table 11. The ideal HRSPSS system size, as determined by the HOMER software optimization program.

Setup of the System	Objective Purpose	Design Restrictions	Tariff for Electricity	Ref.
PV/micro hydro/DG	Operating expenses and return on capital invested	Power distribution and wiggle room	Real-time pricing in steps	[140]
PV/BES/DG	COE and NPC	Load demand, construction cost, accessible energy supplies	Useful time	[141]
Wind/PV/BES/FC	NPC and COE	Power distribution and wiggle room	Useful time	[142]
PV/wind/DG/BES	NPC and LCOE	Production of electricity, emissions, operational costs, and fuel usage	Useful time	[143]
PV/wind/DG/BES	NPC, COE, and RF	Diesel fuel cost and project lifespan	Useful time	[144]

summarizes the optimal HRSPSS system size, as estimated by the HOMER program optimization tool.

3.5. Techniques for Metaheuristic Optimization

Of all the optimization techniques available for renewable energy systems, metaheuristic optimization is the most accurate and popular. The WT-PV-BES system that the researcher built in [145] was optimized in an ensemble of twenty houses, leading to economical, emission-free power production with lower energy prices. In other research, like [146], there were four distinct algorithms employed to look at the effectiveness of the metaheuristic method for hybrid renewable energy’s ideal sizing foliage.

Table 12. Meta-heuristic optimization approaches are used to optimize single and multi-objective capacity for HRSPSS.

Setup of the System	Optimization Method	Advantages	Design Constraints	Ref.
PV/wind/BG/DG/BES	PSO-GWO PSO Hybrid	Optimizes algorithm effectively	Optimal configuration in terms of cost	[147]
PV/BES/wind/FC	PPO stands for proximal policy optimization.	Overcomes the challenges	Overall cost savings and decrease in carbon emissions	[148]
PV/FC/wind	Hybrid search optimization with Firefly and Harmony	Provides a significant amount of energy exchange	Techno-economic and power dynamics	[149]
PV/PHS/wind	Genetic algorithm	Achieves 100% energy autonomy	Power dynamics and techno-economic	[150]
PV/DG/wind/BES	NSGA-II technique	Performs a sensitivity analysis	Power dynamics and techno-economic	[151]

displays the reference number, design constraints, decision variables, optimization techniques, and electricity tariff of the metaheuristic studies that are currently available on both single- and multi-objective optimum design of HRSPSS.

The primary advantage of every research activity is the researchers’ strategy for resolving the community’s electricity-related issues, as indicated in the review articles in

Table 13. Studies involving centralized control scheme.

Setup of the System	Purposes/Goals	Descriptions	Advantages	Limitations	Ref.
PV/battery/ hydrogen	Assesses the techno-economic viability	Developing and analyzing three microgrid scenarios by considering energy balance and techno-economic optimization, using the “HOMER Pro” software	Reducing carbon footprint, contributing to sustainable energy solutions	The H2 production requires additional engineering controls for safe utilization	[152]
PV/wind/ battery	Evaluate and contrast the HESS control techniques	Exploring, dividing, and examining the impact of communication-system time delay on controllers and presenting a novel droop coordinated control method	Simplifying and adding multi-functionality to the HESS controller	Need for improved control accuracy and dynamic performance tracking	[153]
PV/FC/electrolyzer /battery	Regulating energy flow throughout the system	Presenting a hybrid system for home micro-grid customers, optimizing energy flow based on energy availability	Improves the system’s dependability and energy balance	The higher average cost of batteries compared to the power supplied by fuel cells	[154]
PV grid-connected	Developed a modified control strategy	The modified perturb and observed MPPT are utilized for precise power tracking in variable irradiation, while the voltage source inverter control synchronizes grid and HRGS voltages	The proposed strategy enhances system efficiency and stability	The efficiency of the PV array used in the system is low. Additionally, the systems require a storage unit	[155]
PV/battery/ Converter	Utilizing multi-optimization techniques	The study formulates an optimization problem using linear programming to ensure optimal usage of the PV system and BESS, avoiding excess AC grid power consumption	Higher efficiency and reliability of DC microgrid operation	Lack of comprehensive experimental validation	[156]

. In a hybrid system, diesel fuel-based generators are not viable from an environmental and financial standpoint. Furthermore, BES is not economically feasible. Subsequent investigations ought to encompass the previously stated advantages and disadvantages, ensuring that they are explicitly tackled in the green hybrid system configuration that will be utilized on the system.

4. Control System of HRSPSS

To provide a constant energy supply for load demand, energy flow management is essential. An economical, dependable RE-based system with great efficiency is ensured by an ideal energy management plan. Power quality concerns like voltage and frequency control at the user end are caused by the dynamic interplay between renewable energy sources and load demand. Consequently, to combat the transient response in the energy distribution network, management and supervision of systems based on renewable energy are required. In general, the control strategies of such hybrid systems are divided into four categories, as seen in Figure 7: centralized/master, hybrid control paradigms, distributed, and multilevel control paradigms. In all three scenarios, it is assumed that each power-generating source has a local controller that may decide how to proceed most effectively with the relevant unit based on available data.

4.1. Centralized Control Paradigm

Centralized control in small-scale hybrid renewable and sustainable power supply systems offers superior efficiency, enhanced performance, and a simplified setup, while also being more cost-effective than dispersed or hybrid control. It acts as an energy supervisor for the entire system, making control choices based on the measured signals from all units [158]. The main goal of the control paradigm is to maximize the system’s usage of its multiple power sources and storage capacity. The appropriate output power is then ensured by sending the control signals to the storage device and the related power sources. The technique is vulnerable to single-point failure and has a high computational overhead.

Many of the papers have used a centralized control scheme to control the hybrid renewable energy-based system. Mainly, centralized controllers are used in standalone systems. Table 13 presents various studies that have been conducted utilizing a centralized control scheme.

4.2. Distributed Control Paradigm

In a completely distributed control paradigm, the hybrid system’s energy sources provide measurement signals to their local controllers. The controllers interact with each other to make compromised (Pareto) operational decisions and achieve global optimization. One advantage of this technique is the ease of “plug-and-play operation”. This control structure reduces each controller’s calculation burden significantly and eliminates single-point failure concerns, yet the potential complexity of its communication system remains negative. Fuzzy logic, neural networks, evolutionary algorithms, and hybrid combinations are examples of intelligent, model-free algorithms that may be used to solve such challenges. The Multi-Agent System (MAS) is a promising solution for distributed control problems. It has been applied to power system integration, restoration, reconfiguration, and microgrid power management, among other uses [159].

Table 14. Studies involving distributed control schemes.

Setup of the System	Purposes/Goals	Descriptions	Advantages	Limitations	Ref.
PV/battery/ converter	Investigate the small-signal stability	Simulations and experimental studies are conducted to show the efficiency of the suggested control strategy in an islanded DC microgrid	The proposed scheme effectively splits power between batteries and SCs, enhancing system performance in compensating for power imbalances	Scalability and performance are not extensively evaluated	[160]
DC microgrid	Performed stability analysis	Investigation of parameter uncertainties on controller performance through simulations	Assess controller effectiveness compared to traditional PID controller	Lack of consideration for parameter uncertainties	[161]
PV/FC/SC/ battery	To achieve excellent load-sharing	The study utilized dynamic models for the solid oxide fuel cell (SOFC) unit and the voltage source inverter (VSI) with an LCL filter	Utilized dynamic models for the SOFC unit and VSI with an LCL filter, enhancing the accuracy of the study	Did not address economic feasibility, scalability, or integration	[162]
PV/wind/MT	To improve the microgrids’ performance	Ensures the stability of voltage/frequency (V/F) and active/reactive (P/Q) power parameters of the microgrids	Modifying and restoring the control system, enhancing the overall efficiency of the microgrid units	Complexity and difficulty in management	[163]
Hybrid AC /DC microgrid	To achieve power sharing in a distributed manner	The proposed method significantly improves the robustness in terms of communication delay and variation in the status of DGs	The effectiveness of the proposed method is verified through small-signal analysis and controller-hardware-in-the-loop verification	Risk of single-point failure (SPF) which degrades system reliability	[164]

presents various studies that have been conducted utilizing a distributed control scheme.

4.3. Hybrid Control Model: Combining Centralized and Distributed Elements

A hybrid control paradigm incorporates both distributed and centralized control mechanisms. This proposal combines renewable energy-based sources into an integrated system. Every group uses a distributed control method for coordination, whereas each group uses a centralized control scheme. Within each group, local optimization is accomplished using centralized control in this hybrid control paradigm, while distributed control is used to provide global coordination across the various groups [165]. Hybrid control schemes are applied where both standalone and grid-connected systems are present.

Table 15. Studies involving hybrid control schemes.

System Configuration	Purpose/Goal	Noteworthy Involvement	Practical Implications	Limitations	Ref.
PV/wind/battery	Creating a RE-based plant with low volatility	Examines and evaluates RE-based power plant’s economic performance using a profit index, based on Italian electricity exchange rules	Provides guidelines for designers and researchers in renewable energy	Lower efficiency and higher capital cost	[166]
PV/wind/battery/converter	Design an EMS to achieve power exchange balance	The developed control scheme aims to improve dynamic behaviors and reduce the computational burden	Discusses limitations and solutions for enhancing smart microgrid performance	Wind generation system is complex and challenging to implement	[167]
PV/wind/DG	Testing a 14-busbar IEEE distribution system	Providing a detailed model of this MG using the MATLAB/Simulink vR2024a simulation platform, offering a base case for various studies	Offers tools for stability analysis, demand response, and energy storage strategies	The microgrid model does not involve any physical implementation or experimental validation	[168]
Microgrid clusters	To maximize energy storage and the utilization of RE sources	Proposes an optimum stochastic control technique for islanded microgrid clusters, based on jump linear theory, to maximize energy storage and renewable use	Ensures connective stability and maximizes energy storage in islanded clusters.	Complexity and potential challenges in terms of system integration and communication	[169]
MG network model	Enhance the resilience of networked microgrids (MGs)	Utilizing a stochastic linearized OPF for flexibility and a consensus approach for power exchange	Enhances resilience of networked MGs with MESSs through a three-stage approach	Communication delays or failures did not consider	[170]
PV/wind/converter	Study and analyze a hybrid RE system	Presenting a comprehensive study and review of architectures of hybrid RE systems, specifically focusing on the coupling of solar and wind energy with storage (battery)	Control systems enhance efficiency and energy conversion in hybrid systems	Excess power generated by the system that cannot be stored in the battery	[171]

presents various studies that have been conducted utilizing a hybrid control scheme.

4.4. Multilevel Control Paradigm

A multilevel control strategic level is presented in Figure 8. Although there is an additional strategic (supervisory) control level, this strategy is nevertheless identical to the hybrid control scheme outlined above. At the operational level, this scheme’s fundamental decisions about real-time operation are made, and each power-generating unit’s actual control is carried out according to its control objectives extremely quickly, i.e., within a millisecond range. Also, this system has two-way communication so that choices may be carried out at many levels.

Table 16. Studies involving multilevel control scheme.

System Configuration	Purposes/Goals	Description	Practical Implications	Limitation	Ref.
PV/DG/ storage/ converter	Developing and testing of a hierarchical power control and management scheme	The proposed FLC controller demonstrated superior performance in voltage regulation compared to the PI controller in various simulated scenarios.	Enhances reliability and efficiency of off-grid hybrid power systems	The FLC requires a linearized model of the system, which is difficult to obtain	[174]
PV/wind/ battery/ converter	Develops the manner of operation for solar and wind power systems	A hybrid control architecture for cascaded multilevel inverters (CMLI) with a grid-dependent hybrid system that includes wind and solar generating subsystems is suggested	Minimize the variation in system parameters and external disturbances	Algorithms (RSA and GBDT) may increase the computational overhead	[175]
Grid-connected power injection method	Enhanced CHBMLI-based grid-connected hybrid system (HWSECS)	This research proposes a five-level CHBMLI that converts HWSECS electricity into alternating current, optimizing the use of wind power sources and PV arrays independently	Addressing capacitor balancing issues in multilevel inverter topologies	May not be directly applicable to a system with more than two DC sources connected	[176]
FC-MLI/ DC-MLI/CHB-MLI	Control and design of a switched-diode dual-source single-switch MLI	The study explores a modified fish swarm optimization technique for estimating optimal switching angles for SDDS MLI, enhancing voltage quality by removing low-order harmonics	Reduces harmonic distortion and switching losses in industrial energy applications	Lacks a comprehensive analysis of multilevel inverter topologies and modulation techniques	[177]
PV/wind/ converter	Improving cascaded H-Bridge multilevel inverter	Balancing the DC link capacitors and supplying the power grid with a minimal ripple sinusoidal current ensures good power quality	Improved power quality in grid-connected hybrid wind–solar energy systems	Systems require complex control strategies	[178]

presents various studies that have been conducted utilizing a multilevel control scheme. From

Table 17. Comparison of different control paradigms [179,180].

Features	Centralized Control Paradigm	Distributed Control Paradigm	Hybrid Control Paradigm
Multi-objective energy management	Possible	Possible	Possible
Computation burden	Very high	Low	Medium
Single-point failure	Occurs	No occurrence	No occurrence
Multi-agent system (MAS)	Not possible	Possible	Possible
Control approaches	Conventional	Fuzzy logic, genetic algorithm, etc.	Conventional and advanced both
Energy flow management	Medium	Hard	Easy
Inter-communication	Not possible	Possible	Possible
Advantages	Energy savings	Failure is minimal	Allows performance at the local level
Drawbacks	Unified system	Complexity	Complexity

, it is quite clear that concentrated solar hybrid systems and wind–hydro hybrid systems are the two most popular types of hybrid power plants.

5. Energy Management System

Renewable energy sources are unpredictable and intermittent, requiring careful planning and construction to mitigate their limitations. Solutions include integrating multiple sources, providing alternative sources, and battery storage, but there are no universal solutions [181]. Renewable system prices may increase due to extra design concerns and energy flow regulation. To minimize system costs and maximize positive effects, component size optimization and energy management strategies can be implemented, ensuring the system’s efficiency and effectiveness [182]. HRSPSS requires optimal power management and control strategy due to intermittent electrical power from renewable sources and varying consumer load demand. The goal is to satisfy peak load demand, ensure system reliability and efficiency, and minimize costs [183,184].

This review paper discusses various energy management system (EMS) strategies, including conventional, AI, and real-time or online methods, used in hybrid renewable energy-based systems, as shown in Figure 9. It covers independent and grid-connected hybrid setups, highlighting the potential for learning from optimal setups in any HRSPSS.

5.1. Linear and Non-Linear Programming-Based Energy Management

Table 18. Summarizes several works on linear and non-linear programming based on energy management systems.

System Configuration	EMS Approach	Remarks	Practical Implications	Limitation	Ref.
PV/FC/wind /battery	Linear programming using Simulink	Calculating the repositioning costs of the components, considering the hours of operation and power profiles they are subjected to.	Validates EMS performance through long-term simulation in MATLAB-Simulink	Requires backup storage systems due to the fluctuating nature of RE sources	[189]
Wind/DG/battery/converter	Non-linear programming	The application and investigation communities have shown a strong interest in controlling electrical power and energy while utilizing distributed generating systems.	Enhancing technical objectives such as stability, flexibility, reliability, and quality.	The short life cycle of lead–acid batteries and low energy density	[190]
PV/micro-gas turbine linked to the power grid	Linear programming	The proposed energy management strategy ensures reliable hourly forecasting of plant electric power, addressing PV power unpredictability and primary fuel reduction, replacing imbalances with grid substitution.	Optimizes operation cost of hybrid systems for standalone applications	The paper ignores degradation in the sizing of the batteries, resulting in less investment	[191]
PV/DG/ battery/FC/ MT	Non-linear programming	The study employs the branch and reduce optimization navigator (BARON) method, a global optimization approach.	Enhances supply reliability, reduces GHG emissions, and generation costs.	Neglects important factors, such as system reliability, power quality, and grid stability	[192]
PV/diesel/ battery	Non-linear programming	This study proposes a methodology for determining optimal current in transmission and distribution lines, considering load demand, to minimize power losses during peak demand hours.	Challenges discussed for widespread application of energy and power management systems	Systems require significant computational resources	[193]
PV/ultra-capacitor /FC	Linear programming using Simulink	The EM strategy aims to continuously meet load requirements by utilizing PV energy, generating hydrogen, and directing excess energy to the ultra-capacitor when the system is full.	Validates system performance through simulations with solar data and load profile	Low efficiency, as well as complex and expensive system, compared to a single-source solution	[194]
PV/wind/ battery	Linear programming	The energy management strategy considers renewable generation forecasts, component in/out power, and battery state, with grid compensating imbalances and battery adjusting daily fluctuations.	Validates strategy at the microgrid installed in the Renewable Energy Laboratory	Simulation and experimental validation may not capture all possible scenarios	[195]
MMGs with RE and EVs	Linear programming	Proposes a tri-layer framework with iteration-free scheduling for robust energy management using a global positive power factor (GPPF).	Enhances computational efficiency and robustness for MMGs with high EV integration	Focuses on short-term operational efficiency; lacks discussion on long-term energy strategy and infrastructure scalability	[196]
Hybrid AC/DC networked microgrids (NMGs)	Linear stochastic energy management	Proposes a convex model for bi-directional converters (BDCs) using least squares approximation to enhance computational efficiency.	Enables high-efficiency power exchange in hybrid AC/DC microgrids with reduced complexity	Limited focus on real-time constraints and dynamic power behavior	[197]

summarizes numerous works on linear and non-linear programming based on energy management systems. Using Mixed-Integer Linear Programming (MILP), Ahmad et al. [186] offered a technological and economic approach to optimizing a hybrid system. In [187], presented a robust method for a standalone hybrid system that makes use of a predictive control model. Cost, energy use, and gas emissions from diesel production in the hybrid system are all reduced because the model incorporates multi-objective optimization using MILP. Dursun and Kilic [188] compared the efficacy of three different power management techniques for a PV/wind/PEMFC standalone hybrid power system, in which the PV and wind sources provide primary power and the PEMFC provides backup power.

5.2. Metaheuristic-Based Energy Management with Grid-Connected Application

The generation, transmission, and use of electricity have all seen significant shifts during the last several decades. Most nations have made the development of a sustainable, fossil-fuel-free energy infrastructure a top priority. This objective [197] refers to smart networks that include renewable energy sources. The effects of incorporating internal combustion engines and gas turbines into a hybrid system consisting of solar modules were investigated by Das et al. [198]. The system was optimized via a multi-objective genetic algorithm, utilizing electric and thermal methods to monitor load and meet heating and cooling power needs. The GEP issue was given a new multi-objective framework by Gitizadeh et al. [199]. The suggested MOGEP model incorporates the cost function, pollution, and fuel price risk as competing objective functions and uses the MILP formulation. Nivedha et al. [200] looked at how a microgrid with wind power production, fuel cells, an electrolyzer, and a diesel generator all worked together. Fuel cells maintain energy balance in diesel generators, reducing operational expenses and achieving 70% cost savings through Particle Swarm Optimization for peak demand.

5.3. Dynamic Programming-Based Energy Management and Multi-Agent Systems

An energy management system for a wind/solar/storage hybrid power system at laboratory size was suggested by Marabet et al. [201]. The energy management system utilizes real-time control and data capture to optimize system performance by regulating and monitoring power production, storage, and load components according to predetermined rules.

Shuai et al. [202] presented an energy management system for microgrids utilizing dynamic programming and mixed-integer non-linear programming optimization, utilizing offline data for energy flow and battery storage limits. Zhuo [203] proposed a method for managing energy using dynamic programming, renewable power sources, and storage batteries, aiming to maximize profits and minimize expenditures in a decentralized energy market.

The energy management system of a hybrid renewable energy-based power system was presented by Anvari-Moghaddam et al. [204], with an emphasis on distributed generation and demand response coordination, reducing operating costs, and meeting customer expectations using HTTP communication infrastructure.

5.4. Energy Management Based on Artificial Intelligence Techniques

Genetic algorithms (GAs), differential evolution (DE), artificial neural networks (ANNs), Wavelength Transformation (WT), fuzzy logic controllers (FLCs), and Neuro-Fuzzy Systems (NFSs) are only some of the intelligent approaches that have been substantial research in the past few years for their application to energy management in hybrid systems. A brief review of the AI approach is presented by Lanre and Saad [181]. In this section, the notable ones and the most recent ones will be reviewed. Using a game-theoretic, decentralized approach, Mondal et al. [205] suggested an energy management model for a smart microgrid. In this plan, the microgrid opts for a method that will allow it to reap the greatest financial and efficiency rewards.

An energy management system for an MG was presented by Prathyush and Jasmn [206], and it makes use of a fuzzy logic controller with 25 different rules. The primary goal is to maintain the battery’s state of charge while minimizing the grid power variation. A voting-based smart energy management system (VSEMS) concept for a small community’s grid-connected solar, wind, and biomass hybrid energy system is proposed in [207]. The author shows how a simulation model of the planned HES is run using hourly data from the location’s solar and wind resources and biomass to achieve the load requirements described earlier. In [208], the authors describe the design and implementation of a fuzzy control-based EMS for DC microgrid systems. LabVIEW was used to create the integrated monitoring EMS, while the MATLAB/Simulink environment was used for modeling, control, and analysis of the decentralized energy sources and energy storage devices. Using energy storage components, Jia et al. [209] developed an adaptive intelligence approach for energy management. Volatility in the load must be kept to a minimum as a result of unknowns in renewable energy production. Ultracapacitors and other forms of energy storage regulate the load profile.

Table 19. The synopsis of research on artificial intelligence based on energy management system.

System Configuration	EMS Approach	Remarks	Advantages	Limitations	Ref.
PV/battery/ fuel cell	Model predictive control (MPC) logic	The control strategy, based on weather forecasts, aims to optimize renewable source usage while enhancing house comfort.	Minimizing energy costs while optimizing renewable source use	Implementation of MPC requires a validated numerical model of the entire system	[210]
PV/micro-hydro power/wind/ DG/battery	Theory of non-cooperative games	The proposed distributed energy management system architecture utilizes multi-agents to control each energy source or load in the microgrid system.	The MAS-based system allows for easy integration of new DER units or loads	Management systems find it difficult to satisfy tolerance and adaptability etc. criteria	[211]
Thermal (PV/T)/PV	Solar energy system modeling	Artificial neural networks (ANNs) are utilized as a system-based technique for forecasting and optimizing the performance of various solar energy devices.	Providing a detailed insight into artificial neural network types of applications	ANN is time-consuming and requires an iterative training process	[212]
PV/standalone wind/ bioethanol	State-machine methodology	In times of limited solar and wind radiation, fuel cells prioritize wind energy, utilizing energy management techniques to ensure the system’s functionality.	Offering a novel concept of multiple power sources to enhance efficiency	The system requires a backup system and energy storage to ensure reliability.	[213]
Standalone PV/diesel/ wind/battery	Biogeography, genetic algorithms, and Particle Swarm Optimization	Three energy management techniques were examined in the study: load following, peak shaving, and cycle charging. Cycle charging was shown to be the most successful.	Develop a cost-effective solution as well as a reliable and economically feasible alternative to conventional system	Does not address the acceptance and affordability of implementing the hybrid RE based system in the rural area	[214]
Microgrid linked to the main grid	A new method of double-layer coordinated control	The proposed approach uses a scheduling layer for forecasting data and a dispatch layer for real-time energy provision, allowing for the replacement of deficits between real and forecasting.	The paper optimizes power flow in a microgrid, reducing costs significantly	The paper lacks a detailed analysis of the economic impacts and feasibility	[215]
PV/battery/wind/FC	Real-time prediction is achieved by using neural networks.	The proposed strategy involved forecasting renewable sources and loads, with the power management system continuously updated to address decision time intervals and hardware sensor lags.	PMS successfully controls the switching and allows the lab-scale standalone power system to fulfill demand during both transient and steady-state stages.	The hybrid energy system in the paper may not accurately represent the fluctuations of real sustainable power sources	[216]

presents an overview of research on artificial intelligence based on EMS.

5.5. Energy Management Using Predictive Control Methods

The predictive robust control described by García et al. [217] maybe used with a hybrid system operating independently. Mixed integer programming was used in the management model. Loads, batteries, and PV and wind generators make up the system. Model predictive control (MPC) was introduced by Zhang et al. [218] as a means of combining decentralized and renewable power sources. The goal of the model is to decrease generating costs and demand limitations.

An overview of current software-based algorithm-based EMS research is demonstrated in

Table 20. A summary of current energy management system investigations based on software algorithms.

System Configuration	EMS Approach	Remarks	Practical Implications	Limitations	Ref.
Different configurations for standalone systems	HOMER	HOMER replicates system functionality through annual energy balance computations, determining component sizes to minimize COE production and efficiently utilize available and demand energies.	Enhances dynamic response, stability, and voltage control during climatic changes	Higher initial implementation costs	[220,221,222,223]
Grid-connected-various arrangements	HOMER	HOMER replicates system functioning through annual hourly energy balance computations, calculating energy flow between components based on accessible and demand energies.	Grid-connected systems can be profitable by selling excess power	High capital cost of PV panels	[224,225,226]
Standalone fuel cell, battery, or PV	TRNSYS software	Three EM techniques were studied, each examining surplus energy and making decisions on hydrogen creation, battery charging, or both, following the strategies suggested.	Offers insights into unit-sizing approaches for hybrid renewable energy-based systems	Fuel cells have lower efficiency compared to other energy conversion technologies	[227]

. A mathematical model of a self-contained hybrid system’s intelligent loads and energy management was provided by Solanki et al. [219]. Neural networks are used to model loads. Predictive control, a strategy used in energy management, optimizes power dispatch by factoring in weather conditions and other modifiable demands.

A multi-step predictive control model for an MG was presented by Oh et al. [228] to operate across a time horizon/period of 180 min, with each step lasting 15 min. This includes both traditional and renewable power sources, various forms of energy storage, and loads of varying criticality. In the formulation of the objective function for this study, we incorporated key variables, including fuel consumption, the state of charge (SOC) of the batteries, and the reduction of renewable energy usage to minimize the extent of load shedding.

Table 21. The advantages and disadvantages of all the energy management systems are given below. Demonstrates effective dampening [229,230,231,232].

Energy Management Systems	Advantage	Disadvantage
Conventional approach	Demonstrates effective dampening and quick recovery from various disturbances. Easy to control as a simple control technique is used. Installation cost is low. Increased storage system flexibility. Capable of controlling unidirectional and bi-directional DC-DC converters with high efficiency. Can protect the battery from overcharging without using the damp load. Capable of supplying load without interruption.	Low power quality. Low efficiency. Increase the operating cost. Implementation of a multi-energy system is difficult. Manual monitoring and restoration. Only one-way communication is possible. Limited control. Limited customer choice. More failures and blackouts occur. Centralized generation system.
Artificial intelligent approach	Efficiency is high. Good power quality. Significantly reduce the operating cost. Shows a good dynamic performance. Reduces overall system costs, and fuel uses and unmet load. Each of the energy sources can be controlled individually and efficiently. The forecasting can be performed both for the renewable energy sources and the loads. This method can be used to extend the battery life and improve its performance. Controllers like fuzzy controllers and model predictive controllers (MPCs) can be used in this approach to reduce the net present value of the entire hybrid system. Reduce the energy-utilization cost, as well as CO2 emissions. Storage systems such as battery charging can be controlled more efficiently according to the load demand. Implementation of a multi-energy system is possible.	Less adaptable to load fluctuation. Execution time is long. Real-time control is not possible.
Real-time/online approach	Relationship between consumer and utility has been possible. It improves grid reliability, efficiency, and security. Effective energy efficiency, maximizing utilization, cost reduction, and emission control have been possible. Self-monitoring. Optimize asset utilization and self-healing. Increased power reliability and power quality. Reduce greenhouse gas emissions. Distributed generation system. Two-way communication is possible. Advanced meter infrastructure.	Biggest concern is security and privacy as it can be hacked sometimes. Costly, as an advanced sensor is needed. Expensive for customers, as operating costs are high. Many sophisticated components are used here. High-speed internet is needed. If the system fails, the whole network is shut down. Experts are needed for controlling and operating the management systems.

displays the benefits and drawbacks of each energy management method.

6. Discussion

This paper provides a detailed and insightful summary of the methods that researchers have been employing for years to enhance HRSPSS, irrespective of their connectivity to the grid. The evaluation of this research highlights that conventional methods can resolve multi-objective (MO) issues, deliver faster solutions, and potentially optimize system sizing more effectively than commercial tools. However, they lack flexibility, and it can be challenging to adapt them to changing conditions and new technologies. Additionally, this paper reviews various optimization techniques and software that have been utilized in several HRSPSS configurations, such as photovoltaic (PV) and wind systems, over the past two years. The analysis reveals that hybrid approaches, which combine the strengths of classical and advanced optimization techniques, effectively reduce processing times and enhance system performance. Despite these benefits, the main challenge for hybrid methods lies in the complexity of algorithm development and system design. Overall, the areas of system sizing, optimization, control, and power management within HRSPSS are predominantly served by mixed heuristic (MH) methods and HOMER software, underscoring their critical role in advancing the field.

6.1. Unit Sizing

This section discusses various methods for sizing a hybrid renewable and sustainable power supply system, which is a crucial step in determining the capacity of the generators in such a system. Effective sizing methods can significantly influence the efficiency and sustainability of the system. This section discusses different methods, such as the following:

• Conventional methods: These methods encompass analytical and probabilistic approaches that are straightforward and well-understood. While they are easy to implement and interpret, their adaptability to complex or highly variable conditions is limited. They are best suited for small-scale or less complex systems where environmental conditions are stable.
• AI techniques: This category includes dynamic sizing options like neural networks and genetic algorithms, which are highly adaptable and capable of handling complex system dynamics. Despite requiring significant computational resources and sophisticated data handling, they are ideal for complex, large-scale systems or systems operating in highly variable environments.
• Software tools (e.g., HOMER and RETScreen): These tools provide a comprehensive approach to sizing by integrating multiple factors, including economic considerations and environmental impact. They offer a holistic analysis backed by robust databases but may require customization for non-standard scenarios. They are best suited for comprehensive feasibility studies and initial system design where multiple variables must be considered.

Among the methods for sizing power supply systems, AI techniques such as neural networks and genetic algorithms are the most robust, offering high adaptability for complex and large-scale systems. In contrast, conventional methods, although easy to implement, are less effective for dynamic or extensive applications due to their limited adaptability. Software tools like HOMER and RETScreen are best suited for comprehensive feasibility studies and initial designs, providing detailed integration of economic and environmental factors, but may require customization for unique scenarios.

6.2. System Optimization

An extensive summary of the many optimization techniques applied to the design and analysis of HRSPSS is provided in this section:

• Classical methods: Including linear and non-linear programming, these methods are efficient and straightforward in scenarios with well-defined objectives and constraints. However, they are not suitable for non-linear or complex interactions within hybrid systems and are best used for traditional energy systems with minimal integration of renewable sources.
• Hybrid methods: These methods combine classical and AI techniques to optimize both efficiency and computational resource usage. They balance accuracy and computational efficiency but can be complex to configure initially. They are suitable for medium-to-large-scale systems requiring robust optimization under varied conditions.
• AI-based methods: Techniques such as machine-learning models that predict and optimize real-time data inputs are highly effective in dynamic environments with many interacting variables. Despite their high computational cost and complexity, they are best suited for advanced systems with high variability in input conditions and operational demands.

Among these optimization techniques, AI-based methods are the most effective for HRSPSS due to their ability to handle dynamic environments and complex interactions with high variability. On the other hand, classical methods are the least effective, as they struggle with the non-linear complexities and the integration of renewable energy sources, making them suitable only for more traditional and less complex systems.

6.3. Control System

An extensive summary of the various control paradigms utilized in HRSPSS is given in this section:

• Centralized control: Simplifies management but increases risk due to single points of failure. Easier to manage and implement, these systems are vulnerable to system-wide failures if the central node fails, making them suitable for smaller or less complex systems where central oversight is feasible.
• Distributed control: Enhances system reliability by distributing control functions across multiple nodes. While increasing system resilience and reliability, these systems require more complex communication needs and system management. They are best suited for large, complex systems or systems spread over a large geographic area.
• Hybrid control: Combines elements of both centralized and distributed systems to optimize both control and reliability. Balances ease of management with system resilience but may still carry complexities associated with distributed systems. Suitable for systems that require robust control but also need to mitigate risks of centralized control failure.

Among these control paradigms, hybrid control emerges as the best method for HRSPSS due to its ability to balance ease of management with enhanced system resilience, effectively mitigating risks associated with centralized control failures. Conversely, centralized control is the least effective, primarily due to its vulnerability to system-wide failures from single points of failure, making it less suitable for complex or geographically widespread systems.

6.4. Energy Management System

This study provides an in-depth analysis of the many energy management system (EMS) techniques applied to HRSPSS. It encompasses a wide range of methodologies, such as the following:

• Conventional methods: Typically handle predictable load and generation scenarios and are well-understood and straightforward to implement. However, they lack flexibility in response to renewable variability and are best suited for systems with stable demand and supply patterns.
• AI and real-time systems: Use predictive analytics to manage and optimize power flow dynamically. Capable of preemptively adjusting to changes in load and generation, enhancing efficiency, these systems require advanced technology and are more expensive to implement. They are ideal for complex systems where load and generation are highly variable and difficult to predict.

Among these energy management system (EMS) techniques, AI and real-time systems stand out as the best methods due to their ability to dynamically manage and optimize power flow, adapting effectively to changes in load and generation. This capability is crucial for complex systems with high variability, enhancing overall efficiency. Conversely, conventional methods are the least effective for HRSPSS, as their lack of flexibility in responding to renewable variability makes them poorly suited to environments where demand-and-supply patterns frequently change.

7. Challenges and Future Research Vision of HRPSS

It has been observed that, in the last three decades, an increased number of researchers have been dealing with HRPSS, but still there are certain research challenges regarding their optimal use and efficiency. In this section, a limited list of obstacles experienced by the researchers and the future vision of the research topic areas are presented.

7.1. Challenges

HRSPSS encounters a range of operational and environmental challenges that can impede their efficiency and widespread adoption. As these systems become increasingly integral to the global energy infrastructure, identifying and understanding these challenges is crucial. This section outlines the primary obstacles that HRSPSS must overcome to achieve their full potential, such as the following:

• Integration and intermittency: The variability and unpredictability of renewable energy sources, such as solar and wind, remain a major challenge. Effective energy storage solutions and control mechanisms are required to balance supply and demand, especially during peak and off-peak hours.
• Energy storage limitations: Existing storage technologies, such as batteries and hydrogen-based solutions, face issues related to efficiency, cost, and lifespan. These constraints limit the reliability and scalability of HRSPSS.
• Economic viability: The high initial costs associated with HRSPSS, including component procurement, system installation, and maintenance, often deter investment. Moreover, the long payback periods make these systems less attractive in low-income regions.
• Technological compatibility: Ensuring seamless integration of diverse renewable energy sources into a unified system is technologically complex. Compatibility issues among components and subsystems often lead to inefficiencies.
• Regulatory and policy frameworks: The absence of robust regulatory frameworks and incentive mechanisms in many regions impedes the adoption of HRSPSS. Policies that promote renewable integration into the energy grid are essential for large-scale deployment.
• Environmental concerns: The production and disposal of renewable energy components, including solar panels and batteries, raise environmental concerns that must be addressed through sustainable design and recycling mechanisms.

7.2. Future Research Vision

Navigating the challenges posed by HRSPSS highlights the need for strategic, innovative research to evolve and enhance their efficacy and sustainability. This section proposes a vision for future research aimed at resolving these challenges and paving the way for more robust, efficient, and sustainable hybrid systems, such as those listed below.

7.2.1. Integration of Advanced Optimization Techniques

• AI-driven hybrid models: Explore the potential of integrating AI-based optimization models, such as hybrid machine-learning frameworks or evolutionary algorithms, to enhance the precision of HRSPSS in sizing, optimization, and control. These advanced computational models can learn from historical data and improve decision-making processes, optimizing energy flow and resource allocation.
• Quantum computing applications: Investigate the application of quantum computing for solving complex optimization problems in HRSPSS, enabling faster and more accurate decision-making. Quantum algorithms could potentially solve optimization problems more efficiently than classical computers, offering significant advantages in terms of speed and accuracy.
• Decentralized algorithms: Emphasize the development of decentralized optimization algorithms suitable for distributed systems to reduce computational overhead. These algorithms can facilitate the effective management of distributed energy resources, enhancing system flexibility and reducing reliance on centralized control.

7.2.2. Enhanced Control Strategies

• Multi-Agent Control Systems: Develop sophisticated control strategies incorporating Multi-Agent Systems (MASs) for seamless coordination between various renewable energy sources and storage units in HRSPSS. This approach helps to manage interactions and energy flows within the system dynamically, improving efficiency and reliability.
• Adaptive control mechanisms: Focus on adaptive control systems that can dynamically respond to real-time changes in load demand and energy production within HRSPSS. These systems adjust operational parameters on-the-fly, enhancing system responsiveness and stability.
• Resilience-oriented control: Design control strategies to improve the resilience of HRSPSS against natural disasters and cybersecurity threats. These strategies are crucial for maintaining continuous operation and ensuring energy security in adverse conditions.

7.2.3. Innovative Energy Storage Solutions

• Hybrid storage technologies: Explore combining traditional battery systems with advanced storage solutions like supercapacitors and hydrogen fuel cells to improve energy density and reliability in HRSPSS. This hybrid approach can enhance the system’s ability to handle peak loads and provide backup power during outages.
• Life-cycle analysis: Conduct comprehensive life-cycle analyses for emerging storage technologies to evaluate their economic and environmental viability within HRSPSS. Understanding the full life-cycle impacts helps in selecting the most sustainable and cost-effective storage solutions.
• Decentralized storage management: Propose frameworks for managing distributed energy storage systems within HRSPSS to optimize local consumption and reduce dependency on centralized grids. This strategy supports energy self-sufficiency and enhances system sustainability.

7.2.4. Policy and Socioeconomic Considerations

• Energy access and equity: Research the impact of HRSPSS on underserved regions, emphasizing cost reduction and scalability for remote areas. This research can help in formulating strategies to extend reliable and affordable energy access to all communities.
• Policy frameworks: Develop adaptive policy frameworks that encourage the integration of hybrid systems with existing energy infrastructures. These policies should be flexible to accommodate ongoing technological and regulatory changes.
• Community engagement: Promote participatory approaches involving local communities in the design and implementation of HRSPSS to ensure sustainability and acceptance. Engaging communities can also facilitate the adaptation of technologies to meet local needs and preferences.

7.2.5. Grid-Integration and Smart Technologies

• Bi-directional energy flow: Develop systems capable of managing energy flow in both directions between the grid and HRSPSS. This enhances the ability to balance supply and demand dynamically, improving grid stability and reducing energy wastage.
• IoT and smart grid technologies: Integrate advanced sensor technologies and smart grid systems to enable real-time monitoring and management of HRSPSS components. This integration can lead to better system performance and proactive maintenance scheduling.
• Blockchain for energy management: Explore the potential of blockchain technology to create transparent, secure, and efficient platforms for energy trading within decentralized HRSPSS. This could facilitate peer-to-peer energy transactions and more autonomous management of energy distribution.

7.2.6. Emerging Hybrid Configurations

• Hybrid offshore systems: Investigate the feasibility of offshore HRSPSS that harness wind, wave, and tidal energy resources. This research should focus on the integration challenges and potential energy yield of combining multiple types of marine-based energy generation technologies, which could lead to more consistent and reliable energy production offshore.
• Urban HRSPSS: Study the integration of HRSPSS within urban environments, addressing specific challenges, such as limited space, urban regulations, and noise concerns. The focus would be on designing systems that can operate efficiently within the constraints of urban infrastructure and contribute to the reduction of urban carbon footprints.
• Agrivoltaics: Explore the concept of agrivoltaics within the context of HRSPSS, combining agricultural use with solar energy production on the same land. Research should aim at optimizing land use for dual purposes—enhancing energy production while maintaining or even increasing agricultural productivity.

7.2.7. Environmental and Climate Adaptation

• Climate-resilient design: Develop HRSPSS designs that are capable of withstanding extreme weather conditions without losing operational efficiency. This involves incorporating materials, components, and system layouts that are durable and adaptable to changing climate patterns, ensuring system robustness and longevity.
• Carbon footprint reduction: Advance technologies and methodologies that minimize the carbon footprint of HRSPSS throughout their life cycle—from production through operational phases. This includes improving energy efficiency, utilizing low-carbon materials, and optimizing operational strategies to reduce overall emissions.
• Circular economy models: Implement circular-economy principles in the design, operation, and decommissioning of HRSPSS components. This approach focuses on the reuse, recycling, and repurposing of system components and materials to minimize waste and environmental impact, promoting a sustainable life cycle for HRSPSS materials and technologies.

8. Conclusions

This review paper provides a comprehensive overview of HRSPSS, focusing on various critical aspects such as configurations, unit sizing, optimization, modeling of components, control strategies, and power management. Additionally, it identifies key challenges and outlines directions for future research. The major findings of this study are as follows:

• The study evaluates various tools and methods, such as HOMER and RETScreen, for sizing, optimizing, controlling, and managing HRSPSS. The findings show that these tools are instrumental for accurate system analysis, enabling efficient design and development by integrating both traditional and AI-based techniques.
• The study assesses tools like HOMER and RETScreen, noting their ease of use for modeling but highlighting their limitations in dynamic scenarios. AI techniques (e.g., ANN and GA) offer higher accuracy but require significant computational resources. Integrating these methods enhances system efficiency, especially in complex setups.
• Classical methods are suitable for simple systems but fall short in complex configurations. AI techniques (e.g., ANN, GA, and fuzzy logic) are effective for real-time optimization but are resource intensive. Hybrid methods balance efficiency and accuracy, making them ideal for complex, large-scale systems.
• Combining classical and AI methods (e.g., genetic algorithms with fuzzy logic) improves performance and reduces processing times. These approaches are effective for dynamic systems but involve complex design requirements.
• Traditional methods are recommended for stable, small-scale systems, while AI-based and hybrid methods suit large, complex systems. Software tools like HOMER are useful for initial studies, especially when combined with AI for improved flexibility and response.

In general, this review paper offers critical insights into the control and management of HRSPSS, highlighting the effectiveness of centralized, distributed, and hybrid approaches for enhancing system efficiency, reducing failure risks, and extending lifespan. By emphasizing techno-economic optimization, the paper guides stakeholders through advanced management strategies, including AI-based algorithms like fuzzy logic and PSO, as well as commercial tools such as HOMER, to configure and analyze HRSPSS effectively. It serves as a valuable resource for utility companies, policymakers, and researchers by providing evidence-based strategies for integrating renewable energy sources, improving grid stability, and minimizing costs. Governments and private entities can leverage these insights to support and expand HRSPSS implementation, addressing environmental and economic challenges while enhancing energy access and development in remote regions. Overall, this review fosters innovation and strategic planning in the power sector, contributing to the global shift toward sustainable energy solutions.