Advancing Hybrid AC/DC Microgrid Converters: Modeling, Control Strategies, and Fault Behavior Analysis

Abstract

Hybrid AC/DC microgrids (HMGs) are pivotal for integrating renewable resources, yet their stability and resilience are fundamentally constrained by the power electronic converters that interface them. This paper provides a critical review and synthesis of the co-dependent advancements in HMG converter topologies, control strategies, and fault management. Through a systematic analysis of the state of the art, this review examines the evolution from classical control to intelligent, software-defined converter functions. The analysis reveals a fundamental bifurcation in design philosophy between low-voltage (LV) and medium-voltage (MV) systems, driven by a trade-off between power density Gallium Nitride (GaN) and systemic reliability silicon carbide (SiC). Furthermore, it highlights the rise of virtualization, namely virtual Inertia control (VIC) and adaptive virtual impedance control (AVIDC), as a dominant paradigm to compensate for the physical limitations of low-inertia, resistive grids. Finally, this review identifies a critical, synergistic dependency in fault management, where ultra-fast solid-state circuit breakers (SSCBs) guarantee the survivability of vulnerable voltage source converters (VSCs), which in turn enables software-based resilience via fault ride-through (FRT). This synthesis concludes that the converter has become the intelligent nexus of the HMG and identifies the primary barriers to widespread adoption as the computational, economic, and standardization gaps in this new cyber–physical domain.

1. Introduction

The past decade has demonstrated the necessity of improving power system efficiency and sustainability through the integration of distributed energy generators (DGs) due to increasing energy demands. Traditional energy sources, particularly fossil fuels, contribute to various environmental and operational challenges. Therefore, a broader adoption of renewable energy sources (RES) has become essential [1].

The advantages of AC microgrids (MGs) and DC MGs differ based on specific application needs and system demands. AC MGs work well with existing power setups, which simplifies grid integration and supports smooth connections between traditional AC loads and generators. They support long-distance power transmission more efficiently due to the well-established AC grid infrastructure and can utilize transformers for voltage regulation. On the other hand, DC MGs are more efficient for applications involving renewable energy sources like photovoltaic (PV) system and battery storage, as these naturally generate DC power. They minimize conversion losses by reducing the need for AC-DC and DC-AC conversions, improving overall system efficiency [2,3]. DC MGs also support faster response times in energy management and are particularly beneficial for data centers, electric vehicle charging, and low-voltage applications where direct DC supply is advantageous. They eliminate reactive power circulation, resulting in higher efficiency with fewer converters. Additionally, they simplify voltage regulation through DC-DC converters and facilitate the integration of multiple DC-based units. The increasing demand for DC loads, such as electric vehicles (EVs), further strengthens the appeal of DC-based MGs. As a result, DC-based MGs have become a preferred choice for modern power generation and distribution systems [4,5].

A hybrid AC/DC MGs is often the best choice because it combines the advantages of both AC and DC MGs while minimizing their weaknesses. Hybrid AC-DC MGs provides several benefits that improve the performance and reliability of electrical power systems. A hybrid AC/DC MG includes both AC and DC networks, with DGs and loads. It connects to the utility grid through bidirectional power converters, allowing electricity to flow in both directions. This two-way power exchange improves system stability [6,7].

Hybrid MGs also enhance grid reliability by operating independently when needed and managing power flow efficiently, which helps reduce power losses. These benefits make the system more efficient and environmentally friendly while supporting different energy sources and load demands. Additionally, combining AC and DC networks simplifies rather than complicates operation [8]. A standard AC/DC hybrid MGs configuration, illustrated in Figure 1, consists primarily of an AC subnet, a DC subnet, and an interlinking power converter connecting both areas.

Hybrid AC/DC MGs offer many advantages but also present several challenges that impact their stability, reliability, and efficiency. Protection challenges arise due to faults, short circuits, and reliability issues in both AC and DC systems, requiring effective fault detection and coordination. Operational challenges involve power control, stability, and grid integration, where managing the interaction between AC and DC components, ensuring smooth energy transfer, and addressing communication delays are critical. Additionally, power quality issues such as voltage stability, harmonics, and load balancing must be carefully managed to maintain overall system performance. Overcoming these challenges is essential for developing resilient and efficient hybrid MGs. In general, all challenges in hybrid AC/DC MGs can be divided into three main categories [9,10,11]. These challenges, shown in Figure 2, include both technical and operational factors that are essential for ensuring the efficient and reliable performance of a hybrid MG.

As illustrated in Figure 2, modern power converters play a crucial role in various MG applications, significantly enhancing their functionality beyond basic power management. Hybrid AC/DC MGs integrate advantages from AC and DC MGs and are connected together with interlinking converters. They can increase reliability and stability. The use of DC-DC, AC-DC, and DC-AC converters optimizes power conversion by minimizing conversion stages, which reduces energy losses and enhances overall efficiency. These systems simplify control strategies while enabling seamless voltage regulation through transformers and power electronics. Furthermore, these converters support the efficient integration of RESs and EVs into existing distribution networks with minimal infrastructure modifications, enhancing adaptability and scalability [7].

Researchers have developed advanced converters that improve the efficiency, reliability, and adaptability of MGs. The choice of converter depends on the characteristics of the power generated RESs. In DC MGs, AC/DC or DC/DC converters are utilized, whereas AC MGs require DC/AC, AC/AC, or multilevel converters to effectively regulate power flow. In addition to power conversion, these converters incorporate features such as fault current management, harmonic suppression, power quality improvement, and efficiency optimization. These advancements contribute to the enhanced reliability and sustainability of MGs within modern energy systems.

Although prior surveys have addressed individual challenges in converter modeling, control strategies, or protection, a critical gap remains in synthesizing their co-dependent limitations. Most existing works analyze these domains in isolation, which obscures several unresolved issues:

• The control–hardware dependency, where advanced controller designs such as MPC, virtual inertia, or adaptive droop are constrained by semiconductor limits and computational cost.
• The survivability–resilience linkage, in which software-based fault ride-through capabilities fundamentally rely on ultra-fast hardware protection such as SSCBs to prevent destructive VSC capacitor discharge.
• The application-driven bifurcation between LV and MV systems, where different operational requirements lead to divergent trade-offs in semiconductor technology (GaN vs. SiC), control bandwidth, and protection philosophy.

This review addresses these interconnected gaps by providing an integrated synthesis of converter topologies, modern control strategies, and advanced fault management, highlighting their mutual constraints and co-evolution.

This review is organized into seven sections. Section 1 introduces hybrid AC/DC MGs and their key advantages. Section 2 discusses the main converter types and their functional roles. Section 3 reviews converter topologies such as AC–DC, DC–DC, interlinking, and multilevel configurations. Section 4 analyzes conventional and advanced control strategies applied to these converters. Section 5 compares converter design and applications across low- and medium-voltage systems. Section 6 examines fault behavior and modern protection schemes, followed by a conclusion in Section 7 summarizing the key insights and outlining future research directions for advancing hybrid AC/DC MGs.

2. Power Converters Utilized in AC/DC MGs

In hybrid AC/DC networks, power electronic converters enable efficient electrical power conversion, ensuring effective control and transfer between AC and DC systems while enhancing overall system stability and performance [8,10]. These converters support bidirectional power flow, effectively integrating AC and DC subgrids with each other and the main grid. Their primary function is to ensure stable and efficient power distribution, enhancing the overall performance of the hybrid network.

2.1. An Introduction to Power Converter Selection

Selecting the most suitable power converter for an MG depends on the specific application requirements. An ideal power converter should possess the following key characteristics:

• Low impedance: This reduces power losses and enhances voltage regulation, improving overall system efficiency [12].
• Bidirectional power flow: Enables efficient energy exchange between AC and DC sub-grids, ensuring seamless operation [13].
• Flexible control strategies: Supports techniques like droop control to effectively balance power distribution between sources and loads [8,14].
• Stable voltage and frequency regulation: Maintains consistent voltage and frequency levels, ensuring the reliability of the MG [14].

Generally, power converters in AC/DC MGs can be classified in four groups, based on their operational characteristics:

• Grid-forming power converter (GFPC)

These converters are responsible for generating reference voltage and frequency in the network and are mainly used with energy storage systems (ESS). These converters act as controlled voltage sources with low series impedance, commonly used in centrally controlled MGs. These converters have different performances in IS and GC mode. In IS mode, they control voltage and frequency to maintain system stability. In GC mode, they regulate power to manage battery charge levels and improve power quality in both AC and DC networks [15].

2.

Grid-following power converter (GFLPC)

These converters are used with renewable energy sources like solar panels and wind turbines, and they follow the voltage and frequency of the grid, injecting power accordingly. These converters have different performances in AC and DC network. In an AC network, they inject active and reactive power to maintain an optimal power factor. In a DC network, they supply current or power while following the reference voltage of the grid. This kind of converter acts as controlled current sources with high parallel impedance, meaning they follow the grid but do not regulate it [16].

3.

Grid-supporting power converter (GSPC)

When grid-forming converters cannot maintain the required voltage and frequency, grid-supporting converters assist by using additional energy sources. These converters have different performances in IS and GC mode. In IS mode, they help maintain voltage and frequency stability. In GC mode, they enhance power quality and voltage regulation. These converters can operate in both voltage control and current control modes, using droop control strategies to balance the power demand and supply [17].

4.

Interlinking power converter (IPC)

Interlinking power converters are used to connect AC and DC sub-grids, enabling seamless power transfer between them. These converters allow bidirectional power flow, meaning they can transfer energy from AC to DC and vice versa, depending on the power availability and demand in each sub-grid. In IS mode, IPCs facilitate smooth power exchange between AC and DC systems, ensuring stable and balanced operation. In GC mode, they help optimize power flow, provide ancillary services such as voltage and frequency support, and enhance overall system efficiency. IPCs act as controlled voltage or current sources with medium impedance and operate using droop control strategies to maintain system stability across different sub-grids. Despite their advantages, IPCs may face challenges such as circulating currents between parallel converters, synchronization issues after faults, and non-linear load behaviour, which can affect performance and efficiency [18,19,20].

Table 1. Comparison of power converters in AC/DC MGs.

Features	Grid-Forming Converter	Grid-Following Converter	Grid-Supporting Converter	Interlinking Converter
Source type	Controlled voltage source	Controlled current source	Controlled voltage/current source	Controlled voltage/current source
Output impedance	Low	High	Medium	Medium
Control strategy	Constant V/f (AC), Constant Voltage (DC)	PQ control (AC), Current control (DC)	Droop control (P/f, Q/V for AC; V-I or V-P for DC)	Bidirectional droop control
Associated sources	Dispatchable (ESS)	Renewable (Solar, Wind)	Dispatchable DGs/ESS	Both AC and DC grids
Voltage and frequency stability	Fixed	Synchronized with the grid	Regulated	Regulated
Operational modes	Islanded (IS)	Grid-connected (GC)	Both IS/GC	Both IS/GC
Power flow control	Two-way	One-way	Mostly two-way	Two-way

provides an overview of different power converters and their specifics used in AC/DC MGs.

The common challenges faced by MG operators/engineers are unequal load sharing, power quality issues, synchronization failure, safety concerns, protection issues, stability problems [8]. Among the listed converters, grid-forming converters are the most critical for MG stability as they provide reference voltage and frequency, making them essential for islanded operation. Interlinking converters, however, offer the most flexibility as they connect AC and DC networks, allowing bidirectional power flow and supporting various grid conditions. Grid-following converters are best suited for renewable energy integration, while grid-supporting converters provide additional stability when the primary grid is under stress [15,16,17,18,19,20].

2.2. Different Types of Converters in Hybrid AC/DC MGs

Several power converters are needed in hybrid AC/DC MGs to manage AC to DC energy transfers between subsystems control voltage levels and maintain operational efficiency. These converters can be categorized into the following types:

2.2.1. AC-DC and DC-AC Converters (Rectifiers and Inverters)

AC-DC converters convert AC power into DC to supply DC loads or charge energy storage systems. They can be diode-based, thyristor-based, or PWM rectifiers, with varying levels of efficiency and control [21,22]. On the other hand, DC-AC converters transform DC power into AC, enabling integration with AC loads and the main grid.

Table 2. Overview of AC-DC and DC-AC converters.

Converter Type	Description	Advantage	Disadvantage	Direction	Power Control Approach
Diode Rectifier (Uncontrolled)	Uses diodes for AC-DC conversion	Simple, robust	No voltage control	One-way (Unidirectional)	Grid-Following
Thyristor Rectifier (Controlled)	Uses thyristors for controlled rectification	Voltage control possible	Generates harmonics	One-way (Unidirectional)	Grid-Following, Limited Grid-Supporting
Pulse width modulation (PWM) Rectifier (Active Rectifier)	Uses IGBTs/MOSFETs for controlled rectification	High efficiency, low harmonics, bidirectional	Complex control system	Two-way (Bidirectional)	Interlinking, Grid-Supporting
Inverter (DC-AC Converter)	Converts DC to AC	High efficiency in power conversion	Generate harmonics if not properly controlled	One-way (Unidirectional)	Grid-Following
Bidirectional AC-DC Converter	AC to DC and DC to AC	Supports grid-tied energy storage	High control complexity	Two-way (Bidirectional)	Interlinking, Grid-Supporting

illustrates overview of AC-DC and DC-AC converters, their advantages, disadvantages, directionality, and power control approaches.

2.2.2. DC-DC Converters

DC-DC converters regulate voltage levels within the DC subgrid, enabling power flow between different voltage domains. Buck converters step down voltage, while boost converters step it up. Buck-boost converters offer flexibility by handling both step-up and step-down operations [23,24]. Bidirectional DC-DC converters enable efficient energy transfer between storage units and loads, significantly enhancing MG energy management by optimizing power distribution and utilization [25].

Table 3. Overview of DC-DC converters.

Converter Type	Description	Advantage	Disadvantage	Direction	Power Control Approach
Buck Converter	Steps down DC voltage	Simple, efficient	Cannot boost voltage	One-way (Unidirectional)	Grid-Following
Boost Converter	Steps up DC voltage	Increases voltage efficiently	Sensitive to load variations	One-way (Unidirectional)	Grid-Following
Bidirectional DC-DC Converter	Allows bidirectional power flow	Ideal for battery charging/discharging	Complex control	Two-way (Bidirectional)	Interlinking and Grid-Supporting

illustrates overview of DC-DC converters, their advantages, disadvantages, directionality, and power control approaches.

Beyond the fundamental Buck and Boost topologies, higher-order converters such as Ćuk, SEPIC (Single-Ended Primary-Inductor Converter), and ZETA are gaining prominence in hybrid microgrid applications. Unlike basic topologies, the Ćuk and SEPIC converters offer continuous input current, which significantly reduces current ripple, a critical factor for extending the lifespan of sensitive sources like fuel cells and PV arrays. Additionally, SEPIC and ZETA topologies provide the advantage of non-inverted output voltage with buck-boost capability, simplifying the integration of RES with varying voltage levels.

Parallel to these topological advancements, control strategies are also evolving towards more robust, data-independent frameworks. For instance, recent state-of-the-art research has successfully applied data-driven control strategies to converters modeled as Switched Affine Systems (SAS). As demonstrated by Xu et al. [26], these methods can effectively manage time-varying affine terms and ensure convergence to desired limit cycles without requiring precise system dynamics, offering a promising direction for the resilient control of both AC-DC and DC-DC interfaces in microgrids.

2.2.3. Control of Interlinking Converters

Interlinking converters enable bidirectional power transfer between AC and DC parts of hybrid AC/DC MGs [18]. Through controlled power flow management voltage source converters (VSCs) and current source converters (CSCs) maintain energy distribution stability and efficiency [27,28].

Table 4. Overview of interlinking converters.

Converter Type	Description	Advantage	Disadvantage	Direction	Power Control Approach
Voltage Source Converter (VSC)	Converts between AC and DC	High efficiency, power quality control	Requires complex control	Two-way (Bidirectional)	Always Interlinking, Grid-Supporting
Current Source Converter (CSC)	Uses inductors for energy storage	High reliability	Requires large inductors	Two-way (Bidirectional)	Interlinking, Grid-Supporting

illustrates overview of interlinking converters, their advantages, disadvantages, directionality, and power control approaches.

A comprehensive overview of recent studies addressing various control strategies and topologies for interlinking converters in hybrid AC/DC MGs is presented in

Table 5. (a,b) Comparative Overview of Interlinking and bidirectional Converters.

(a)
Ref.	Converter Topology	Control Strategies	Power Rating	THD	Response Time
[20]	LCL filter-based voltage source converter (VSC)	Multi-resonant (3R) Controller-Repetitive Controller (RC)-Droop Control-PI Controller-Phase-Locked Loop (PLL)	1.35 kW (for nonlinear load condition)	Reduced from 12.10% to 3.51% after compensation	Approximately 5 cycles
[29]	Bidirectional Interface Converter (BIC)	Virtual Inertial Control Strategy based on Virtual Synchronous Generator (VSG)	40 kW (DC load) and 30 kW (AC load)	N/A	N/A
[30]	Interlinking Converter (IC) with Voltage Source Inverter (VSI)	Symmetrical Impedance Network (SIN) and Dedicated Modulation Scheme	1.4 kW total	N/A	Not mentioned
[31]	Interlinking Converter (IC) with Voltage Source Inverter (VSI)	Large-Bandwidth Triple-Loop Control (LBTLC)	approximately 2.4 kW	N/A	Islanded mode detection within 1.5 cycles
[32]	Interlinking Converter (ILC) with Hybrid AC/DC Grid	Passivity-Based Control (PBC) with Dual-Droop Control	4 MVA Synchronous Generator, 1 MW & 3 MW DC Sources	N/A	N/A
[33]	Interlinking Converter (ILC) for Hybrid MG	Virtual Synchronous Machine (VSM) Based Droop Control	50 kW hybrid MG system	N/A	N/A
[34]	voltage Source Converter (VSC)-based Interlinking Converter (IC)	Parabolic Relaxation Method for Optimal Power Flow (OPF)	1 MW (DC), 1 MVA (AC) hybrid MG	N/A	Sequential penalization takes 7.34 s; real-time OPF updates occur every 10 s
[35]	Interlinking Converter (VSC)	Local Control Method, Normalized Droop Control	16 kW	N/A	N/A
[36]	Interlinking Converter (VSC)	Voltage-Controlled Method (VCM) with PR Controllers	15 kW	N/A	N/A
[37]	Bidirectional Interface Converter (BIC)	Dual-Side Virtual Inertia Control with Virtual Inductors	1.167 MW (DC MG) and 2 MW (AC MG)	N/A	N/A
[38]	Interlinking Converter (VSC)	Droop Control with Active Power Filtering (APF)	10 kW	Reduced from 13.08% to 4.09%	Improved dynamic response with reduced overshoot
[39]	Parallel LCC-VSC Interlinking Converter	Unified Control with Droop-Based Coordination	N/A	N/A	N/A
(b)
Ref.	limitations		Experimental or Simulation	Goal
[20]	The IC controller requires precise tuning to effectively eliminate harmonics. High-order harmonics (above 18th) may not be fully.		Both Simulation and Experiment	Improve Power Quality-Enhance power flow management
[29]	The effectiveness of virtual inertia control is dependent on the tuning of virtual capacitance parameters. High virtual capacitance values may slow down dynamic response. Implementation may require high computational effort for parallel BICs.		Simulation	Improving AC bus frequency and DC bus voltage inertia to enhance stability of AC-DC hybrid MG
[30]	Requires precise tuning of impedance network parameters. Leakage current suppression is limited to the low-voltage DC side. Power decoupling strategies may be needed for stable operation.		Experimental	Flexible integration of renewable energy into hybrid AC/DC grids with high power density, low leakage currents, and controllable power flow
[31]	Requires precise tuning for seamless transitions between grid-tied and islanded modes. Low-voltage ride-through (LVRT) capabilities depend on predefined threshold settings. Complex control structure may increase implementation difficulty.		Experimental	Achieving multifunctional flexible control of interlinking converters for hybrid AC/DC MGs, ensuring power quality, seamless mode transitions, and low-voltage ride-through capability.
[32]	The passivity-based approach is a sufficient condition but may not always be necessary for stability. Requires accurate modeling of AC and DC bus dynamics. Control tuning is complex due to decentralized nature.		Simulation	Ensuring stable power sharing and voltage/frequency regulation in hybrid AC/DC grids
[33]	Virtual inertia parameter tuning affects system stability and dynamic response. High computational requirements for real-time implementation. Effectiveness under large disturbances needs further validation.		Simulation	Enhancing frequency stability and voltage regulation in hybrid MG systems
[34]	Nonconvex nature of OPF requires sequential penalization for feasible solutions. Voltage phase-angle constraints must be carefully tuned to ensure accuracy. Computational complexity increases for large-scale systems.		Simulation and Hardware-in-the-Loop (HIL) validation	Enhancing computational efficiency of OPF in hybrid AC/DC MGs while ensuring power balance and voltage regulation.
[35]	Normalized droop control has limitations in achieving accurate GPS; Large droop constants may cause instability		Simulation and Hardware-in-the-Loop (HIL) Experiments	Improve power sharing accuracy, enhance system stability
[36]	Requires precise tuning of PR controllers; Complexity increases with additional MGs		Both Simulation and Experiment	Improve power quality by reducing frequency deviation and voltage unbalance
[37]	Requires precise tuning of virtual inertia parameters; Complex implementation for real-world deployment		Simulation-based study with small-signal modeling	Enhance stability of hybrid AC-DC MGs by mitigating frequency fluctuations and voltage dips
[38]	Requires precise tuning of droop coefficients; Performance is affected by nonlinear load variations		Simulation and Real-Time Digital Simulator (RTDS) Experiments	Improve power quality by mitigating harmonics and ensuring stable power sharing
[39]	Requires precise droop coefficient tuning; Complex implementation for hybrid AC/DC networks.		Real-Time Simulation	Improve power sharing, enhance stability, mitigate commutation failure in LCC

N/A = Not Applicable.

a,b. The table summarizes the most relevant works reported in the literature [20,29,30,31,32,33,34,35,36,37,38,39], covering converter configurations, applied control methods, system ratings, performance metrics, and implementation frameworks, which collectively provide valuable insights into the evolution of interlinking converter designs and their operational objectives in modern hybrid MGs.

2.2.4. Multilevel Converters (MLCs)

Multilevel converters enhance power quality and efficiency which is especially beneficial for high-power applications [40]. The neutral point clamped (NPC) [41,42], flying capacitor (FC) and flying capacitor dual-output (FCDO) [43,44], and cascaded H-Bridge (CHB) [45] converters produce various voltage levels which minimize harmonic distortion. The modular nature and scalable design of modular multilevel converters (MMCs) make them ideal for use in large-scale grid systems [46].

Table 6. Overview of multilevel converters.

Converter Type	Description	Advantage	Disadvantage	Direction
Neutral Point Clamped (NPC) Converter	Uses diodes to create multiple voltage levels	Lower switching losses	Requires additional components	Two-way (Bidirectional)
Flying Capacitor (FC) Converter	Uses capacitors for voltage levels	Better voltage balancing	More complex	Two-way (Bidirectional)
Cascaded H-Bridge (CHB) Converter	Uses series-connected H-bridges	Modular, scalable	Requires isolated power sources	Two-way (Bidirectional)
Modular Multilevel Converter (MMC)	Uses multiple submodules	High efficiency, reduced harmonics	High cost, complexity	Two-way (Bidirectional)

illustrates overview of multilevel converters, their advantages, disadvantages, directionality, and power control approaches.

To provide a coherent comparison of the converter technologies most widely deployed in hybrid AC/DC MGs,

Table 7. Comparative analysis of different power converters in previous work.

Converter Type	Load Dynamics	Distribution Performance	Harmonic Mitigation	Voltage Stability	Bidirectional Capability	Fault Tolerance	Validation	Key Feature
CSC [46]	√	✗	√	√	✗	✗	Sim	Power stability
VSC [53]	√	√	√	√	✗	√	Both	High quality
Boost [56]	√	√	✗	√	√	✗	Sim	PV extraction
Bidirectional DC-DC [59]	√	√	✗	√	√	✗	Sim	Power management
NPC [66]	√	√	√	√	✗	√	Both	Voltage balancing
FC [73]	√	√	√	√	✗	√	Both	DC-link quality
CHB [75]	√	√	√	√	√	✗	Sim	Voltage balancing
MMC [82]	√	√	√	√	√	√	Sim	Modular stability

Validation: Simulation/Experimental. √: Available. ✗: Not available

summarizes their performance across four primary operational criteria, namely load-dynamic behavior, distribution-side performance, harmonic mitigation, and voltage-stability support. In addition, bidirectional capability, fault tolerance, and the availability of experimental or simulation validation are included to highlight their practical applicability. For clarity and conciseness, only representative converter families are presented in the main text, while the complete extended table is provided in Appendix A. The detailed comparative analysis in Appendix A compiles the key results reported in previous studies [46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87].

The summarized results indicate that VSC-based and multilevel converter topologies (NPC, FC, CHB, and MMC) provide the strongest overall performance, offering excellent harmonic mitigation, voltage stability, and scalability. In contrast, CSCs deliver reliable current regulations but exhibit weaker distribution-side flexibility and limited bidirectional operation. Simpler single-switch converters such as Boost and conventional bidirectional DC-DC converters maintain good load-dynamic response but inherently lack strong harmonic suppression and fault-tolerant capability. These observations align with the broader trend toward multilevel and modular topologies in medium- and high-power hybrid microgrid applications.

3. Control Strategies for Power Converters in AC/DC MGs

The combination of renewable energy systems, storage, and diverse loads in HMGs requires advanced converter control to ensure efficient energy management, operational stability, and responsiveness to hybrid dynamics. The strategy that regulates a converter’s power output is often termed local, primary, or power-converter control [88]. The control strategies applied to these converters can be categorized into conventional control strategies and advanced or modern control strategies, as illustrated in Figure 3.

From a practical implementation standpoint, the control strategies used in hybrid AC/DC MGs are executed on dedicated digital control hardware to satisfy strict real-time requirements. Primary control loops, responsible for fast current and voltage regulation are typically implemented on high-performance microcontrollers (MCUs), Digital Signal Processors (DSPs), or increasingly on FPGA/SoC platforms when computationally intensive algorithms such as MPC or VSM-based methods are employed. These controllers operate at sampling frequencies synchronized with the converter switching frequency, commonly ranging from 10 kHz to 50 kHz. In contrast, secondary and tertiary control layers, which address voltage/frequency restoration and optimal power flow, operate at slower timescales (milliseconds to minutes) and can be implemented on supervisory controllers or industrial PCs. This hierarchy ensures that fast dynamics are handled locally while higher-level coordination functions operate under less stringent real-time constraints.

3.1. Hierarchical Control Architecture and Resiliency

HMG control is typically structured into three layers separating instantaneous regulation from slow economic coordination:

• Primary (Local) Control: Operating at millisecond scale, it handles inner current loops, voltage/frequency formation and stabilization, and fast transients. At this layer, converters are categorized by function:
  • GF set V/f references
  • GFL inject current to track the grid
  • IPCs manage bidirectional AC/DC power exchange
• Secondary and Tertiary Control: Running at seconds–minutes, these layers use communication (for instance, IEC 61850 [89], TCP/IP) to restore nominal V/f (secondary) and execute optimal power flow and economic dispatch (tertiary).

Because secondary/tertiary layers depend on external communication, cyber-physical threats such as false data injection (FDI) can corrupt reference signals, causing power-sharing errors and instability; resilience must therefore be embedded into the control design [90,91,92].

Table 8. Hierarchical control layers and their functional characteristics in MG operation.

Control Layer	Function and Scope	Typical Variables Controlled	Control Speed
Primary (Local) Control	Voltage/frequency formation, current/power injection, transient response.	Active/Reactive Power (P/Q), Local Voltage (V), Current (I).	Milliseconds
Voltage/frequency formation, current/power injection, transient response.	Voltage/frequency restoration, accurate power sharing, error elimination.	Voltage Reference Adjustments, Frequency Correction Signal.	Seconds
Active/Reactive Power (P/Q), Local Voltage (V), Current (I).	Optimal power flow (OPF), economic dispatch, inter-MG power trading.	Real/Reactive Power Flow Reference to Main Grid/Utility.	Minutes/Hours

shows the control layers for HMG in power systems.

A critical aspect of hierarchical control in hybrid microgrids is the interaction between fast converter-level control and higher-level decision-making functions, particularly under massive RES penetration. Converter-based strategies such as droop control, VSM, and MPC operate at fast time scales to ensure dynamic stability, but they ultimately act on the setpoints generated by Tertiary-level economic dispatch and scheduling. Under high-RES uncertainty, the formulation of these higher-level decisions becomes as influential as the converter dynamics themselves.

Recent work by Liang et al. [93] demonstrates this coupling through a data-driven quad-level robust optimization framework that explicitly models RES uncertainty and identifies the least conservative operating strategy for hybrid AC/DC microgrids. Their results show that system-level decisions, such as day-ahead commitments and interlink converter scheduling are tightly constrained by converter operating limits and protection mechanisms. This highlights that converter-level control and system-level optimization cannot be treated independently; reliable HMG operation requires a robust economic plan from the upper layers and stable, resilience-oriented execution from the converter controllers.

3.2. Conventional Limitations and the Need for Virtualization

3.2.1. Limitations of Classical Droop Control

Classical P–f and Q–V droop enable autonomous sharing without high-bandwidth communications, but they implicitly assume inductive line impedance. In LV/MV MG, resistance is often comparable to or larger than inductance, breaking active/reactive decoupling and leading to inaccurate sharing, circulating currents, and potential instability [94]. Similarly, DC-voltage droop applied to multiple IPCs can induce circulating power and overstress due to line resistance [95].

3.2.2. Adaptive and Virtualization Techniques

To resolve the line impedance dilemma and compensate for the lack of mechanical inertia in converter-dominated systems, research has focused on control methods that synthesize the necessary electrical properties.

• Adaptive Droop Control: This technique dynamically adjusts droop coefficients based on real-time system states (for instance, voltage deviation), overcoming the static nature of classical droop. Adaptive methods demonstrate significant improvements in current and power sharing accuracy, stability margins, and faster transient response times compared to static methods [96].
• Virtual Impedance: The most direct solution to impedance mismatch is the Adaptive Virtual Impedance Droop Control (AVIDC). This mechanism dynamically introduces a synthetic (usually inductive) impedance into the control loop. This action effectively decouples $P$ and $Q$ control, ensuring accurate reactive power sharing by balancing the converter’s output with the virtual line drop [97].
• Virtual Inertia Control (VIC): To enhance damping and mitigate rapid frequency deviations, the Virtual Synchronous Machine (VSM) or Virtual Synchronous Generator (VSG) is implemented in GFCs and IPCs. VSM algorithms digitally emulate the mechanical inertia of traditional generators, providing active support to the AC bus frequency and DC bus voltage stability.

3.3. Modern Predictive and Optimized Control

The complexity and high functional demands of adaptive/virtual control have driven research toward optimization and learning-based techniques that offer superior dynamic performance and native constraint management.

3.3.1. Model Predictive Control (MPC)

MPC is an advanced model-based technique that predicts the system’s future state over a finite horizon and selects the optimal control action by solving an optimization problem subject to operational constraints.

• Performance: MPC, particularly Finite Control Set MPC (FCS-MPC), demonstrates improved transient and steady-state voltage/frequency responses, along with reduced Total Harmonic Distortion (THD) compared to conventional methods [98].
• Limitation: MPC’s main barrier to widespread scalability is its high online computational burden, which often involves solving complex mixed-integer linear/quadratic programs (MILP/MIQP) in real-time.

3.3.2. Data-Driven and Hybrid Learning

Reinforcement Learning (RL) and other learning methods offer an alternative by learning optimal policies through interaction with the environment, making them highly adaptive to dynamic conditions. RL can be preferable for its rapid decision-making ability in managing a large number of energy systems (scalability), although it requires substantial training data [99].

The current trend is Hybridization (Learning-Enhanced MPC). By integrating RL into the MPC framework, the model can simplify the computationally intensive online optimization problem (such as reducing an MILP/MIQP to a standard LP/QP). This synergy significantly reduces the computational time, enabling the real-time deployment of complex optimal control strategies that MPC alone could not handle [100].

Table 9. Comparison of advanced control strategies for data-driven and hybrid learning in energy systems.

Strategy	Key Advantage	Primary Limitation	Computational Burden	Typical Application
Model Predictive Control (MPC)	Excellent dynamic response, handles constraints naturally, low THD [98]	High computational complexity, sensitivity to model accuracy	High	High-power VSCs, stringent power quality demands
Reinforcement Learning (RL)	Highly adaptive, quick decision-making post-training [99]	Requires extensive training data, lack of guaranteed stability	High (Training), Medium (Online)	Energy management, complex systems where scalability is prioritized
Sliding Mode Control (SMC)	High robustness against uncertainties and disturbances	Chattering phenomenon (rapid switching), requires detailed system knowledge	Medium	Fault-tolerant control, systems with high parameter variations

compares three main control methods in HMGs.

3.4. Robust Control and Resilience

To ensure performance despite internal physical faults, load fluctuations, and external cyber intrusions, robust and non-linear control methods are essential.

• Robust non-linear control: Methodologies like $H_{\infty }$ control mathematically guarantee performance by minimizing the maximum possible effect of disturbances, a critical requirement for coordinating complex IPCs under uncertainty [101]. Passivity-Based Control (PBC) ensures stability by leveraging the inherent energy conservation properties of the physical system.
• Sliding mode observer (SMO) for fault-tolerance: A key resilience application is integrating observers into the secondary control layer. A Sliding Mode Observer (SMO) can accurately reconstruct sensor or actuator faults (for instance, magnitude or harmonics) in real-time, functioning as a diagnostic tool. This Fault-Tolerant Control (FTC) architecture uses the reconstructed fault value to actively offset the influence of the fault, ensuring the system maintains operational performance, restores voltage, and maintains proportional current sharing even when components or communication links are compromised [92].

Beyond internal fault tolerance, these robust frameworks exhibit intrinsic capabilities to reject external stochastic disturbances. For instance, $H_{\infty }$ control strategies are particularly effective at attenuating grid-side harmonics and mitigating the impact of high-frequency noise on control performance1. Similarly, the inherent robustness of Passivity-Based Control (PBC) and Sliding Mode Control (SMC) provides a natural buffer against the rapid variability of RES generation, maintaining stability even when operating points shift abruptly. Furthermore, observer-based techniques (such as SMO) extend this resilience to the cyber-physical layer; by treating measurement anomalies and mild cyber events such as False Data Injection (FDI), as bounded disturbances, they can reconstruct and compensate for these errors in real-time, ensuring continuous operation without triggering unnecessary protection trips.

3.5. Specialized Interlinking Converter (IPC) Control

As the nexus of the HMG, the IPC requires complex, multi-functional control:

• Coordinated control and mitigation: IPCs must manage smooth, bidirectional power transfer and seamless mode transition between grid-tied and islanded operation (for instance, rapid islanded detection within 1.5 cycles). Research has developed solutions for highly coordinated systems, such as employing a superimposed frequency on the DC sub-grid to achieve autonomous, proportional power sharing in systems with multiple IPCs, effectively mitigating circulating power flows caused by line resistance [95].
• Power quality: IPC control also integrates selective controllers (for instance, Multi-resonant, Repetitive Controllers) with primary control to actively track and eliminate harmonic distortions (documented THD reduction from 12.10% to 3.51%), ensuring the entire system meets power quality standards.

3.6. Quantitative Synthesis of Performance

To move beyond qualitative descriptions and enable objective comparisons, this section presents a quantitative synthesis of performance metrics for key converter topologies and control strategies found in hybrid AC/DC MGs. The evaluation of power electronic converters hinges on a set of Key performance indicators (KPIs) that dictate their suitability for specific applications. These KPIs include steady-state efficiency, which governs operational losses; THD, which quantifies the impact on power quality; and dynamic response time, which determines the converter’s ability to react to transient events and maintain system stability [102]. This analysis focuses on the IC, the pivotal component managing power flow between the AC and DC subgrids, as well as other significant topologies such as the MMC, which is increasingly favored in high-power applications for its modularity, scalability, and superior harmonic performance [102].

A holistic examination of these metrics, as consolidated in

Table 10. (a,b) The fundamental engineering trade-offs inherent in converters according to previous works.

(a)
Ref.	Converter Topology/Application	Control Strategy	Rated Power (kW/MVA)	Voltage Level (V/kV)
[13]	Bidirectional Interlinking Converter (BIC)	Virtual Inertia Emulation	Not Specified	Weak Grid Conditions
[36]	Single Interlinking Converter (IC)	PLL-less Voltage Controlled Method (VCM)	~5 kVA (simulation)	220 V (AC), 400 V (DC)
[103]	High Step-Up DC-DC Converter (SISO)	Dynamic Modeling, Control Strategy	500 W (prototype)	Not Specified
[104]	Smart Grid Inverter	Hierarchical Control Strategy (HCS)	Not Specified	Not Specified
[105]	Unified Power Quality Conditioner (UPQC)	Gated Recurrent Unit (GRU) Controller	Not Specified	Not Specified
[106]	Bidirectional Interlinking Converter (IC)	Decentralized Control with APF	100 kVA (simulation)	415 V (AC), 700 V (DC)
[107]	Modular Multilevel Converter (MMC)	Interleaved Half-Bridge Sub-Modules	Not Specified	Low/Medium Voltage
[6]	Solid-State Transformer (SST)	Not Specified	1000 kVA	MV/LV
(b)
Ref.	Efficiency (%)	THD (%)	Response/Settling Time (ms)	Switching Frequency (kHz)
[13]	>97% (implied)	<5% (implied)	>45% improvement in frequency deviation	Not Specified
[36]	Not Specified	<2% (under load transient)	~20 ms	10
[103]	96.4% (at 200 W)	Not Applicable	Not Specified	50
[104]	Not Specified	<5% (oscillation)	<400	Not Specified
[105]	Not Specified	0.04, 0.25, 0.98	Not Specified	Not Specified
[106]	Not Specified	<5% (as per IEEE 519)	~20 ms (for THD correction)	10
[107]	>98% (typical for MMC)	<3% (typical for MMC)	Not Specified	Varies (low effective freq.)
[6]	Slightly lower than passive transformer	Not Specified	Not Specified	High (for HF transformer)

a,b, reveals the fundamental engineering trade-offs inherent in converter design. For example, the pursuit of superior dynamic performance (for instance, faster settling times) and lower THD often necessitates higher switching frequencies. While this allows for smaller passive components (inductors and capacitors) and improved control bandwidth, it concurrently leads to increased switching losses in the semiconductor devices, which can adversely affect the converter’s overall efficiency [103]. Conversely, optimizing for maximum efficiency might involve lowering the switching frequency, which could compromise power quality and dynamic response. This interplay is not merely theoretical; it is a practical design constraint demonstrated by the empirical data extracted from the literature. By juxtaposing these, performance metrics for various converter designs, these trades-offs become explicit, providing a data-driven foundation for a nuanced discussion on converter selection and optimization. This approach transforms a simple catalog of technologies into a practical tool for engineering analysis, allowing researchers and designers to weigh competing objectives based on the specific requirements of their target application.

Given the diversity of control architectures discussed, selecting the optimal strategy requires a careful evaluation of specific system constraints, such as the available computational power, the resistive nature of the network, and the necessity for constraint handling. To facilitate this decision-making process for practitioners,

Table 11. Summary of Applicability and Preference for Advanced Control Strategies.

Control Strategy	Primary Use Case	When Is It Preferable?	Key Trade-Off
Adaptive Droop Control	Power sharing in systems with varying load conditions or line impedances.	Preferable when high sharing accuracy is needed without complex communication infrastructure.	Requires careful tuning of adaptation gains to avoid instability during large transients.
Adaptive Virtual Impedance (AVIDC)	Low-voltage (resistive) microgrids where P/Q coupling is an issue.	Preferable for decoupling active/reactive power control and ensuring accurate reactive power sharing in resistive lines	Can slightly reduce the effective voltage range at the point of common coupling.
Model Predictive Control (MPC)	High-performance VSCs requiring fast dynamic response and constraint handling.	Preferable when the system has strict operational constraints such as, current limits and requires excellent transient response and low THD.	High computational burden; performance is highly sensitive to parameter mismatches in the internal model.
Hybrid Learning (RL + MPC)	Complex, large-scale systems where standard MPC is too slow.	Preferable for real-time implementation of optimal control where online optimization such as MILP is computationally prohibitive.	Requires extensive offline training data and a robust training phase before deployment.

provides a comparative synthesis of the advanced control methods reviewed. It outlines the primary use cases, preferable application scenarios, and critical trade-offs for each strategy, distinguishing between methods best suited for decentralized power sharing such as, Adaptive Droop, versus optimized for high-performance dynamic regulation such as, MPC and Hybrid Learning.

4. Converters for Low Voltage (LV) and Medium Voltage (MV) Applications

The design of hybrid AC/DC MGs necessitates various converters that depend on voltage levels as well as power requirements and system applications. LV converters operate below 1 kV to support power conversion and distribution in residential, industrial and commercial MGs. Power distribution networks as well as renewable energy systems and industrial automation applications rely heavily on MV converters that operate between 1 kV and 35 kV. Selecting a converter requires consideration of the system’s unique requirements including AC-DC conversion capabilities, DC voltage regulation needs, bidirectional energy flow capabilities, and AC-DC interconnection features for effective power management [88].

Rectifiers represent essential converters that transform AC power into DC power and find widespread application in LV systems to energize battery chargers, uninterruptible power supplies (UPS), electric vehicle charging stations (EVCs), and PV systems [108]. DC storage systems dominate many LVMGs, which makes rectifiers vital for achieving efficient energy conversion. Inverters serve as the conversion devices when DC power must be transformed into AC power. Inverters change DC power into AC to support grid-connected renewable energy systems as well as motor drives and power substations that demand stable AC-DC power transfer. In medium voltage applications, inverters become vital for integrating grid-connected renewable energy systems and managing industrial motor drive control.

DC-DC converters regulate voltage levels within DC subgrids while rectifiers and inverters handle AC-DC conversion. DC-DC converters play a crucial role in managing battery storage systems and power conversion for electric vehicles and they ensure voltage regulation in MGs for LV applications [109]. DC-DC converters are rarely used in MV systems because transformers and AC-DC conversion methods are more commonly used for controlling voltage and distributing power. However, in high-power systems like MVDC grids, MV DC-DC converters help directly transfer DC power, making the system more efficient.

Bidirectional DC-DC converters stand as an essential category of converters because they enable efficient power transfer between LVMGs and storage units like battery energy storage systems (BESS), super capacitor-based storage solutions, and renewable energy systems. The converters provide essential support for bi-directional power flow, which guarantees uninterrupted energy transfer among distributed energy resources and the power grid. Bidirectional DC-DC converters remain uncommon in MV applications due to the preference for AC-coupled systems with transformers and high-power inverters for energy storage at medium voltage levels [110,111].

VSCs serve as fundamental elements in LV and MV applications beyond basic power electronic devices like rectifiers, inverters, and DC-DC converters. They serve as standard components in UPS systems and solar inverters for LVMGs by ensuring dependable and efficient AC-DC interconnections. VSCs serve as essential components in MV applications such as HVDC systems and FACTS while also playing a key role in maintaining power quality and voltage regulation for large-scale industrial processes [112]. On the other hand, CSCs serve as the primary solution for MV industrial motor drives and grid control applications because they deliver dependable operation in high-power situations [113].

MLCs improve power quality in MV systems by producing multiple voltage levels, which diminish harmonics and enhance system efficiency. MLCs do not see much use in LV applications but become essential for high-voltage DC transmission systems as well as industrial motor drives and STATCOM grid stabilization. The MMC represents the forefront of multilevel converter technology and demonstrates superior performance in HVDC transmission alongside MV motor drives and grid-forming functions [114]. The choice of converters for LV and MV MG systems relies on specific system requirements along with voltage regulation requirements and power conversion characteristics.

Table 12. Comparison of converter types for LV and MV applications.

Converter Type	LV Applications (≤1 kV)	MV Applications (1 kV–35 kV)
AC-DC Converters (Rectifiers)	Used in LV MGs, battery chargers, power supplies	Used in MV substations, industrial applications
DC-AC Converters (Inverters)	Used in solar inverters, UPS, motor drives	Used in MV grid-tied renewable systems, industrial motor drives
DC-DC Converters	Used in LV battery storage, EVs, MGs	Rare, but sometimes used in MV DC grids
Bidirectional DC-DC Converters	Used in energy storage (batteries, super capacitors), EVs	Rare, as MV systems typically use transformers for power conversion
Voltage Source Converters (VSCs)	Used in LV MG interlinking, UPS, and PV inverters	Used in MV FACTS (Flexible AC Transmission Systems), HVDC links
Current Source Converters (CSCs)	Rare in LV applications	Used in MV motor drives, grid control
Multilevel Converters (MLCs)	Rarely used in LV	Common in MV applications (HVDC, STATCOM, industrial motor drives)
Modular Multilevel Converter (MMC)	Not typically used in LV	Used in HVDC, STATCOM, large MV drives

summarizes the usage of different converter types in LV and MV applications.

The distinction between LV and MV systems in hybrid MGs extends far beyond a simple numerical difference in operating voltage. It represents a fundamental divergence in design philosophy, technological implementation, safety protocols, and regulatory compliance. As defined by international standards bodies such as the International Electro technical Commission (IEC) and the American National Standards Institute (ANSI), LV systems typically operate at voltages below 1 kV AC, while MV systems span the range from 1 kV to approximately 36 kV [115]. This classification dictates a cascade of engineering decisions, from the choice of semiconductor materials to the type of protection equipment employed.

MV systems are the backbone of industrial power distribution and utility-scale renewable energy integration. Their design is dominated by the challenges of managing high electric fields. This necessitates sophisticated insulation media such as sulfur hexafluoride (SF6) gas or vacuum, along with significantly larger physical clearances (creep age and clearance distances) to prevent electrical breakdown [116]. The immense energy associated with MV faults demands robust and highly reliable protection schemes, typically involving vacuum or gas circuit breakers controlled by advanced digital relays [115]. Consequently, the design, installation, and maintenance of MV systems require specialized knowledge and certified personnel, reflecting their critical role in high-power infrastructure [116].

In contrast, LV systems are tailored for end-user applications, such as residential buildings, commercial complexes, and light industrial facilities. Here, the design priorities shift towards safety for non-specialist users, modularity for ease of installation, and cost-effectiveness for mass deployment. Insulation is typically air-based, and protection is achieved through standardized and widely available devices like Molded Case Circuit Breakers (MCCBs) and Miniature Circuit Breakers (MCBs) [116].

This top-level system classification has a profound and direct impact on the selection of power semiconductor technology. The distinct operational environments of LV and MV converters create separate ecological niches where different wide-bandgap (WBG) semiconductor materials excel. For LV converters, particularly in applications like solar micro inverters or electric vehicle chargers, there is a strong drive towards higher power density, which requires increasing the switching frequency to reduce the size of passive components like inductors and capacitors [117]. Gallium Nitride (GaN) devices, with their exceptionally high electron mobility and lower switching losses compared to silicon (Si) and even Silicon Carbide (SiC) in the sub-1 kV range, are ideally suited for these high-frequency applications [118]. The superior switching performance of GaN enables the design of more compact, efficient, and lightweight LV power converters.

Conversely, the primary challenge for MV converters is the ability to block high voltages and conduct large currents while managing thermal dissipation. SiC possesses material properties that are far superior to both Si and GaN in this domain. Its wide bandgap results in a much higher critical breakdown electric field (approximately 10 times that of Si), and its high thermal conductivity allows for more effective heat removal from the device junction [118]. These characteristics enable the fabrication of SiC MOSFETs and diodes that can operate at higher voltages, temperatures, and power levels than their counterparts, making SiC the enabling technology for the next generation of compact and efficient MV converters and solid-state transformers [119]. Thus, the initial system-level decision to operate in the LV or MV domain sets in motion a chain of technological consequences that reaches down to the fundamental choice of semiconductor material. The comparative framework is summarized in

Table 13. Comparative analysis of LV and MV converter design considerations.

Parameter	Low-Voltage (LV) Systems	Medium-Voltage (MV) Systems
Voltage Class (IEC/ANSI)	Up to 1 kV AC/1.5 kV DC [116]	1 kV to 36 kV (up to 72.5 kV in some standards) [115]
Typical Power Ratings	Watts to ~250 kW (e.g., residential/commercial) [116]	>250 kW to Multi-MVA (e.g., industrial, utility-scale) [115]
Dominant Semiconductor Tech.	Si-MOSFET/IGBT. Increasingly Gallium Nitride (GaN) for high-frequency, high-efficiency applications [117]	Si-IGBT/GTO. Increasingly Silicon Carbide (SiC) for higher voltage/temperature and IGCTs for high power [119]
Key Interconnection Standards	IEEE 1547 for Distributed Energy Resources (DERs) interconnected with distribution networks [120]	Utility-specific transmission and distribution grid codes (e.g., BDEW, ENTSO-E); IEEE C37 series for switchgear [116]
Insulation Requirements	Primarily air-insulated with standard component clearances. Focus on user safety and accessibility [116]	Requires specialized insulation media (SF6, vacuum, oil) and larger creepage/clearance distances to manage high electric fields [116]
Primary Protection Devices	Fuses, Miniature Circuit Breakers (MCBs), Molded Case Circuit Breakers (MCCBs) [116]	Vacuum/SF6 Circuit Breakers, advanced digital protection relays, high-rupturing capacity fuses [115]
Design Philosophy	Focus on modularity, cost-effectiveness, high power density, and compliance with standardized interconnection rules for mass deployment.	Focus on high reliability, robustness, operational safety, and bespoke engineering for critical, high-power infrastructure.

.

4.1. Case Studies in Application

To illustrate the practical implications of the LV and MV design paradigms, this section presents two representative case studies drawn from the literature. These examples highlight how the application context dictates the specific technological choices and performance priorities for the converter system.

4.1.1. LV Residential PV Micro Inverter

The residential PV sector provides a quintessential example of an LV hybrid MG application. These systems are designed to operate on the end-user side of the distribution network, typically interfacing with single-phase or three-phase LV grids (for instance, 120/240 V or 230/400 V). The power ratings of these systems are modest, with the vast majority of installations falling between 3 kW and 5 kW [121].

The design of converters for this application is driven by several key factors: cost, efficiency, reliability, and strict adherence to grid interconnection standards. The primary governing standard in North America and influential globally, is IEEE 1547 [122]. This standard mandates specific requirements for the converter’s response to abnormal grid voltage and frequency conditions, anti-islanding protection (ceasing to energize the grid within 2 s of a utility outage), and power quality (for instance, limits on DC current injection and harmonic distortion) [120].

A significant trend in this sector is the shift from centralized string inverters to module-level power electronics (MLPE), such as micro inverters or DC power optimizers [123]. A micro inverter is a small, dedicated converter attached to each individual PV panel. This architecture allows for maximum power point tracking (MPPT) on a per-panel basis, which maximizes the overall energy harvest, especially in installations with partial shading or varying module orientations [124]. While this approach increases the number of converters, it improves system resilience and energy yield.

The performance of these systems is not judged solely on peak converter efficiency but on the total annual energy production relative to the available solar resource. This is quantified by the performance ratio (PR), which accounts for all system losses, including those from the inverter, cabling, temperature effects, and soiling. A large-scale study of residential PV systems in Belgium found a mean PR of 78%, indicating that a typical system produces 22% less energy than an ideal, lossless system under the same conditions [121]. This case study exemplifies the LV design philosophy: a mass-market, standards-driven product where overall energy yield and safety, governed by standards like IEEE 1547 [125] are paramount.

4.1.2. MV Converter for Offshore Wind Farm Collection

Offshore wind farms represent a frontier application for MV power conversion technology, operating at a scale and in an environment that places extreme demands on equipment. Modern offshore wind turbines are massive power generation units, with individual ratings reaching and exceeding 10 MW. To transmit this power back to shore efficiently, the energy from multiple turbines is aggregated in an internal medium-voltage grid before being stepped up to high-voltage for the main transmission link. Increasingly, this internal collection grid is being designed as an MVDC system to reduce cable losses and improve controllability [126].

A key enabling component in this architecture is the high-power DC-DC converter located at each turbine. This converter must step up the variable DC voltage from the turbine’s generator-side rectifier (typically in the range of 1 kV to 6 kV DC) to the MVDC collection bus voltage (for instance, 30 kV to 60 kV DC) [126]. The power rating of this converter matches that of the turbine, meaning it must handle multi-megawatt power flows continuously.

The primary design driver for this application is reliability. The cost of maintenance and the economic loss from downtime in an offshore environment are prohibitive, making component robustness and longevity critical [127]. To achieve the required power and voltage ratings, designers often employ advanced semiconductor devices like Integrated Gate-Commutated Thyristors (IGCTs). These press-pack devices can handle very high currents (up to 5000 A) and voltages (up to 6.5 kV) without the need for paralleling, which simplifies the converter design and eliminates a common failure mode associated with paralleled IGBT modules. The design must also account for the harsh marine environment, with special considerations for humidity, salt spray, and mechanical vibrations. Redundancy is often built into critical subsystems, such as the water-cooling pumps, to ensure continuous operation [127].

This case study is a clear illustration of the MV design philosophy. It involves bespoke, high-value engineering for a critical infrastructure application. The focus is on maximizing reliability and availability over a long service life, where the initial capital cost of the highly robust converter system is justified by the avoidance of extremely expensive offshore interventions and lost revenue from energy production.

5. Fault Behavior and Advanced Protection Schemes

The reliable operation of hybrid AC/DC MGs is contingent upon robust and rapid fault management systems. The unique characteristics of DC circuits, namely the lack of a natural current zero-crossing and the low impedance of the network, present significant challenges that render traditional AC protection schemes inadequate [128]. This section provides a systematic taxonomy of fault types, analyzes the inherent response of different converter topologies, and evaluates the performance of advanced protection devices and control strategies that are essential for ensuring the safety and stability of these complex systems.

5.1. Fault Characteristics in HMGs

5.1.1. Classification of Fault Types

Faults within the DC portion of a HMG can be broadly categorized into two main types, each with distinct electrical characteristics and implications for the protection system [129].

• Pole-to-Pole (PP) Faults: A pole-to-pole fault is a low-impedance, or bolted, short circuit that occurs directly between the positive and negative DC conductors. This is the most severe type of DC fault. Upon fault inception, the large DC-link capacitors associated with the power converters discharge rapidly and uncontrollably into the fault path. This results in an extremely high-current surge with a very steep rate of rise $\frac{di}{dt}$, capable of causing catastrophic damage to semiconductor devices within microseconds if not interrupted swiftly [130].
• Pole-to-Ground (PG) Faults: A pole-to-ground fault occurs when either the positive or the negative conductor makes an unintended connection to the system ground. These faults are typically of higher impedance than PP faults, as the fault path may include resistive elements. While the resulting fault current is generally lower, PG faults can be more difficult to detect, particularly in ungrounded or high-resistance grounded systems [131]. The ability to detect these faults is critical for personnel safety and to prevent the evolution of a PG fault into a more severe PP fault.

5.1.2. Inherent Converter Fault Response

The intrinsic topology of a power converter fundamentally dictates its behavior and vulnerability during a DC-side fault. The two primary classes of converters, VSCs and CSCs, exhibit starkly different responses.

• Voltage Source Converters (VSCs): VSCs, which are ubiquitous in modern Mgs due to their flexible control capabilities, are exceptionally vulnerable to DC faults. A VSC topology includes a large DC-link capacitor bank and uses switches (like IGBTs) with anti-parallel diodes. During a DC-side fault, even if the IGBTs are immediately turned off, the fault creates a direct path for the DC-link capacitors to discharge through the anti-parallel diodes of the converter bridge [130]. This process is completely uncontrolled, leading to a massive current spike that can easily exceed the surge current rating of the diodes, causing their destruction. The fault current rises extremely rapidly, driven by the low impedance of the capacitor discharge path [132]. This makes the protection of VSCs one of the most critical challenges in DC MG design.
• Current Source Converters (CSCs): In contrast, CSCs are inherently more robust to DC faults. The defining feature of a CSC is the large DC-link inductor connected in series with the DC source. This inductor serves to maintain a relatively constant DC current. During a fault, this inductor naturally opposes any rapid change in current, thereby significantly limiting the rate of rise of the fault current $\frac{di}{dt}$ [132]. This inherent current-limiting behavior provides a crucial time window, typically milliseconds instead of microseconds for the protection system to detect the fault and actuate a breaker, preventing the fault current from reaching destructive levels. This characteristic makes CSCs a more fault-tolerant option, though they are often less flexible in control and have higher conduction losses compared to VSCs.

5.2. Advanced Protection Devices and Control Actions

The demanding protection requirements of DC MGs, particularly those employing VSCs, have driven the development of new hardware and control philosophies that move beyond simple overcurrent tripping.

5.2.1. The Rise of Solid-State Circuit Breakers (SSCBs)

Traditional protection devices like fuses and mechanical circuit breakers (MCBs) are fundamentally too slow to protect VSC-based systems. The tripping time of an MCB, which involves mechanical movement and arc quenching, is on the order of tens of milliseconds. In that time, the fault current from a VSC’s capacitor discharge would have already destroyed the converter’s semiconductor devices [130].

The enabling technology for reliable DC MG protection is the SSCB [133]. SSCBs utilize fast-switching power semiconductors, such as IGBTs, IGCTs, or emerging SiC and GaN devices to interrupt the current path. Lacking any moving parts, their response is exceptionally fast, with fault clearing times ranging from a few microseconds down to hundreds of nanoseconds [134]. This ultra-fast interruption capability allows the SSCB to extinguish the fault current before it rises to a level that could damage the sensitive power electronics in the converters. The development and deployment of cost-effective and reliable SSCBs are therefore considered a critical prerequisite for the widespread adoption of DC MGs [133].

5.2.2. Control Actions from Protection to Resilience

Modern grid codes and the operational needs of MGs demand more than just fault isolation; they require resilience. The concept of FRT has become a mandatory requirement for grid-connected resources in many jurisdictions [135]. FRT dictates that a MG’s converters must not disconnect from the utility grid during transient voltage sags caused by external faults but must instead remain online and actively support the grid by, for example, injecting reactive power to help stabilize the voltage [135]. This shifts the paradigm from a simple protect and disconnect philosophy to a survive and support control action.

Following the clearance of a fault, the MG’s control system must manage the post-fault recovery process. This involves rapidly and stably restoring the MG’s internal voltage and frequency (in islanded mode) to their nominal values. The recovery time is a critical measure of system resilience and has been shown to be highly sensitive to network parameters, particularly line inductance, which can affect the damping of post-fault oscillations [136]. Advanced control strategies, sometimes coordinated through multi-agent systems, are being developed to optimize this restoration process, ensuring a swift return to normal operation while maintaining power supply to critical loads [137].

The relationship between advanced protection hardware and advanced control software is deeply synergistic. The ultra-fast fault isolation provided by an SSCB is the critical first step that ensures the physical survival of the power converters. By interrupting the destructive fault current in microseconds, the SSCB preserves the integrity of the system’s hardware [134]. This action creates a stable, albeit post-disturbance, environment in which the converter’s control system can then execute its pre-programmed FRT strategy. Without the SSCB’s rapid intervention, the converter would be destroyed, and any FRT control algorithm would be moot [130]. Therefore, the SSCB provides the survivability that enables the FRT control to provide resilience and grid support. This two-stage response, fast hardware protection followed by intelligent software control is the cornerstone of modern fault management in hybrid MGs. The matrix in

Table 14. Matrix of fault characteristics and protection schemes for HMG converters.

Fault Type	Affected Component(s)	Inherent Fault Response/Key Challenge	Primary Protection Device	Device Technology	Typical Control Action	Reported Clearing Time (µs)
DC Pole-to-Pole (Low Impedance)	VSC-based Converters, DC Bus Capacitors	Uncontrolled capacitor discharge through anti-parallel diodes; extremely high di/dt [130]	Solid-State Circuit Breaker (SSCB)	SiC MOSFET, IGBT, IGCT	Ultra-fast Isolation	0.2 (200 ns) [138], 3.6, 10 and160 [134]
DC Pole-to-Pole (Low Impedance)	CSC-based Converters	Inherent fault current limiting due to large DC-link inductor; slower di/dt [133].	Mechanical DC Circuit Breaker (MCB) or Hybrid CB	Mechanical/Solid-State Hybrid	Isolation	>1000 (ms range)
DC Pole-to-Ground (High Impedance)	DC Bus, Grounding System	Low fault current magnitude; difficult to detect in ungrounded/high-Z grounded systems [131].	Ground Fault Detector, Differential Protection Relay	Digital Relay	Alarm/Coordinated Trip	Detection-dependent
AC Symmetrical Fault (3-Phase)	Grid Interface, AC-side of ILC	Causes severe voltage sag at PCC, potential for VSC overcurrent and DC-link overvoltage [139]	STATCOM, DVR, Fault Current Limiter (FCL) [135]	FACTS Devices, Superconductors	Fault Ride-Through (FRT): Reactive power injection to support grid voltage [135]	N/A (Support Action)
AC Asymmetrical Fault (L-G, L-L)	Grid Interface, AC-side of ILC	Creates negative sequence voltage/current, causing torque pulsations in machines and DC-link voltage ripple [139]	Crowbar Circuit, Series Dynamic Resistor (SDR) [135]	Thyristor, Resistor Grid	FRT: Negative sequence current injection to balance grid voltages [139]	N/A (Support Action)
Post-Fault Recovery	Entire Microgrid	System oscillations, voltage/frequency instability post-clearance. Recovery time is sensitive to line impedance [136]	Microgrid Central Controller (MGCC)	Multi-Agent System (MAS)	Coordinated restoration, load shedding, resynchronization [137]	Seconds to minutes
Fault Type	Affected Component(s)	Inherent Fault Response/Key Challenge	Primary Protection Device	Device Technology	Typical Control Action	Reported Clearing Time (µs)
DC Pole-to-Pole (Low Impedance)	VSC-based Converters, DC Bus Capacitors	Uncontrolled capacitor discharge through anti-parallel diodes; extremely high di/dt [130]	Solid-State Circuit Breaker (SSCB)	SiC MOSFET, IGBT, IGCT	Ultra-fast Isolation	0.2 (200 ns) [137], 3.6, 10 and160 [134]
DC Pole-to-Pole (Low Impedance)	CSC-based Converters	Inherent fault current limiting due to large DC-link inductor; slower di/dt [133].	Mechanical DC Circuit Breaker (MCB) or Hybrid CB	Mechanical/Solid-State Hybrid	Isolation	>1000 (ms range)
DC Pole-to-Ground (High Impedance)	DC Bus, Grounding System	Low fault current magnitude; difficult to detect in ungrounded/high-Z grounded systems [131].	Ground Fault Detector, Differential Protection Relay	Digital Relay	Alarm/Coordinated Trip	Detection-dependent
AC Symmetrical Fault (3-Phase)	Grid Interface, AC-side of ILC	Causes severe voltage sag at PCC, potential for VSC overcurrent and DC-link overvoltage [139]	STATCOM, DVR, Fault Current Limiter (FCL) [135]	FACTS Devices, Superconductors	Fault Ride-Through (FRT): Reactive power injection to support grid voltage [135]	N/A (Support Action)
AC Asymmetrical Fault (L-G, L-L)	Grid Interface, AC-side of ILC	Creates negative sequence voltage/current, causing torque pulsations in machines and DC-link voltage ripple [139]	Crowbar Circuit, Series Dynamic Resistor (SDR) [135]	Thyristor, Resistor Grid	FRT: Negative sequence current injection to balance grid voltages [139]	N/A (Support Action)
Post-Fault Recovery	Entire Microgrid	System oscillations, voltage/frequency instability post-clearance. Recovery time is sensitive to line impedance [136]	Microgrid Central Controller (MGCC)	Multi-Agent System (MAS)	Coordinated restoration, load shedding, resynchronization [137]	Seconds to minutes

synthesizes these concepts, linking fault types to specific protection and control responses.

6. Discussion

This comprehensive review of the state-of-the-art in HMGs reveals a fundamental shift in the role of the power electronic converter. It has definitively evolved from a passive hardware interface for power processing into a central, intelligent, and cyber-physical nexus. The findings compiled from this analysis (Section 2, Section 3, Section 4 and Section 5) consistently demonstrate that the stability, efficiency, and fault-resilience of the entire HMG are not merely dependent on individual components but are an emergent property of the deeply synergistic interplay between converter topology (hardware), control strategy (software), and the physical realities of the grid (line characteristics).

A primary insight from this analysis is the clear and accelerating divergence of converter design philosophy based on the application, as highlighted by the detailed comparison of LV and MV systems in Section 4. This is not simply a quantitative difference in voltage but a qualitative and fundamental bifurcation in engineering priorities. As Table 13 illustrates, LV systems, driven by mass-market deployment and end-user safety, prioritize high power density, cost-effectiveness, and strict adherence to interconnection standards like IEEE 1547. This economic and technical environment has fueled the adoption of high-frequency GaN devices. Conversely, MV systems for critical infrastructure (such as offshore wind, utility-scale storage) prioritize extreme reliability and operational robustness above all else, justifying the use of specialized SiC or IGCT devices that can withstand higher voltage and thermal stresses. This technological specialization signifies a maturing field where one-size-fits-all solutions are obsolete, replaced by highly optimized, application-specific designs.

The most powerful theme emerging from this review is the ascendancy of software-defined hardware, where advanced control strategies are deployed to compensate for, or entirely mask, the physical limitations of both the converters and the grid. This review identifies two critical examples in Section 4:

• Virtual Inertia: The inherent lack of mechanical inertia in converter-dominated grids is a primary source of frequency instability. The implementation of VSM or VIC in grid-forming and interlinking converters is a direct software-based solution, synthesizing the stabilizing properties of a rotating mass where none physically exists.
• Virtual Impedance: Classical droop control, while appealing for its autonomy, fundamentally fails in the resistive LV/MV lines common in HMGs, leading to power-sharing inaccuracies. AVIDC acts as a software-based fix, digitally reshaping the converter’s perceived output impedance to effectively decouple active and reactive power control and force the system to behave as the classical inductive model predicts.

This trend implies that future innovation in power electronics will be as much a function of control theory and computational science as it is of traditional hardware topology and semiconductor physics.

This software-defined paradigm extends most critically into fault management, which Section 5 identifies as a key vulnerability, particularly for the ubiquitous VSC. The analysis of fault behavior as illustrated in Table 14, establishes that the VSC’s intrinsic topology, with its large DC-link capacitor and anti-parallel diodes, makes it exceptionally vulnerable to DC-side faults, leading to uncontrolled and destructive capacitor discharge currents. This finding underscores a critical and non-negotiable technological dependency: VSC-based HMGs are only viable through the synergistic pairing of ultra-fast hardware protection and intelligent software resilience. The SSCB provides the survivability by interrupting the fault in microseconds, thereby protecting the hardware. This action, in turn, enables the implementation of advanced software control for resilience, such as the FRT strategies that shift the grid-support philosophy from protect and disconnect to survive and support. This two-stage, hardware-software response is the cornerstone of modern fault management.

Finally, while this review highlights a clear trajectory toward more intelligent and optimal control, namely MPC and RL (Table 8), it also implicitly points to significant implementation gaps that define the next research frontier.

• The Computational Barrier: The superior performance of MPC is consistently hampered by its high computational burden, which is often untenable for the microsecond-level execution required for converter switching. The true research gap is not just faster algorithms, but the co-design of control algorithms and dedicated hardware accelerators (for instance, FPGAs, Systems-on-a-Chip) to make real-time optimization viable at the speed of power electronics.
• The Economic Barrier: The proliferation of advanced hardware (MMCs, SiC devices, SSCBs) identified in Section 3 and Section 5 creates a significant economic barrier to entry. The field urgently requires the development of robust techno-economic models and lifecycle cost analyses to quantify the long-term operational benefits (for example, higher efficiency, reduced downtime, ancillary service revenue) against the immediate capital expenditure (CAPEX), thereby justifying their adoption.
• The Standardization Barrier: The very success of proprietary virtual controls (VSM, AVIDC) creates a new, pressing challenge interoperability. As different vendors implement their own virtual inertia algorithms, the risk of negative control interactions and system-wide instability in multi-vendor HMGs becomes significant. A critical future direction is the development of new industry standards (akin to IEEE 1547) to define the external behavior and communication protocols of these virtualized functions, ensuring plug-and-play stability and moving the field from bespoke projects to a standardized, scalable industry.

7. Conclusions and Future Perspectives

This review has systematically analyzed the pivotal role of power electronic converters in the design and operation of HMGs. Our analysis demonstrates that the converter has transcended its traditional function of power processing to become the intelligent, cyber-physical core of the modern grid. We have synthesized three dominant trends:

• The software-defined hardware paradigm, where virtualization techniques like VSM and AVIDC are essential for stability
• The fundamental design bifurcation between LV (cost/density-driven) and MV (reliability-driven) applications
• The non-negotiable, synergistic pairing of ultra-fast hardware protection SSCBs for survivability with advanced software FRT for resilience.

These findings show that HMG design is no longer a sequential optimization problem but a holistic co-design challenge in which converter hardware, control intelligence and the physical network must evolve together.

Building on this synthesis, three forward-looking research pathways emerge as critical enablers of scalable and reliable HMG deployment.

• Research Direction 1: The Computational Frontier

Advanced controllers such as MPC deliver superior dynamic performance but remain constrained by computational complexity. Future work should focus on hardware–software co-design, including dedicated accelerators (FPGAs, SoCs) and reduced-order models for real-time feasibility. An emerging direction is the use of AI and reinforcement learning to approximate MPC-like optimality with substantially lower online computational burden.

• Research Direction 2: The Economic Frontier

Wide-bandgap semiconductors (SiC/GaN) and SSCBs offer transformative performance but face high CAPEX barriers. Research should therefore emphasize techno-economic lifecycle models that quantify long-term value, including reduced downtime and improved resilience. A promising trend is the monetization of converter-based services, where VSM and FRT functions can participate in ancillary markets (e.g., synthetic inertia, fast frequency response), enabling new revenue streams that offset initial investment.

• Research Direction 3: The Standardization Frontier

The rapid proliferation of proprietary virtual control schemes introduces risks of negative interoperability. A practical pathway forward is the development of Digital-Twin-based validation environments that allow multi-vendor testing of virtualized controls under standardized conditions. Such platforms can accelerate the creation of next-generation interoperability standards (similar to IEEE 1547) for intelligent, software-defined converter assets. In summary, the future of HMGs will rely on coordinated advancements in computational efficiency, techno-economic viability, and robust standardization frameworks. Addressing these frontiers will be essential for translating emerging converter technologies into practical, scalable and resilient hybrid microgrid deployments.