A Literature Review on Strategic, Tactical, and Operational Perspectives in EV Charging Station Planning and Scheduling

Abstract

Before the onset of global warming concerns, the idea of manufacturing electric vehicles on a large scale was not widely considered. However, electric vehicles offer several advantages that have garnered attention. They are environmentally friendly, with simpler drive systems compared to traditional fossil fuel vehicles. Additionally, electric vehicles are highly efficient, with an efficiency of around 90%, in contrast to fossil fuel vehicles, which have an efficiency of about 30% to 35%. The higher energy efficiency of electric vehicles contributes to lower operational costs, which, alongside regulatory incentives and shifting consumer preferences, has increased their strategic importance for many vehicle manufacturers. In this paper, we present a thematic literature review on electric vehicles charging station location planning and scheduling. A systematic literature review across various data sources in the area yielded ninety five research papers for the final review. The research results were analyzed thematically, and three key directions were identified, namely charging station deployment and placement, optimal allocation and scheduling of EV parking lots, and V2G and smart charging systems as the top three themes. Each theme was further investigated to identify key topics, ongoing works, and future trends. It has been found that optimization methods followed by simulation and multi-criteria decision-making are most commonly used for EV infrastructure planning. A multistakeholder perspective is often adopted in these decisions to minimize costs and address the range anxiety of users. The future trend is towards the integration of renewable energy in smart grids, uncertainty modeling of user demand, and use of artificial intelligence for service quality improvement.

1. Introduction

The history of electric vehicles (EVs), with a particular focus on passenger cars, dates back to the early 19th century, when inventors in Hungary, the Netherlands, and the United States began experimenting with battery-powered transportation. By the late 1800s, EVs had gained popularity due to their quiet operation and ease of use compared to gasoline-powered cars, which were noisy and required manual gear shifting. In fact, by the early 20th century, electric cars accounted for a significant portion of the USA automobile market. However, the mass production of the Ford Model T in 1908, combined with the discovery of vast petroleum reserves and the development of better road infrastructure, led to the decline of EVs in favor of internal combustion engine vehicles [1,2].

Interest in EVs resurged in the late 20th century, driven by growing concerns over air pollution, oil dependency, and climate change. While early efforts such as the GM EV1 in the 1990s faced challenges related to cost and infrastructure, the 2010s and 2020s have seen transformative progress. A major catalyst in this shift is the rise of Tesla, which introduced the Roadster in 2008 and later revolutionized the EV market with the Model S, the Model 3, and a vertically integrated approach to battery production and charging infrastructure. Tesla’s innovations in performance, range, and autonomous driving features helped reshape public perception of EVs and spurred competition across the automotive industry [3,4]. Advances in lithium-ion battery technology, reductions in battery costs, and supportive government policies have further accelerated EV adoption globally. Recent studies have highlighted the role of EVs in achieving net-zero emissions targets, with countries like Norway and China leading in market penetration [5,6]. Moreover, the integration of EVs with smart grids and renewable energy sources is becoming a focal point of research, aiming to enhance grid stability and sustainability [7,8]. The increasing focus on clean energy in recent years has led to a widespread embrace of electric passenger cars as a key solution to reducing dependence on fossil fuels in transportation. While this review focuses on battery electric vehicles (BEVs), it is important to acknowledge the parallel development of hydrogen fuel cell electric vehicles (FCEVs). FCEVs offer advantages such as faster refueling times and longer driving ranges, particularly suited for heavy-duty and long-haul applications. However, their adoption remains limited due to high infrastructure costs, lower energy efficiency compared to BEVs, and limited hydrogen refueling networks. In contrast, BEVs benefit from a more mature charging infrastructure and higher energy conversion efficiency, making them the dominant technology in current urban and passenger transport planning. Future research may explore integrated infrastructure planning that considers both BEV and FCEV technologies, especially in regions investing in hydrogen ecosystems [3,9]. This shift has been enthusiastically supported by both individuals and governments around the world, leading to a dramatic increase in the number of electric vehicles on the roads. According to the International Energy Agency’s Global EV Outlook 2024 [10], global electric car sales exceeded 15 million units in 2024, marking a 10% increase from 2023. The report highlights that China, Europe, and the United States accounted for over 90% of global EV sales, with emerging markets such as India and Southeast Asia showing rapid growth. For instance, India’s EV sales doubled compared to 2022, driven by government incentives and infrastructure expansion. These trends underscore the accelerating global shift toward electrified transportation, supported by falling battery costs and stronger policy frameworks. Similar projections from BloombergNEF’s 2025 EV Outlook [11] suggest that EVs could make up over 30% of new car sales globally by 2030, reinforcing the long-term trajectory toward decarbonization in the transport sector.

This review specifically focuses on the planning and scheduling of charging infrastructure for private battery electric vehicles (BEVs), particularly passenger cars, which currently represent the majority of EV adoption globally. Public transit and commercial EV applications are beyond the scope of this study. Electric vehicles offer the advantage of being able to charge from electricity generated by various sources, including wind, solar, nuclear, water, and biofuels [12]. This diversity of sources helps reduce dependence on oil and gasoline, leading to less imported fuel and lower costs associated with it. The adoption of electric vehicles in various transportation sectors, including general and goods transportation, is increasing, prompting numerous studies on their design and performance optimization.

While most research and infrastructure development efforts for electric vehicles (EVs) have been concentrated in developed regions like Europe, the U.S., and China, emerging economies are becoming increasingly influential in the global shift toward electric mobility. Countries such as India have launched major programs like the FAME II initiative to accelerate EV adoption, especially for two- and three-wheeled vehicles. In Southeast Asia, nations like Thailand and Indonesia are actively investing in local EV manufacturing and charging networks. Meanwhile, African countries including Kenya and Rwanda are exploring solar-powered charging solutions to overcome limitations in grid access. Although these regions face obstacles such as high initial investment costs, unreliable electricity supply, and limited financing options, they also present unique opportunities for innovation—particularly in integrating renewable energy and customizing solutions to local transportation needs. According to the World Bank, over half of the 20 developing countries it assessed could gain economic benefits from transitioning to electric mobility, especially where electricity is subsidized and fuel is heavily taxed [13,14].

From the perspective of EV owners, the lower fuel and operational costs of EVs compared to vehicles with internal combustion engines are significant advantages. This cost reduction is primarily attributed to the higher efficiency of electric engines in contrast to internal combustion engines. While internal combustion vehicles typically operate at an efficiency of 15–18%, electric vehicles boast efficiencies ranging from 60 to 70% [15]. Furthermore, Vehicle-to-Grid (V2G) technology presents an additional opportunity for EV owners to generate revenue by leveraging the energy stored in their batteries to exchange with the grid [16]. However, from the standpoint of the power supply network, the presence of EVs can lead to increased losses and costs within the entire system. Nevertheless, employing appropriate charging methods can effectively mitigate these negative effects. Controlled charging methods, for instance, have been shown to reduce system costs and peak demand by over 50% compared to uncontrolled charging [17,18]. This underscores the importance of implementing smart charging strategies to optimize the integration of EVs into the power grid while minimizing adverse impacts.

Despite these advantages, electric vehicles face limitations such as the lower energy density of batteries compared to fossil fuels, which affects vehicle range and charging infrastructure requirements [19,20]. Unique refueling equipment and a scarcity of charging stations further restrict their adoption. Additionally, energy consumption in electric vehicles is dependent on the vehicle’s load, which poses practical limitations. However, due to their positive impact on reducing air pollution, there is a growing effort to design and establish more charging stations. It is important to note that the reduced fuel consumption of electric vehicles leads to lower service costs and, consequently, higher customer satisfaction [21].

From the standpoint of the power supply network, the presence of EVs can lead to increased losses and costs within the entire system. Nevertheless, employing appropriate charging methods can effectively mitigate these negative effects. Controlled charging methods, for instance, have been shown to reduce system costs and peak demand by over 50% compared to uncontrolled charging [17,18]. Controlled charging, also known as smart charging, includes strategies such as time-of-use pricing, load shifting, and coordinated charging. These methods help align EV charging with grid capacity, reduce peak demand, and lower electricity costs. In contrast, uncontrolled charging—where EVs begin charging immediately upon connection—can lead to grid congestion and increased operational costs. Studies have shown that smart charging can significantly improve grid efficiency and reduce peak loads [22]. This underscores the importance of implementing smart charging strategies to optimize the integration of EVs into the power grid while minimizing adverse impacts. Currently, home charging is the most prevalent method, with private chargers outnumbering public ones by nearly ten to one. Most EV owners rely on charging their vehicles at home.

Contribution of This Review: This review advances the academic discourse by systematically synthesizing recent studies related to electric vehicle charging station planning and scheduling. Unlike previous reviews that focus narrowly on either location planning or charging technologies, this study integrates strategic, tactical, and operational perspectives—covering location models, scheduling algorithms, and emerging technologies such as V2G and AI-based optimization. It highlights methodological trends, identifies research gaps, and proposes a structured framework for future research. This comprehensive approach provides researchers and practitioners with a clearer understanding of the current landscape and actionable directions for advancing EV infrastructure planning.

In this paper, our objective is to conduct a comprehensive literature review on electric vehicles infrastructure planning with a focus on station location, charging scheduling, and parking lots sizing and allocation for smart grid purposes.

Despite the growing body of research, several challenges remain unresolved. These include the lack of standardized methodologies for multi-objective optimization, limited integration of real-time data in scheduling models, and insufficient attention to infrastructure planning in emerging markets. Moreover, existing reviews often focus narrowly on either location planning or charging technologies, without offering a comprehensive thematic synthesis. This review addresses these gaps by categorizing and analyzing the recent literature across strategic, tactical, and operational dimensions, thereby offering a holistic view of the current landscape and future directions.

2. Background

EV charging is the process of replenishing an electric vehicle’s battery by connecting it to an electric power source [23]. Since electric vehicles rely entirely on electricity for power, this process is vital to their operation. Below are the key aspects of EV charging:

2.1. Charging Levels

The charging levels for electric vehicles can be categorized into the following [24,25,26]:

• Level 1 Charging: Level 1 is the slowest and simplest charging method, utilizing a standard 120 volt household outlet. It is typically used for overnight charging at home. The charging rate for Level 1 is around 2 to 5 miles of range per hour.
• Level 2 Charging: Level 2 chargers operate at 240 volts and are commonly found in residential garages, workplaces, and public charging stations. They offer much faster charging than Level 1, providing approximately 10 to 30 miles of range per hour of charging.
• Level 3 Charging (DC Fast Charging): Also known as rapid charging or DC fast charging, Level 3 chargers use high-voltage Direct Current (DC) to charge vehicles significantly faster than Levels 1 and 2. These chargers are often located along highways and major routes, allowing EVs to gain up to 100 miles of range in just 20–30 min.

Table 1. Summary of comparison of charging levels.

Charging	Voltage	Current	Useful	Maximum	Charging	Connector
Level		Type	Type	Output	Time	Types
Level 1	120 V	AC	1.4 kW	1.9 kW	12 h	J1772
Level 2	208–240 V	AC	7.2 kW	19.2 kW	3 h	J1772
Level 3	100–450 V	DC	50 kW	150 kW	20 min	J1772/CHAdeMO/Supercharger

presents a summary of various charging levels.

2.2. Charging Connectors

Charging connectors [27,28] vary depending on the region and manufacturer.

• Level 1 and Level 2 Connectors: The SAE J1772 EV plug is the most common connector used for Level 1 and Level 2 charging in North America. It is a standardized interface that ensures compatibility between EVs and charging stations. While Tesla vehicles use their own proprietary connector for Tesla Supercharger stations, they also come with an adapter to charge using the SAE J1772 plug, ensuring compatibility with other charging stations. SAE J1772 connectors are primarily for Level 1 and Level 2 charging, which are suited for slower, overnight charging.
• Level 3 Connectors: For rapid DC fast charging (Level 3), the most common connectors are CHAdeMO and the SAE Combo (also known as CCS or Combo Charging System). These two connectors are not interchangeable, meaning that a vehicle with a CHAdeMO port cannot use an SAE Combo plug, and vice versa. Tesla also has its own exclusive connector for its Supercharger stations, which is only compatible with Tesla vehicles. Although SAE J1772 specifies six charging levels, only three are widely used in North America: Level 1 (120 VAC), Level 2 (208–240 VAC), and fast charging (200–450 VDC). Tesla’s proprietary Supercharger network is another DC fast-charging system used exclusively for Tesla vehicles [29].

2.3. Wireless Charging of Electric Vehicles

Wireless charging, or wireless power transfer (WPT), is increasingly being explored as a convenient alternative to plug-in charging for electric vehicles. This includes both stationary systems—where vehicles charge while parked—and dynamic systems, which allow for charging while driving over embedded coils in the road. For example, pilot projects in Sweden and South Korea have demonstrated dynamic wireless charging for buses and trucks, showing its potential to reduce downtime and improve energy efficiency. Additionally, vehicle-to-vehicle (V2V) wireless charging is being researched as a way for EVs to share energy without relying on fixed infrastructure [30]. Wireless charging technology represents a significant advancement in the charging infrastructure for electric vehicles (EVs). Unlike traditional charging systems, which require physical connections between the vehicle and a charging station, wireless charging allows EVs to charge without the need for direct contact. This technology promises to address several limitations of conventional charging methods, such as time lost during charging, waiting in line, and energy wastage due to connection inefficiencies [31,32].

A typical EV wireless charging system includes the following steps:

• AC to DC Conversion: AC power is first converted into DC power using an AC to DC converter.
• Power Transmission: The DC power is then converted into a high-frequency signal, which drives the transmission signal through a compensating network.
• Safety Features: A high-frequency isolation transformer is used to protect the system by preventing insulation failure.
• Magnetic Field Induction: The transmitter coil generates an alternating magnetic field, which induces an AC voltage in the receiver coil.
• Rectification: The induced power is rectified to DC power to charge the vehicle’s battery [33,34].
• Wireless power transmission systems consist of components like rectifiers, power factor correctors, inverters, network compensators, and magnetic couplers (transmitter and receiver coils).

Additionally, topologies such as series–series, series–parallel, and LCL or LCC (inductor–capacitor–inductor or inductor–capacitor–capacitor) configurations are used to improve the system’s performance and efficiency [35]. The LCL topology offers the advantage of current independence from load conditions, while the LCC topology reduces size and cost by using a series capacitor with the coil.

Dynamic Electric Vehicle Charging (DEVC) is a promising solution that could address many of the challenges faced by electric vehicles. Unlike traditional charging or Static Electric Vehicle Charging (SEVC), which occurs when the vehicle is stationary, DEVC allows for wireless charging while the vehicle is in motion. This system would enable EVs to travel unlimited distances, continuously charging as they move along specially equipped roads.

Qualcomm, in collaboration with the French company Vedcom, has been researching this technology. They set up a 100 m test track in Versailles to evaluate the DEVC system. In these tests, vehicles were charged wirelessly while in motion, significantly reducing the need for charging stops and offering an efficient solution for long-distance travel.

3. Data Collection

The data for literature review was selected using online resources and research databases such as Web of Science, Proquest, Scopus, IEEE Explore, and Google Scholar. Figure 1 presents the study flow diagram using the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA). A preliminary search in the domain of electric vehicles yielded 400 articles. These articles were then narrowed down using keyword search of electric vehicles and station location planning, charging stations, charging scheduling, parking lot allocation, parking lot sizing, smart grids, Vehicle-to-Grid (V2G) [36], and distribution networks. This finally yielded 95 articles, which are the source for the review proposed in this paper. The yearwise distribution is presented in Figure 2. It can be seen that there is a major growth in the publication of articles in the last three years and the trend is growing. The sources are presented in Figure 3.

Among the collected literature, the majority of articles were published in peer-reviewed journals and conference proceedings. Notable journals include IEEE Access, Energies, Applied Energy, IEEE Transactions on Industrial Informatics, IEEE Transactions on Smart Grid, IEEE Transactions on Sustainable Energy, Journal of Energy Storage, Sustainable Energy, Grids and Networks, Transportation Research Part C: Emerging Technologies, World Electric Vehicle Journal, and IEEE Transactions on Vehicular Technology. Most conferences were organized by IEEE. While this list includes several high-impact (Q1) journals, it is not exhaustive; other leading journals such as Renewable and Sustainable Energy Reviews, Energy Conversion and Management, and Energy Policy are also highly influential in this domain and may be underrepresented due to the specific focus or timeframe of the dataset. Other sources in the dataset include dissertations, blogs, and technical reports.

A thematic analysis of these 95 papers yielded 10 key themes as shown in Figure 4. The top three themes that have received attention in the literature on electric vehicle charging infrastructure planning are charging station deployment and placement, optimal allocation and scheduling of EV parking lots, and V2G and smart charging systems. These three themes are the subject of a detailed investigation in subsequent sections in this paper.

4. Charging Station Deployment and Placement

Figure 5 presents the key themes in the area of charging station deployment and placement. It can be seen that there are five broad categories, namely location planning for EV charging stations, optimization models and techniques, urban traffic and demand considerations, GIS and advanced tools for planning, and a combination of charging techniques (slow and fast charging). These categories are explained in detail as follows:

4.1. Location Planning for EV Charging Stations

The first category of studies is dedicated to location planning for electric vehicle (EV) charging stations, which is a critical aspect of the deployment and optimization of EV infrastructure. These studies typically take a multistakeholder approach, focusing on minimizing costs for both EV users and service providers, while ensuring that the supply and demand of charging stations are balanced. The planning of these locations is often influenced by various factors, including the anticipated utilization of the charging stations, proximity to key public areas, and the presence of multimodal transport links. Celik and Ok [37] conducted a comprehensive literature review on EV charging stations’ location and capacity planning, focusing on optimization approaches to address the problem of selecting the most strategic locations for charging stations.

4.2. Optimization Models and Techniques

According to [37], station location planning can be categorized into three broad approaches:

• Mathematical models,
• Simulation models, and
• Meta-heuristic algorithms.

Various optimization methods are employed in EV charging station planning, each suited to different problem types. Linear and non-linear programming techniques are typically used when the problem structure is well-defined and continuous. LP is suitable for linear problems, while NLP handles more complex, non-linear constraints. MILP is often applied when binary decisions are involved, such as whether to build a station at a specific location. For more complex or large-scale problems, meta-heuristic algorithms like GA and PSO are preferred. GA is effective for multi-objective problems, while PSO is known for its fast convergence in dynamic environments. These methods are especially useful when traditional optimization becomes computationally intensive. Various optimization methods are employed in EV charging station planning, each suited to different problem types. Linear and non-linear programming techniques are typically used when the problem structure is well-defined and continuous. LP is suitable for linear problems, while NLP handles more complex, non-linear constraints. MILP is often applied when binary decisions are involved, such as whether to build a station at a specific location. For more complex or large-scale problems, meta-heuristic algorithms like GA and PSO are preferred. GA is effective for multi-objective problems, while PSO is known for its fast convergence in dynamic environments. These methods are especially useful when traditional optimization becomes computationally intensive [38,39,40]. Mathematical optimization models, including linear programming (LP), non-linear programming (NLP), integer programming (IP), and stochastic programming, have been widely used in EV station location planning. These models are often employed to maximize revenue and minimize costs. For instance, Hu et al. [41] and Hussain et al. [42] both investigated location planning using LP and NLP techniques to minimize costs associated with land prices and connection fees, while ensuring that stations are optimally located to maximize their utilization by EV owners. Moreover, Yang [43] applied stochastic programming to account for uncertain factors such as varying customer demand and future EV adoption rates. This approach adds flexibility to the location optimization process, making it more adaptive to changes in external conditions like urban development and the introduction of new EV models. Multicriteria models such as AHP, TOPSIS, and delphi techniques have also been reported by several authors [33,44,45]. Mortaz et al. [46] proposed a stochastic model for siting and sizing Vehicle-to-Grid (V2G) facilities within microgrids using large-scale Mixed-Integer Non-linear Programming.

Table 2. Comparison of optimization methods for EV charging station planning.

Model Type	Objective	Constraints	Suitable Scenarios	Strengths/Limitations
Linear Programming (LP)	Minimize cost	Resource limits, demand	Deterministic demand, urban areas	Simple, fast, but limited to linear relationships
Non-Linear Programming (NLP)	Minimize cost	Non-linear constraints, demand	Complex relationships, urban areas	Handles non-linear relationships but computationally intensive
Mixed-Integer Linear Programming (MILP)	Minimize cost	Binary decisions, resource limits	Urban areas, large-scale planning	Combines integer and linear programming but can be slow for large problems
Genetic Algorithm (GA)	Optimize multi-objective	Resource limits, demand	Complex, multi-objective scenarios	Flexible, handles non-linearities but may not find global optimum
Particle Swarm Optimization (PSO)	Optimize multi-objective	Resource limits, demand	Dynamic, uncertain environments	Fast convergence but can get stuck in local optima

provides an illustrative comparison of selected optimization methods frequently applied in EV charging station planning. While informative, it is not intended to be an exhaustive list of all possible approaches.

Simulation models are another valuable tool in EV station location planning. These models are particularly useful for addressing customer demand, EV usage, and charging/discharging behaviors in different scenarios. Key techniques used in simulation models include discrete-event simulation, agent-based modeling, and Monte Carlo simulations. Lam and Leung [47] focused on EV charging station placement using simulation techniques to capture the complexities of user behavior and charging needs. Elnozahy and Salama [48] investigated the impacts of Plug-in Hybrid Electric Vehicles (PHEVs) on residential distribution networks, focusing on how the charging behavior of PHEVs can affect grid stability using Monte Carlo simulation. Wang and Paranjpe [49] conducted a multi-agent-based simulation model to simulate how different levels of EV penetration impact grid stability and the overall charging process. Similarly, Xi et al. [50] explored simulation-optimization methods to design public EV charging infrastructure, optimizing the placement of stations while considering factors like peak demand and customer satisfaction. Yi et al. [51] took a similar approach by developing agent-based models for urban-scale charging station optimization.

Meta-heuristic algorithms have become increasingly popular in solving complex EV station location problems, particularly in large, uncertain, and dynamic environments. Techniques such as tabu search, Genetic Algorithms (GAs), Particle Swarm Optimization (PSO), simulated annealing, and ant colony optimization (ACO) are commonly applied to optimize station placement. For example, Lie et al. [52] proposed the use of PSO for EV charging station planning, where the algorithm optimizes station locations by evaluating the performance of various placements based on the total cost, including land prices, infrastructure costs, and expected demand. Similarly, You and Hsieh [53] developed a hybrid heuristic approach that combines Genetic Algorithms with other local search techniques to enhance the search for optimal locations in a complex, multi-objective environment. Wang and Lin [1] presented a study on multiple recharging station types, optimizing the placement of different types of charging stations (e.g., fast chargers, slow chargers) using a Genetic Algorithm.

Table 3. Comparison of optimization models and techniques for EV charging station planning in selected studies.

Ref.	Model		Focus Area	Advantages	Limitations	Algorithm	Simulation	Implication	Application Notes
	Objective Function/Formula	Type
[41]	$p(C,L)=\left\{\begin{array}{cc}{\displaystyle \frac{C_j}{{\sum }_{j=0}^k{C}_j}}·\left(1-{\displaystyle \frac{L_j}{{\sum }_{j=0}^k{L}_j}}\right)\hfill & \mathrm{if} k\ge 2\hfill \\ 1\hfill & \mathrm{if} k=1\hfill \\ 0\hfill & \mathrm{if} k=0\hfill \end{array}\right.$	–	Station location planning	Simple, fast computation	Limited real-world validation	Network Model of E-GPSR	Network Simulator 2 (NS-2)	Avoids routing through congested nodes	Suitable for small-scale urban networks with predictable demand
[42]	${min}_{n_i\in N, t\in T, {\tilde{p}}_i\in P}W{T}_i(n_i,t,{\tilde{p}}_i)$	FILP	User-centric scheduling	Captures user uncertainty; improves experience	Real-time implementation challenges	FISA	Java	Improves EV user charging experience	Applied in real time Java-based platforms
[33]	SCLP: ${\sum }_{j\in J}{y}_j$ MCLP: ${\sum }_{i\in I}{w}_i{x}_i$	Dubai strategic transportation model	EVCS equity–efficiency	Balances station coverage with demand equity	SCLP may underutilize low-demand stations	MCLP	–	Balances equity and efficiency in EV planning	Strategic deployment of city-wide charging stations
[54]	$min\left(T{C}_r\left(p^{*}\right)+T{D}_r\left(p^{*}\right)\right)$	AHP	Station placement	Integrates expert judgment; scalable	NP-hard p-median problem	Iterative p-median + AHP	–	Optimizes EV station number and location	Used for determining station quantity and optimal distribution
[32]	$\omega =\left\{\begin{array}{cc}{\omega }_{\max }-{\displaystyle \frac{({\omega }_{\max }-{\omega }_{\min })(f-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}}}\hfill & f>f_{\mathrm{avg}}\hfill \\ {\omega }_{\max }\hfill & f<f_{\mathrm{avg}}\hfill \end{array}\right.$	APSO	Holistic planning	Strong convergence speed; adaptive to system changes	Assumes fixed EV behavior	PSO	–	Holistic planning tool for EV infrastructure	Suitable for evolving grid environments
[46]	SCLP: $min\{{\sum }_{k}F{P}_{ik}+{\sum }_m(P_{t2,m}+P_{d2,m}+{\sum }_{j}S{1}_{ij,t}$ $+S{2}_{ij,t,k}+f_{j,t,k})+{\sum }_{k}BatteryWearCos{t}_{ik,t,k}\}$	MINLP	V2G siting in microgrids	Accurate; captures battery wear and costs	High computational complexity	Two-stage heuristic	–	V2G reduces microgrid costs.	Tailored for microgrids with energy storage
[49]	$min\left({\sum }_{t_0}^{t_1}P\left(t\right)M\left(t\right)\right) t_0\le t<t_1$	MAS	Residential charging	Reflects household consumption patterns	Focuses on residential charging only	–	TEEMA	MAS models heterogeneous agents	Home-based, agent-specific scheduling
[1]	$\mathrm{Minimize}{\sum }_{i\in N}{\sum }_{k\in K}{c}_i^k{X}_i^k$	MIP	Charging station layout optimization	Effective for integrated system coverage	Fixed path flows; no stochastic demand	Genetic Algorithm	–	Combines SRS, FRS, BES for better coverage.	Robust solution for defined grid layouts
[50]	$max{\sum }_{j\in J}{\sum }_{g\in G}{\tilde{f}}_i(h_j^g,\zeta ·P_g)$	IP	Location optimization	Easy to compute; straightforward structure	Independent location simulation	–	–	Optimal locations remain under budget changes	General location planning under cost variation
[43]	$\mathrm{Maximize}{\sum }_{q\in Q}{\lambda }_q{\overline{X}}_q$	IP	Budget-constrained allocation	Maintains stable allocations under changing parameters	No partial allocation allowed	Stochastic programming	–	Stable decisions across budgets and congestion	Applicable for staged rollout of EVSE
[51]	$\mathrm{Maximize}{\sum }_{i\in I}{d}_i^{12}{\sum }_{j\in J_i}{z}_{ij}^{12}+{\sum }_{i\in I}{d}_i^{13}{\sum }_{j\in J_i}{z}_{ij}^{13}$	CMCLP	Intracity charging access	Coverage for short urban trips	Focuses on intracity trips only	ABM	MATSim	Generalizable for future EVSE planning	Urban-specific mobility and accessibility design
[53]	${max}_{\mathbf{x}}{\sum }_{m=1}^M{\sum }_{i=1}^N{\sum }_{j=1}^N{f}_{ij}^m{z}_{ij}^m$	MIP	Routing optimization	High-capacity modeling	No adaptive routing	HGA	–	Covers outbound and return trips	Applied to round-trip network problems

presents a comparative summary of various optimization models and techniques employed in EV charging station planning. It highlights the diversity of approaches, including mathematical programming, simulation-based models, and meta-heuristic algorithms, along with their respective objectives, limitations, and practical implications. As shown in Table 3, linear and non-linear programming models are suitable for structured problems but fail to scale in uncertain environments. Meta-heuristics such as PSO and GA are highly flexible and widely used but may converge prematurely or lack theoretical guarantees. Hybrid models (e.g., AHP + p-Median, or AI + MILP) provide multi-criteria adaptability but increase computational cost. Furthermore, user-centric models (like fuzzy or agent-based scheduling) address behavioral uncertainties but require high data granularity. These trade-offs illustrate a gap in robust, real-time scheduling models that balance accuracy with scalability. Figure 6 illustrates the distribution of various modeling and algorithmic approaches employed in the study, categorized into mathematical models, simulation models, and meta-heuristic algorithms.

4.3. Urban Traffic and Demand Considerations

The third category involves urban traffic and demand considerations in location planning. Considering urban traffic will help in reduced hours spent in congestion and prolonged use of car batteries while demand coverage will ensure optimal utilization of EV stations and their profitability. Rajabi-Ghahnavieh and Sadeghi-Barzani [55] both highlighted the need to consider traffic flow and vehicle density in the deployment of fast-charging stations. Pazouki et al. [56] focused on risk management and participation planning for electric vehicles (EVs) in smart grids for demand response by adjusting the charging behavior of EVs in real time.

The location models for charging stations based on demand can be categorized into two main types:

4.3.1. EV Charging Station Location Models

In the charging station location model based on nodal demand, the flow of current through the nodes is taken into account. It assumes that electric vehicle charging, known as Plug-in Electric Vehicles (PEVs), takes place at specific geographical nodes within the target planning area, and charging stations are strategically positioned to meet this demand. However, this approach solely considers the straight-line geographical distance between the charging nodes, overlooking constraints related to the density of the transportation network [57].

Some studies have developed innovative models to address specific challenges in the EV station location problem. Celik and Ok [58] proposed a set cover model for optimal charging station location, which minimizes the number of stations required to cover a specific area while ensuring that all potential users can reach a station within a defined distance. Moreover, Ge et al. [59] presented a grid partition method for charging station planning, where the urban area is divided into grids, and the placement of charging stations is optimized within each grid. This method allows for fine-tuned control over the distribution of charging infrastructure across cities or regions.

He et al. [60] explored optimal location strategies considering factors like EV driving range, aiming to place stations in areas that mitigate range anxiety while ensuring the network’s efficiency. Li et al. [61] focused on dynamic demand, investigating how to optimize the layout of charging stations when demand is not fixed but fluctuates over time due to changing user behavior and seasonal patterns. This model helps create adaptable station networks that can evolve with growing EV adoption. Further studies such as Zeng et al. [62] specifically examined range anxiety by considering how far an EV can travel before needing a recharge. The optimization model incorporates distance between stations to ensure EV owners have access to charging infrastructure without experiencing anxiety about running out of power.

4.3.2. Location Based on Traffic Flow

In this planning model, which relies on traffic simulation, the aim is to estimate the charging requirements for PEVs (Plug-in Electric Vehicles). Typically, these simulations utilize data obtained from real-world trips, which can be costly to acquire in certain regions. Recognizing the dynamic nature of electric vehicle mobility, some researchers have proposed planning methodologies based on traffic flow analysis. In this approach, the flows of vehicles from their origins to destinations are analyzed to gauge the demand for charging, encompassing both the Flow Coverage Location Model (FCLM) and the Fuel-Flow Location Model (FRLM). The FCLM addresses a fundamental challenge of maximizing route coverage by determining the optimal placement of charging stations along flow paths to serve a given number of routes [15]. On the other hand, the FRLM, an adaptation of the FCLM, emphasizes route-based demand maximization to ensure that vehicles can be refueled along their journeys [17]. Unlike the FCLM, where a single charging station along the flow path suffices to cover the flow, the FRLM considers factors such as route distance and vehicle driving range, often necessitating the placement of multiple charging stations to prevent vehicles from running out of fuel mid-journey.

4.3.3. Routing Considerations

One of the primary challenges in electric vehicle routing is managing the battery charge at each node, as EVs’ range is determined by their battery charge level. Unlike traditional vehicles that rely on fuel availability, electric vehicles need precise calculation and management of battery charge to ensure that they can complete their routes. Yang [43] and Campana and Inga [63] also stress the importance of considering congestion, both at the stations and on the surrounding road network. This aspect is critical for optimizing the user experience and minimizing delays. Davidov and Pantos [64] highlight the role of reliability in the planning and operation of fast-charging infrastructure, ensuring that the charging network remains functional even under stress. Gan et al. [65] introduced the concept of elastic demand, focusing on the dynamic nature of charging station use. Their model adapts to changing demand, ensuring the efficient allocation of resources.

4.4. GIS and Advanced Tools for Planning

The use of Geographical Information Systems (GIS) and other advanced tools for electric vehicle (EV) charging station location planning has gained significant attention in recent years. Zafar et al. [66] provided a GIS-based framework for determining the optimal location of these stations. Maximal Covering Location Problem (MCLP) was addressed.

4.5. Combination of Charging Strategies (Slow and Fast)

The fifth category is dedicated to a combination of charging strategies (slow and fast) in location selection. Jia et al. [67] performed urban EV charging facility planning combining slow and fast charging. Demands are classified into slow, regular, and urgent demands. Slow and regular charging demands are met by charging piles in corresponding areas. Fast charging stations are planned according to the distribution of charging demand and the structure of the road network. Xu et al. [68] investigated the allocation of conventional and fast charging methods in distribution networks. Yan et al. [69] reported the optimal planning of electric vehicle charging stations (EVCSs) into a distribution system problem as a complex, non-linear, and combinatorial optimization problem and proposed a multi-objective and multivariate planning model based on Hierarchic Genetic Algorithm (HGA).

Hu and Song [70] developed a model that incorporates distribution network expansion alongside charging station planning. This approach ensures that the energy network is robust enough to support the growing number of EVs and the associated increase in demand for charging services. Rajabi-Ghahnavieh and Sadeghi-Barzani [55] and Rastegarfar et al. [71] took a similar approach by investigating the zonal placement of fast-charging stations. These studies use advanced optimization techniques to place fast-charging stations in strategic locations, ensuring that the network can handle high demand and support long-distance travel. Guojun et al. [72] studied the hierarchical classification of electric vehicles (EVs) by analyzing the differences in the charging behaviors of different types of EVs. Four indexes (the increment of peak load, the duration of peak load, the maximum smoothing index of load curve, and the average smoothing index) were proposed to evaluate the impact on the distribution network.

5. Optimal Allocation and Scheduling of EV Parking Lots

Figure 7 categorizes the various studies conducted on optimal allocation and scheduling of EV parking lots. It can be seen that there are three broad categories:

• Scheduling and optimization of EV charging and discharging,
• Intelligent parking lot management,
• Advanced techniques and cooperative mechanisms.

These categories are described in detail as follows:

5.1. Scheduling and Optimization of EV Charging and Discharging

While several optimization techniques such as MILP, GA, and PSO are also used in charging station location planning, their application in scheduling problems differs significantly in terms of objectives, constraints, and time horizons. Location planning is a strategic-level problem focused on long-term infrastructure deployment, whereas scheduling is a tactical or operational-level problem that deals with short-term decisions such as when and how much to charge or discharge EVs. Therefore, although similar methods may appear in both contexts, their formulations and solution strategies are tailored to the specific characteristics of each problem.

Dynamic Programming (DP) is often used in scenarios where decision-making is sequential, and the goal is to minimize a cost function over time. However, DP struggles with the “curse of dimensionality” as the number of EVs increases, resulting in computational complexity.

Day-ahead scheduling involves predicting the demand and supply of electricity for the next day and optimizing the scheduling accordingly. While effective in many situations, it may not adapt well to sudden fluctuations in demand. Jin et al. [21] investigated how aggregators can manage energy trading in both day-ahead and real-time markets, with an aim to minimize the total cost of charging. A Mixed-Integer Linear Programming (MILP) model is proposed to determine optimal charging schedules for EVs in coordination with energy storage.

Model-based methods are particularly useful for finding near-optimal solutions in complex, high-dimensional problems. Shaheen et al. [73] investigated the performance of Differential Evolution (DE), Particle Swarm Optimization (PSO), Whale Optimization Algorithm (WOA), and Grey Wolf Optimization (GWO) for cost-efficient Vehicle-to-Grid (V2G) scheduling. Rezaei and Akhbari [74] presented a hybrid meta-heuristic approach to EV scheduling, which balances charging and discharging to reduce load fluctuations on the grid. Rezaei and Golkar [75] used machine learning forecasts to predict future load demands and modeled the charge and discharge management of EVs. Sharma et al. [76] explored a multi-objective scheduling problem that maximizes the aggregator’s profit, minimizes charging costs, and optimizes the SOC for EVs using a heuristic dynamic optimization method. Ahmadi et al. [77] addressed the scheduling of EVs in parking lots to flatten the demand curve, using a non-linear optimization model. Shariatzadeh et al. [78] focused on workplace parking lot scheduling and found that the energy price differential significantly impacts the ability to meet desired SOC targets within a defined charging budget. Ostado-Véliz et al. [79] proposed a two-stage stochastic model for scheduling in energy communities. Yang et al. [80] proposed a real-time energy management strategy for parking lots with high EV penetration. Sausen et al. [81] explored prioritization-based residential EV scheduling, where charging priorities are assigned based on various factors such as the EV’s SOC, user preferences, and time of day. While MILP models such as those by [21,26] provide strong optimization under deterministic settings, they often fail to adapt in real-time environments with uncertainty. This gap is addressed by stochastic models like those in [51,82], but at the cost of higher computational complexity. Meta-heuristic methods (e.g., PSO, GA) are commonly used due to their flexibility in multi-objective contexts. However, they often lack convergence guarantees and can suffer from local optima. A summary of the optimization models, their objectives, limitations, and implications in EV charging scheduling is presented in

Table 4. Overview of optimization models and algorithms for EV charging scheduling: Insights from selected studies.

Ref.	Model		Limitations	Algorithm	Implication
	Objective Function/Formula	Type
[77]	$\mathrm{Objective} \mathrm{Function}=min\left[\left(W_1\times {\displaystyle \frac{f_1}{f_1^{\prime }}}+W_2\times {\displaystyle \frac{f_2}{f_2^{\prime }}}\right)+\left(W_3\times {\displaystyle \frac{f_3}{f_3^{\prime }}}+W_4\times {\displaystyle \frac{f_4}{f_4^{\prime }}}+W_5{f}_5\right)\right]$	MINLP	RES uncertainty is ignored; deterministic generation profiles are assumed.	HMA	The proposed method reduces total cost and technical losses compared to standalone GA.
[21]	$max{\sum }_t\left[\mathrm{Revenue} \mathrm{from} \mathrm{EVs}+\mathrm{Regulation} \mathrm{Revenue}-\mathrm{Electricity} \mathrm{Cost}\right]$	MILP	Relies on day-ahead forecasts; real-time adaptability is limited to hourly updates.	LP	ES helps mitigate mismatches between day-ahead forecasts and real-time demand, reducing penalties.
[75]	$L={\displaystyle \frac{\mathrm{Max}\left(P_{ui}\right)+\mathrm{Min}\left(P_{ui}\right)}{2}}$	MINLP	Assumes known arrival/departure times and energy needs; no stochastic modeling.	Support Vector Regression (SVR) using SVM	SVM-based forecasting improves scheduling accuracy (MAPE = 1.53%).
[81]	$max{\varphi }^k={\sum }_{j=k}^J\left[{\sum }_{n=1}^N\left(C_n^j{p}_n^{C,\mathrm{m}á\mathrm{x}}{\rho }_n^{C,j}{C}_j^n{\tau }_n^{C,j}{C}_j^n+D_n^j{p}_n^{D,\mathrm{m}á\mathrm{x}}{\rho }_n^{D,j}{D}_j^n{\tau }_n^{D,j}{D}_j^n\right)\right]$	MINLP	Focuses on transformer capacity but does not model voltage, phase imbalance, or cable limits.	MINOS (via NEOS server)	Prevents overload by scheduling discharging to support transformer capacity.
[73]	$\mathrm{Minimize}{\sum }_{i=1}^N{\sum }_{t=1}^T\left|\left(C^{\left(0\right)}·x_i^{\left(0\right)}·P^{\left(0\right)}\right)-\left(R^{\left(0\right)}·y_t^{\left(0\right)}·P^{\left(0\right)}\right)\right|$	MINLP	Focuses on transformer capacity but omits voltage, frequency, and phase balancing.	PSO, DE, WOA, GWO	WOA had the best convergence; PSO was fastest computationally.
[78]	$\mathrm{Min}{f}_j\left(x\right)={\sum }_{n=1}^N{\sum }_{c=1}^C{\sum }_{t=1}^T\left((C{b}_t\times P_{n,c,t}^{+}\times \Delta t)-(C{s}_t\times P_{n,c,t}^{-}\times \Delta t)\right)$	MILP	Charging behavior (SoC, budget) is predicted but treated as deterministic in optimization.	SVM, MLP	Incorporates user preferences (SoC, budget) into optimization.
[76]	$max\left(\mathrm{Profit}={\sum }_{t=0}^T({\mathrm{Revenue}}_t-{\mathrm{Cos}\mathrm{t}}_t)\right)$	MOOP	Assumes fixed upstream/downstream prices; no bidding or ancillary services.	MOHESA	Balances aggregator profit with user satisfaction (SoC and cost).
[79]	$\mathrm{Cost}={\sum }_{r\in R}\left\{{\pi }_r·{\sum }_{t\in T}\left[E\left({\lambda }_t^{\mathrm{Grid}, \mathrm{buy}}\right)·p_{s|t}^{\mathrm{Grid}, \mathrm{buy}}-E\left({\lambda }_t^{\mathrm{Grid}, \mathrm{sell}}\right)·p_{s|t}^{\mathrm{Grid}, \mathrm{sell}}\right]\right\}$	MILP	Focuses on energy balance; omits voltage, frequency, and line limits.	–	Combines stochastic and IGDT to handle different uncertainty types effectively.
[80]	$S_{fin}^{\prime }=min\left\{S_{ini}+\left(P_{\max }^{ch}·{\eta }^{ch}·t_{stop}\right)/E,S_{max}\right\}$	Fuzzy	Real-time strategy lacks global foresight; may miss optimal long-term schedules.	–	Strategy is computationally light and suitable for real-time deployment.

.

5.2. Intelligent Parking Lot Management

The management of electric vehicle (EV) charging in intelligent parking lots (IPLs) plays a significant role in optimizing the use of parking spaces, reducing costs, and improving user satisfaction while maintaining grid stability. Alinejad et al. [83] presents a model for intelligent parking lot management that considers the random behaviors of EV owners and other real-world conditions. Honarmand et al. [84] explored the self-scheduling of EVs in an intelligent parking lot using stochastic optimization techniques. Nejati et al. [85] investigated the use of Particle Swarm Optimization (PSO) to optimize the scheduling of EVs in smart parking lots (SPLs). Zenhom et al. [86] proposed a stochastic approach to coordinating the charging and discharging of EVs in PV-powered parking lots using Monte Carlo simulation and sample average approximation. Mehrabi and Kim [87] focused on charging/discharging scheduling in multiple homes and common parking lots for smart household prosumers using a mixed optimization model. Manshadi et al. [88] explored the integration of wireless EV charging in transportation networks. Shaheen et al. [73] performed cost reduction in EV parking lot management using meta-heuristic algorithms. Ahmadi et al. [77] proposed hybrid algorithms for optimizing EV parking lot operations.

5.3. Advanced Techniques and Cooperative Mechanisms

The third category of research focuses on advanced techniques and cooperative mechanisms that aim to optimize EV charging scheduling. Zhang et al. [35] introduced a cooperative V2V charging model that enables energy exchange between EVs—one acting as an energy consumer and the other as an energy provider using the concept of stable matching, where the goal is to match EVs that need energy with those capable of providing energy, under a cooperative framework. Baharifard et al. [89] studied the impact of intelligent charging systems on grid imbalance indices and proposed a two-stage framework for EV charging scheduling. The cooperative mechanisms proposed by Zhang et al. [35] also intersect with grid management strategies, such as V2V energy sharing, which can be integrated into the broader grid through the Vehicle-to-Grid (V2G) paradigm. Baharifard et al. [89] addressed intelligent scheduling techniques for V2V whereby EVs are more flexible and capable of responding to both charging demand and grid requirements dynamically.

5.4. Emerging AI Techniques for Demand Forecasting

While this review does not focus on AI methods in depth, it is important to acknowledge their growing role in EV demand forecasting and scheduling. Recent studies have begun integrating artificial intelligence (AI) techniques into EV charging scheduling and demand forecasting. For instance, Long Short-Term Memory (LSTM) networks are used to model temporal patterns in charging behavior, while Deep Reinforcement Learning (DRL) enables adaptive scheduling under dynamic grid conditions. Graph Neural Networks (GNNs) have also been explored for spatial demand prediction across charging networks. These methods offer advantages over traditional models by capturing non-linear dependencies and enabling real-time decision-making. Although AI is not the primary focus of this review, its increasing application in EV infrastructure planning warrants further investigation [90,91,92].

Figure 8 presents a heatmap illustrating the distribution of research methods employed across different years. Some years are not represented on the heatmap axis because the associated research methods were either not specified, categorized under broad labels, or primarily related to the general ‘Simulation’ category.

6. V2G and Smart Charging Systems

Figure 9 presents the various categories reported in the literature for V2G and smart charging systems. They are presented in detail as follows.

6.1. Planning and Siting of EV Parking Lots

The first category of research focuses on the planning and layout of EV parking lots (EVPLs), which is a crucial aspect of the development and optimization of electric vehicle (EV) infrastructure. Rahmani-Andebili [34] investigated the placement and sizing of EV parking lots from the perspective of a local distribution company (DISCO) and minimized the total cost associated with EV parking lot planning, considering not just technical aspects but also social and geographical factors. Mozaffari et al. [93] explored expansion planning for EV parking lot placement and its integration with distribution networks using a probabilistic algorithm to model EVs’ driving, charging, and discharging behaviors. In Moradijoz et al. [94], a two-stage optimization model is proposed for parking lot allocation in smart grids. Mirzaei et al. [82] adopted a probabilistic approach based on the point estimate method to determine the optimal location and capacity of EVPLs in distribution networks. Kazemi et al. [95] focused on optimizing both the location of EV parking lots and the placement of EVs within distribution systems.

6.2. Integration with Distributed Energy Resources and Renewables

The second category of research focuses on the integration of EV parking lots (EVPLs) with distributed energy resources (DERs) and renewable energy sources (RESs), such as solar PV and wind energy. Kashiri et al. [96] explored a hybrid energy management system for intelligent parking lots (IPLs) that are integrated with hydrogen storage systems (HSSs) and renewable energy sources (RESs) such as solar and wind energy using the honey badger algorithm (HBA). Shafie-Khah et al. [97] presented a two-level optimization model for EV parking lot operators (PLOs) integrated with renewables (such as wind and solar PV) within distribution systems. Qi and Li [98] developed an economic operation strategy for EV parking lots integrating V2G and renewable energy sources (RESs). The model uses a hybrid approach, integrating economic dispatch with real-time optimization to ensure that the cost of charging is minimized while also ensuring that renewable energy sources are fully utilized.

6.3. Challenges and Implementation of V2G Systems

The third category focuses on the challenges and implementation of Vehicle-to-Grid (V2G) systems in electric vehicle parking lots (EVPLs). Comi and Elnour [99] identified technical issues such as battery degradation, user behavior variability, and the complexity of bi-directional charging infrastructure as main issues. Moreover, the integration of V2G into the grid requires overcoming regulatory and financial barriers, as well as establishing standards for interoperability between different charging stations and grid operators. AlNahhal et al. [100] optimized the V2G operations to improve the reliability of the distribution network, reducing the risk of blackouts or energy shortages. Al-Obaidi et al. [101] developed a model for the optimal design of a V2G-capable electric vehicle parking lot (EVPL) using a financial model for V2G services and an incentive-participation scheme for EV owners. Li et al. [102] highlight the importance of optimizing the degradation cost associated with the charging/discharging cycles of EVs in V2G applications. Kandil and Osman [103] indicate that the location and profile of the parking lot significantly affect its flexibility in reducing peak load and optimizing charging costs. Shahinzadeh et al. [104] investigated the reliable operation of V2G systems, particularly under probabilistic mobility patterns of Plug-in Hybrid Electric Vehicles (PHEVs).

6.4. Scheduling and Charging Management

Ref. [68] focused on the forecasting and schedulable capacity calculation for EV parking lots using real-time data on EV availability, user behavior, and grid conditions, allowing for dynamic and adaptive scheduling. Zhang et al. [92] explored the integration of Urban Internet of EV (IoEV) Systems for V2G scheduling using distributed algorithms. Jakus et al. [105] utilized Mixed-Integer Linear Programming (MILP) to address the energy supply management for EV parking lots (EVPLs). Morais [106] proposed a multi-objective optimization approach to balance charging costs, scheduling efficiency, and fairness among users. Sadati et al. [90] focused on a cooperative behavior among multiple EVPLs, which helps balance demand and reduce energy costs for the community. Xu et al. Kandil and Osman [103] indicate that some parking lots, especially those with a high density of EVs, are more flexible in implementing strategies like peak load shifting, which can help reduce the overall cost of electricity for both the parking lot owners and EV users.

6.5. Advanced Optimization Models and Frameworks

Ahmadi et al. [107] introduced a two-stage optimization approach that combines Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to optimize the allocation of charging stations, distributed generation (DG), renewable energy, and storage systems in EV parking lots (EVPLs). Shafie-Khah et al. [97] presented an innovative two-level model designed for EV parking lots integrated with renewable energy in distribution networks using forecasted load data for the day-ahead and implement energy management strategies using tools like General Algebraic Modeling System (GAMS) software. Zhao et al. [108] explored uncertainty management in shared parking spaces using an agent-based approach.

6.6. Multi-Objective and Economic Allocation for Parking Lots

The sixth category is dedicated to the multi-objective optimization and economic allocation for EV parking lots (EVPLs), particularly in the context of distributed energy resources (DERs), smart charging, and Vehicle-to-Grid (V2G) integration. Ferraz et al. [109] focused on multi-objective optimization for EV parking lots (EVPLs) in the context of smart charging and distributed energy resources (DERs) using Non-dominated Sorting Genetic Algorithm II (NSGA-II). Pazouki et al. [56] explored the simultaneous optimal planning of charging stations and distributed generation (DG) systems using Genetic Algorithm. Xu et al. [68] proposed an innovative forecasting method for calculating schedulable capacity in EV parking lots using a day-ahead time-of-use clustering strategy to provide more accurate and detailed predictions of EV charging behavior. Qi and Li [98] presented an economic operation strategy for EV parking lots that integrates V2G and renewable energy sources, such as photovoltaic (PV) and wind turbines, along with energy storage systems.

6.7. Scheduling and Charging Management

Efficient scheduling and charging management are crucial for optimizing the operation of electric vehicle parking lots (EVPLs), enhancing the flexibility of charging processes, and ensuring a balanced energy distribution to support both electric vehicles (EVs) and the grid. Zhang et al. [92] proposed a novel approach by integrating an urban Internet of Electric Vehicle Parking System to facilitate V2G scheduling. This system addresses the allocation and scheduling of electric vehicles (EVs) with joint V2G regulation services and edge computing services using Mixed-Integer Quadratic Programming. Kandil and Osman [103] assessed the flexibility of different parking lots in adopting smart EV charging strategies and their potential for peak load reduction and cost optimization.

To provide a clearer understanding of methodological trends across key research themes,

Table 5. Correlation of research methods and thematic focus areas in the EV charging literature.

Research Method	Deployment and Placement	Scheduling and Allocation	V2G and Smart Charging
MILP/MINLP/IP	✓✓	✓	✓
Heuristics (GA, PSO, WOA, etc.)	✓✓	✓✓	✓
Agent-Based Models (ABM)	✓	✓✓	✓✓
Simulation Models (e.g., TEEMA, Monte Carlo)	✓	✓	✓✓
AI and Machine Learning (LSTM, DRL, GNN)	✓	✓✓	✓✓
Multi-Criteria (AHP, Fuzzy, MCDM)	✓✓	✓	–
Hybrid Approaches (e.g., AI + MILP, GA + PSO)	✓	✓	✓

summarizes the correlation between major research methods and the three thematic focus areas identified in this review: deployment and placement, scheduling and allocation, and V2G and smart charging. The symbol ✓✓ indicates methods that are foundational or frequently applied in a given domain. As shown in Table 5, deployment and placement problems are predominantly addressed using optimization-based methods such as MILP, MINLP, and IP. These models offer precise siting and cost minimization capabilities and are often complemented by multi-criteria decision-making approaches like AHP and fuzzy logic when qualitative factors are considered.

In contrast, scheduling and allocation increasingly incorporates agent-based models and AI-driven forecasting techniques such as LSTM and deep reinforcement learning. These methods are well-suited for handling dynamic user behavior, time-dependent charging events, and adaptive control.

For V2G and smart charging, the literature shows a preference for simulation tools and AI models, which are effective in modeling bidirectional power flow, stochastic demand, and complex grid interactions. Hybrid methods are found across all three themes, reflecting the growing trend toward integrating optimization with learning-based models to address real-time, uncertain environments.

7. Conclusions, Limitations, and Future Research Directions

From the perspective of EV owners, the lower fuel and operational costs of EVs compared to vehicles with internal combustion engines are significant advantages. This cost reduction is primarily attributed to the higher efficiency of electric engines in contrast to internal combustion engines. Furthermore, Vehicle-to-Grid (V2G) technology presents an additional opportunity for EV owners to generate revenue by leveraging the energy stored in their batteries to exchange with the grid. This paper presents a thematic literature review on electric vehicle (EV) charging station location planning and scheduling, drawing insights from 95 research papers obtained through a comprehensive literature review. The review focuses on three primary themes (in decreasing order of importance):

• Charging station deployment and placement,
• Optimal allocation and scheduling of EV parking lots,
• V2G (Vehicle-to-Grid) and smart charging systems.

For charging station deployment and placement, key themes are location planning for EV charging stations, optimization models and techniques, urban traffic and demand considerations, GIS and advanced tools for planning, and a combination of charging techniques (slow and fast charging). The review highlights the maximum utilization of optimization, multi-criteria decision-making, and simulation techniques in the planning and scheduling of EV charging stations, with hybrid meta-heuristics such as GA-PSO proving effective in addressing multi-objective problems. Stochastic models have also been shown to be useful in handling uncertainties related to user behavior and renewable energy availability. Commonly used criteria for location planning are costs, accessibility, proximity to user locations, and presence of traffic congestion. The overall focus is to reduce the range anxiety of users.

For optimal allocation and scheduling of EV parking lots, the key themes are the scheduling and optimization of EV charging and discharging, intelligent parking lot management, and advanced techniques and cooperative mechanisms. Different levels of charging and user convenience are considered. The charging could happen in residential zones, en-route, or destination nodes. A variety of optimization techniques including game theory and multistakeholder models are proposed. Emerging trends include the integration of machine learning for predictive modeling and a growing emphasis on user convenience and minimizing waiting times at charging stations.

For V2G and smart charging systems, the key themes are planning and siting of EV parking lots, integration with distributed energy resources and renewables, challenges and implementation of V2G systems, scheduling and charging management, advanced optimizations models and frameworks, multi-objective and economic allocation for parking lots, and scheduling and charging management. Many recent studies have focused on the impact of EV charging on the grid, with V2G systems emerging as a potential solution to improve grid stability. However, these systems require careful planning to avoid negative effects such as voltage imbalances. From a socio-economic perspective, it has been observed that financial incentives and user participation play a crucial role in the success of EV infrastructure deployment. Additionally, some studies emphasize incorporating solar or wind energy into EV parking lot operations to reduce grid dependence and operational costs.

Limitations of the Study: While this review provides a comprehensive overview of EV charging station planning and scheduling, it is limited by the scope of the selected literature (95 papers) and the focus on predominantly urban and developed regions. Some emerging technologies and regional practices may not have been fully captured. Additionally, the review does not include a quantitative meta-analysis, which could have provided deeper statistical insights into model performance and outcomes.

Future research directions include the following:

• Climate and Geographic Adaptability: Investigate how EV and V2G systems perform under diverse climatic conditions and urban forms, especially in developing countries where infrastructure and energy access vary significantly.
• Dynamic Pricing and User Behavior: Explore how real-time and time-of-use pricing models influence user participation in smart charging and V2G programs and develop adaptive scheduling algorithms that respond to these incentives.
• Best Practices and Policy Frameworks: Conduct comparative studies across regions to identify effective regulatory, financial, and technical strategies for EV infrastructure deployment, with a focus on scalability and equity.
• Sustainability and Grid Integration: Assess the environmental and economic impacts of integrating renewable energy sources (e.g., solar, wind) into EV charging networks, including lifecycle emissions and grid stability.
• AI-Driven Predictive Scheduling: Develop machine learning models that forecast demand, user behavior, and renewable generation to optimize real-time charging and discharging schedules, particularly in high-penetration EV scenarios.

These areas provide valuable opportunities for future research, driving the evolution of EV infrastructure and its integration with the energy grid.