A Coordinated Inertia Support Strategy for Wind–PV–Thermal Storage Systems Considering System Inertia Demand

Abstract

To address the challenges to power system frequency stability under high penetration of renewable energy, this paper proposes a coordinated inertia support strategy for wind–PV–thermal storage systems, overcoming the limitations of conventional inertia parameter adjustment. The core of the strategy lies in optimizing unit control activation logic and establishing a scenario-adaptive batch activation mechanism. Specifically, virtual inertia characteristic models for wind, PV, and storage units are developed, with key parameters optimized via fuzzy-logic-based coordinated control. An inertia demand assessment model under frequency security constraints is constructed to quantify the minimum system inertia requirement. Furthermore, disturbance reference power is generated based on the inertia reserve capability of each unit, and disturbance intervals are classified to achieve coordinated optimal allocation of virtual inertia. Simulation results on a built 3-machine, 9-node system demonstrate that the proposed strategy can intelligently coordinate the activation timing, role assignment, and regulation resources of wind, PV, and storage according to the type and severity of disturbances. Under various scenarios such as sudden load increase and decrease, the system effectively mobilizes resources to maintain frequency within the secure range while avoiding frequent actions of any single unit. The results verify that the strategy significantly enhances the system’s capability to handle bidirectional power disturbances and provide frequency support, offering a practical solution for inertia management in renewable-dominated power systems.

1. Introduction

The large-scale integration of wind and PV energy sources has led to a decline in system inertia levels. When power grids experience severe disturbances, frequency instability may occur, potentially causing system collapse [1,2,3]. The 2019 London blackout, triggered by a sudden drop in wind power output, resulted in significant load shedding [4]. Major blackouts such as the 2016 South Australian event and the 2021 Texas power crisis [5,6] further highlighted frequency response challenges in power systems with high renewable penetration. Conventional thermal power generation provides robust inherent inertia support, serving as a fundamental stabilizer for system frequency [7,8]. Meanwhile, battery energy storage systems (BESS) offer rapid response capabilities [9,10]. The integration of wind, photovoltaic, thermal, and storage systems into complementary multi-energy portfolios presents an effective pathway to address inertia deficiency through characteristic complementarity and coordinated control.

Substantial research has been conducted on comprehensive inertia control strategies for various generation portfolios. Reference [11] implemented precise equivalent inertia quantification and control parameter optimization for wind turbines through dynamic modeling and transfer function analysis. Reference [12] enhanced system frequency stability by combining fuzzy adaptive virtual inertia control with state-of-charge-aware ESS recovery strategies. For hybrid PV-battery systems, reference [13] employed genetic algorithm-optimized virtual synchronous generator control to improve transient stability. Reference [14] developed proportional frequency-voltage deviation relationships with segmented DC-side voltage control to effectively support system inertia in PV-ESS.

These methods predominantly rely on accurate system parameters, which are often difficult to obtain in practical engineering applications. Consequently, controllers lack reliable parameter setting basis and struggle to precisely match system frequency regulation requirements. Significant differences exist in inertia response characteristics among different energy sources: nuclear, coal and combined-cycle gas power plants provide stable but relatively slow response; wind power offers wide adjustable range and fast virtual inertia; PV systems respond fastest but with limited adjustability; ESS delivers both rapid and flexible inertia response. References [15,16] demonstrated enhanced frequency support capability through inertia reserve assessment and rational task allocation.

The stable inertia provided by coal and hydropower remains crucial for grid security, effectively mitigating wind and solar fluctuations [17]. Reference [18] quantitatively evaluated inertia contributions of doubly fed induction generators under varying wind conditions, while reference [19] analyzed PV system inertial response capability through frequency-voltage relationships. Reference [20] established energy conversion relationships between battery/capacitor storage and synchronous generator kinetic energy, deriving virtual inertia estimation methods for two storage types. Considering uncertainties like wind and solar irradiance, probabilistic inertia assessment methods based on generation characteristics have been developed [21]. Reference [22] achieved frequency stability and economic operation through coordinated virtual inertia and unit commitment optimization in regional grids. Reference [23] systematically evaluated practical implementation aspects of assessment methods, providing valuable insights for accurate renewable inertia estimation.

During frequency response, both the rate of change in frequency (RoCoF) and frequency deviation are critical security indicators [24,25]. Power source inertia determines not only the initial RoCoF after disturbance but also the frequency nadir. Inertia allocation must satisfy grid frequency security requirements [26,27]. Comprehensive evaluation of the relationship between power disturbances, frequency security, and inertia reserves is essential for implementing coordinated inertia control in integrated energy systems.

The rapid integration of inverter-based renewable energy sources, such as wind and photovoltaics, has progressively reduced the rotational inertia of power systems, posing significant challenges to frequency stability. Existing research has extensively explored virtual inertia control for individual renewable or storage units, and several studies have proposed coordinated strategies among multiple resources. However, most approaches rely on fixed control parameters or simplified activation logic, which lack adaptability to varying disturbance scenarios and fail to fully exploit the complementary potential of hybrid wind-PV-thermal-storage systems in a dynamically optimized manner. To bridge this gap, this paper proposes a scenario-adaptive coordinated inertia support strategy. The main contributions are threefold: (1) establishing virtual inertia models for wind, PV, and storage units, with a fuzzy-logic-based coordinated control to dynamically optimize key evaluation parameters; (2) constructing an inertia demand assessment model under frequency security constraints to quantify the minimum system inertia requirement; and (3) developing a disturbance-interval-based activation logic that coordinates the timing and role of different units according to disturbance severity. Simulation results verify that the proposed coordinated control strategy significantly enhances system frequency stability through rational allocation of inertia resources and adaptive control activation, preventing severe frequency instability caused by low inertia, intermittent sources and non-coordinated inverted-based resources.

2. Inertia Assessment of Integrated Wind-PV-Thermal-Storage Systems

2.1. Inertia Characteristics of Wind-PV-Thermal-Storage Units

To reflect the differences in inertial response characteristics among different units and comprehensively account for the factors influencing power output changes during frequency response, this paper establishes simulation models for the electrical energy conversion characteristics of wind turbines, photovoltaic systems, and battery energy storage systems.

2.1.1. Inertia Characteristics of Thermal Power Units

In power systems, the inherent inertia time constant H_g of a synchronous generator is determined by its rotational kinetic energy and rated capacity. The rotor motion equation of the generator can be expressed as Equation (1) [28]:

$$\left\{\begin{array}{l}{J}_g{d}^2\delta /d{t}^2={\overline{T}}_m-{\overline{T}}_e\\ H(d{\overline{\omega }}_r/dt)={\overline{T}}_m-{\overline{T}}_e\approx {\overline{P}}_m-{\overline{P}}_e\end{array}\right.$$

(1)

where $J_g$ is the total moment of inertia of the generator and prime mover rotor; $\delta$ is the rotor angular displacement relative to the synchronous rotating reference axis; ${\overline{T}}_m$, ${\overline{T}}_e$, ${\overline{P}}_m$, ${\overline{P}}_e$ are the per-unit values of mechanical torque, electromagnetic torque, mechanical power, and electromagnetic power, respectively; $H$ is the inertia time constant; ${\overline{\omega }}_r$ is the per-unit value of rotor angular velocity; considering the negligible variation in mechanical angular velocity $\omega$, the per-unit value of torque is approximately equal to the per-unit value of power.

When the inertia magnitude is known, the inertia time constant of the generating unit can be calculated using the following formula:

$$H=J {{\omega }_e}^2/(2S_N)$$

(2)

where $J$ represents the rotational inertia or virtual inertia of the unit, $S_N$ denotes the rated capacity of the unit, and ${\omega }_e$ is the angular velocity at the standard frequency.

2.1.2. Inertia Characteristics of Wind Turbines

Considering that the inertia reserve of a wind turbine is positively correlated with its power reserve, and the power reserve is positively correlated with wind speed, the calculation of wind turbine inertia under wind speed constraints can be established based on this relationship.

According to the wind power generation characteristics described in [18], the virtual inertia of a wind turbine, which depends on the rotor speed and the system nominal angular velocity, can be expressed as:

$$J_{vw}=J_w{\omega }_{r0}d{\omega }_r/({\omega }_{e}d{\omega }_e)\approx J_w\Delta {\omega }_r{\omega }_{r0}/(\Delta {\omega }_e{\omega }_e)$$

(3)

Based on the relationship between inertia and the inertia time constant, the virtual inertia time constant of a wind turbine can be expressed as:

$$H_{vw}=J_{vw}{{\omega }_e}^2/(2S_{NW})=J_w{{\omega }_e}^2\Delta {\omega }_r{\omega }_{r0}/(2\Delta {\omega }_e{\omega }_e{S}_{NW})=H_w\Delta {\omega }_r{\omega }_{r0}/(\Delta {\omega }_e{\omega }_e)$$

(4)

where $S_{NW}$ represents the rated capacity of the wind turbine. Since the rotor speed of the wind turbine depends on real-time wind conditions, the relationship between wind speed and rotor speed can be expressed as:

$${{\displaystyle \omega }}_{r0}=\left\{\begin{array}{l}{p}_n{N}_t{\lambda }_{opt}\nu /({\omega }_e{R}_t), \begin{array}{c}{\nu }_{\min }<\nu <{\nu }_{\max }\end{array}\\ p_n{N}_t{\omega }_{\max }/{\omega }_e, \begin{array}{c}\nu \ge {\nu }_{\max }\end{array}\end{array}\right.$$

(5)

By substituting Equation (11) into Equation (10), the calculation method for the virtual inertia time constant of wind turbines under different wind speeds can be derived as follows:

$$H_{vw}=\left\{\begin{array}{l}{p}_n{N}_t\Delta {\omega }_r{H}_W{\lambda }_{opt}\nu /({\omega }_e\Delta {\omega }_e{R}_t), \begin{array}{c}{\nu }_{\min }<\nu <{\nu }_{\max }\end{array}\\ p_n{N}_t\Delta {\omega }_r{H}_W{\omega }_{\max }/({\omega }_e\Delta {\omega }_e), \begin{array}{c}\nu >{\nu }_{\max }\end{array}\end{array}\right.$$

(6)

where ${\lambda }_{opt}$ represents the optimal tip-speed ratio of the wind turbine, $N_t$ denotes the gearbox speed ratio, $p_n$ is the number of pole pairs of the generator, $R_t$ signifies the blade radius, $\Delta {\omega }_r$ indicates the variation in rotor angular velocity, $\Delta {\omega }_e$ refers to the change in synchronous angular velocity, and $H_W$ corresponds to the inherent inertia time constant of the wind turbine.

As derived from Equation (5), wind turbines can emulate the inertial response characteristics of synchronous generators by regulating rotor speed variations. The aforementioned equations establish a framework for modeling wind turbine inertia control and evaluating wind power inertia reserves. However, accounting for constraints related to wind speed and rotor speed, the operational limits of wind turbine rotor speed during frequency dips can be expressed as:

$$\Delta {\omega }_{\max }=\left\{\begin{array}{l}{\omega }_{r0}-{\omega }_{\min }, \begin{array}{c}{\nu }_{\min }<\nu <{\nu }_h\end{array}\\ {\omega }_{r0}-g({\omega }_{r0},t_h), \begin{array}{c}{\nu }_h<\nu <{\nu }_{\max }\end{array}\end{array}\right.$$

(7)

where $g({\omega }_{r0},t_h)=\sqrt{t_h(k_{opt}{{\omega }_{r0}}^3-P_n)/H_w+{{\omega }_{r0}}^2}$, ${\nu }_h=0.989{\omega }_e{R}_t/({\lambda }_{opt}{p}_n{N}_t)$, Combining Equations (7) and (2), the maximum inertial response time constant of a wind turbine can be expressed as:

$$H_{vw\text{\_}\max }=\left\{\begin{array}{l}({\omega }_{r0}-{\omega }_{\min }){\omega }_{r0}{H}_W/(0.02{\omega }_e), \begin{array}{c}{\nu }_{\min }<\nu <{\nu }_h\end{array}\\ [{\omega }_{r0}-g({\omega }_{r0},t_h)]{\omega }_{r0}{H}_W/(0.02{\omega }_e), \begin{array}{c}{\nu }_h<\nu <{\nu }_{\max }\end{array}\end{array}\right.$$

(8)

By integrating the methodology from reference [15] with Equation (8), the parameters of the wind turbine inertia controller can be derived as follows:

$$K_{vw}=\left\{\begin{array}{l}2H_w\left(\ln \left|({\omega }_{ref}-{\omega }_{r0})/({\omega }_{ref}-{\omega }_{r1})\right|+\Delta {\omega }_r\right)/t_h, \begin{array}{c}\Delta P_d>0\end{array}\\ 2H_w\left(\ln \left|({\omega }_{ref}-{\omega }_{r0})/({\omega }_{ref}-{\omega }_{r1})\right|-\Delta {\omega }_r\right)/t_h, \begin{array}{c}\Delta P_d<0\end{array}\end{array}\right.$$

(9)

where $H_W$ denotes the inherent inertia time constant of the wind turbine, taken as 2 s; ${\omega }_{ref}$ represents the rated rotational speed, with a value ranging from 1.0 to 1.2 p.u.; ${\omega }_{r0}$ is the currently measured rotational speed; ${\omega }_{r1}$ signifies the rotational speed margin, set between 0.1 and 0.2 p.u.; and $\Delta {\omega }_r$ indicates the variation in rotational speed. Based on the above equations, operational constraints for the wind turbine can be established, and the parameters of the inertia controller satisfying these constraints can be determined.

2.1.3. Inertia Characteristics of Photovoltaic Systems

To participate in inertia assessment, photovoltaic systems must maintain an active power reserve, typically 10% of their rated capacity [29]. By regulating the DC-link voltage, PV systems can emulate the inertial response characteristics of synchronous generators, as expressed by:

The inertia reserve of a photovoltaic system is directly determined by the DC-link voltage level [19]. The virtual inertia provided by the PV system can be expressed as:

$$J_{pv}=\left\{\begin{array}{l}0.5C{U}_{dc}^2-0.5C{U}_{mpp}^2, \begin{array}{c}\Delta f<0\end{array}\\ 0.5C{U}_{\max }^2-0.5C{U}_{dc}^2, \begin{array}{c}\Delta f>0\end{array}\end{array}\right.$$

(10)

The inertia time constant of a photovoltaic system can be expressed as:

$$H_{pv}=J_{pv}{{\omega }_e}^2/(2S_{Npv})=\left\{\begin{array}{l}(0.5C{U}_{dc}^2-0.5C{U}_{mpp}^2){{\omega }_e}^2/(2S_{N\mathrm{pv}}), \begin{array}{c}\Delta f<0\end{array}\\ (0.5C{U}_{\max }^2-0.5C{U}_{dc}^2){{\omega }_e}^2/(2S_{Npv}), \begin{array}{c}\Delta f>0\end{array}\end{array}\right.$$

(11)

where $C$ represents the DC capacitance value, $U_{dc}$ denotes the DC-side voltage, $U_{mpp}$ is the maximum power point tracking (MPPT) control voltage, and $S_{Npv}$ signifies the rated capacity of the photovoltaic system.

Solar irradiance directly affects the output power of photovoltaic systems, consequently influencing their available inertia reserve. The relationship between PV power output and irradiance can be expressed as:

$$P_{mpp}=\eta SE$$

(12)

where $P_{mpp}$ represents the power output under maximum power point tracking (MPPT) control, $\eta$ denotes the photoelectric conversion efficiency, $S$ signifies the effective illumination area, and $E$ indicates the solar irradiance intensity. In the photovoltaic power generation model of this study, to focus on the effectiveness analysis of the control strategy, the irradiance condition is simplified and set to a constant 1000 W/m².

The relationship between the DC-side capacitor voltage and solar irradiance intensity in photovoltaic systems can be expressed as:

$$C{U}_{dc}d{U}_{dc}/dt=P_{mpp}$$

(13)

Referring to [28], the parameter calculation of the PV voltage droop controller is generally closely related to the DC-side voltage. The droop coefficient can be expressed as:

$$K_{pv}=K_{pv\_ref}\cdot \left(1+K_U(U_{dc}-U_{dc\_ref})/U_{dc\_ref}\right)$$

(14)

where $K_{pv\_ref}$ represents the rated droop coefficient, determined based on the frequency regulation capacity; $K_U$ denotes the voltage sensitivity coefficient, typically ranging from 0.5 to 2.0; and $U_{dc\_ref}$ indicates the rated DC-side voltage, set at 700 V.

With reference [16,30], the operational fluctuation range of the PV DC-link voltage is set at ±10%.

2.1.4. Inertia Characteristics of Battery Energy Storage Systems

The inertia reserve of an energy storage system is positively correlated with the stored energy of the battery. Based on this relationship, the calculation conditions for energy storage inertia under capacity constraints can be established. The relationship between the battery reserve energy $W_b$ and its virtual inertia can be expressed as:

$$W_b=\int Q_{N}d(1-{\gamma }_{soc})J_{vb}{\omega }_{e}d{\omega }_e/({\omega }_{e}d{\omega }_e)=\int J_{vb}{\omega }_{e}d{\omega }_e$$

(15)

where $Q_N$ represents the rated capacity of an individual battery unit. Based on the battery discharge characteristics, the virtual inertia of the energy storage system can be expressed as:

$$J_{vb}=J_b{{\omega }_e}^2/(2S_{NB})=-J_g{\omega }_e\Delta {\gamma }_{\mathrm{soc}}{W}_b/(2{\gamma }_{\mathrm{soc}}\Delta {\omega }_e{E}_g)$$

(16)

where $S_{NB}$ denotes the aggregated rated apparent power capacity of the entire Battery Energy Storage System. Substituting Equation (16) into Equation (2), the virtual inertia time constant of the battery energy storage system can be expressed as:

$$H_{vb}=J_b{{\omega }_e}^2/(2S_{NB})=-H_g{\omega }_e\Delta {\gamma }_{\mathrm{soc}}{W}_b/(2{\gamma }_{\mathrm{soc}}\Delta {\omega }_e{E}_g)$$

(17)

where ${\gamma }_{\mathrm{soc}}$ represents the state of charge, and $E_g$ denotes the rotational kinetic energy of an equivalent-capacity synchronous generator reserve. As derived from Equation (17), battery energy storage systems can emulate the response characteristics of synchronous generators by regulating the variation in their state of charge. The above equations establish an inertial control model for battery energy storage systems and enable the evaluation of inertia reserves while accounting for the charge–discharge characteristics of batteries. Based on references [15,20], the relationship between the virtual inertia time constant of the energy storage system and the parameters of the inertia controller can be expressed as:

$$K_{vb}=2H_{vb\max }=\rho n{t}_{h0}/0.0198$$

(18)

where $\rho$ represents the wind and solar penetration rate, $n$ denotes the ratio of energy storage capacity to renewable energy capacity, set at 10%, and $t_{h0}$ indicates the inertial response time when the system relies solely on synchronous generators for inertia. Using Equation (18), the parameters of the energy storage inertia controller can be determined.

By integrating Equations (9), (14) and (18), this paper establishes a controller parameter tuning module that comprehensively considers the operational constraints of wind power, photovoltaic, and battery energy storage units.

2.2. Frequency Response Model of Wind-PV-Thermal Power and Battery Storage

Following a disturbance, the stability of system frequency is influenced by factors including the inertia of traditional synchronous generators, the virtual inertia of renewable energy sources, and load characteristics. The system’s inertia level significantly affects both the RoCoF and the frequency deviation during the frequency response process. By adjusting the power output of individual units, the overall system inertia level can be modified, ultimately achieving inertia support and frequency response. Consequently, the frequency response model of the system integrating wind, PV, thermal power and batteries is illustrated in Figure 1.

Based on the frequency response model, the system inertia time constant quantifying the inertial response characteristics of each unit can be expressed as:

$$H_s=H_g+H_{vw}{\eta }_{vw}+H_{pv}{\eta }_{pv}+H_{vb}{\eta }_s$$

(19)

where $H_{vw}$, $H_{pv}$, and $H_{vb}$ represent the inertia time constants of wind power, photovoltaic, and battery energy storage systems, respectively, in seconds; ${\eta }_{vw}$ is the wind power penetration rate; ${\eta }_{pv}$ is the photovoltaic penetration rate; ${\eta }_s$ is the renewable energy penetration rate, which equals the sum of ${\eta }_{vw}$ and ${\eta }_{pv}$; and $P_{base}$ is the total available power from all units.

The proposed inertial response model remains valid under most operating conditions, as a unit’s inertial characteristics are dictated by its control system, while the inertial capacity is determined by its available power reserve. For example, battery virtual inertia is enabled by the grid-tied converter, not by its charging or discharging state. Thus, it can provide support during charging—such as absorbing excess power during over-frequency events—as well as under power fluctuations and before full discharge.

3. System Frequency Dynamic Characteristics

3.1. System Inertia Requirement Under Frequency Security Constraints

Following a disturbance, the frequency response process is illustrated in Figure 2.

According to the system’s frequency equation:

$$2H_s(df/dt)=-\Delta P_d$$

(20)

where $H_s$ is the system inertia time constant (s). System inertia determines RoCoF. From the perspective of RoCoF security, the relationship between the minimum system inertia requirement and the RoCoF can be expressed as:

$$\min {(H_s)}_2=\Delta P_d/(2{\dot{f}}_{\max })$$

(21)

where ${\dot{f}}_{\max }$ is the maximum rate of frequency change (RoCoF), as determined by security regulations [31,32]. In this paper, it is set to 0.5 Hz/s. Within the secure operational range of the system, during the dynamic process following the initial disturbance, the system frequency drops to its nadir. The frequency deviation at the end of the inertia response phase, transitioning to the frequency nadir, is taken as the maximum frequency deviation. At this point, the minimum system inertia requirement can be expressed as:

$$\left(\begin{array}{l}\min {(H_s)}_1=F(\Delta f_{\max },\Delta P_d)\\ \Delta P_d=P_{mea}-P_{Load}\end{array}\right)$$

(22)

where $\Delta P_d$ is the disturbance power, taken as 0.1–0.16 p.u.; $\Delta f_{\max }$ is the maximum frequency deviation, taken as 0.9–1 Hz; $P_{mea}$ is the measured output power of the units; and $P_{Load}$ is the current load level.

By integrating the relationships among inertia, disturbance power, and frequency response values from Equations (21) and (22), the minimum system inertia requirement is determined by taking the larger value from the two evaluation results. The relational expression satisfying both constraint conditions is derived as follows:

$$\begin{array}{l}{H}_{s\min }=\max (\min {(H_s)}_1,\min {(H_s)}_2)\\ \begin{array}{c}\begin{array}{c} =H_{\max }(\Delta P_d,{\dot{f}}_{\max },\Delta f_{\max })\end{array}\end{array}\end{array}$$

(23)

3.2. Design of the Fuzzy-Logic-Based Coordinated Control

Due to their distinct response speeds, support durations, and energy-supply characteristics, wind power, photovoltaic systems, and energy storage exhibit fundamentally different dynamic behaviors during frequency disturbances. Such heterogeneity results in complex composite frequency responses, which undermine the accuracy of conventional inertia-assessment methods based on synchronous-generator models. To address this limitation, this study proposes a fuzzy logic–based feedback coordination mechanism. The mechanism takes system frequency dynamics—specifically frequency deviation (Δf) and rate of change in frequency (df/dt)—as inputs and generates adjustment commands for the virtual inertia evaluation outcomes of individual units. The proposed approach does not directly compute power correction terms or system inertia, nor does it alter the inherent physical inertia parameters of the generating units. Instead, it functions as a system-level coordinator that dynamically optimizes and aligns the inertia evaluation results from distributed resources, including wind turbines, photovoltaic inverters, and energy storage systems. Through this coordination framework, rapid and adaptive matching between system-wide inertia requirements and unit-level evaluation results is achieved, thereby effectively compensating for overall inertia evaluation discrepancies.

The proposed coordination mechanism employs fuzzy logic for three principal reasons. First, the controller inputs—frequency deviation (Δf) and rate of change in frequency (df/dt)—directly correspond to the key physical variables used in inertia assessment, thereby avoiding the need for intermediate signal transformation or high-fidelity system modeling. Second, in systems with high renewable penetration, the inertial responses of wind, photovoltaic, and storage resources are inherently heterogeneous and exhibit strong nonlinearities. Fuzzy logic provides a structured framework to encode operational heuristics and system knowledge into executable rules, enhancing robustness against model inaccuracies and parametric uncertainties. Third, fuzzy-based adjustment represents an established approach in inertia and damping control applications [33,34,35], offering transparent interpretability and validated adaptability through its rule-based inference mechanism. Consequently, fuzzy logic serves as the core of the coordination strategy, effectively managing the inherent complexity of multi-source inertia support while maintaining favorable implementation feasibility and operational reliability.

A close mathematical relationship exists among frequency deviation, RoCoF, and the inertia constant. Based on Figure 2, the corresponding frequency-response characteristics and inertia-adjustment rules are summarized in

Table 1. Frequency Response Interval Characteristics and System Inertia Adjustment Principles.

Scenario	Oscillation Interval	df/dt	Δf	Hs
Load Sudden Increase	[tstart, th]	<0	<0	Increase
	[th, tend]	>0	<0	Decrease
Load Sudden Decrease	[tstart, th]	>0	>0	Increase
	[th, tend]	<0	>0	Decrease

.

Based on a fuzzy-logic framework, the system is designed with a dual-input, single-output structure: real-time frequency deviation (Δf) and rate of change in frequency (df/dt) serve as the input features, while the inertia adjustment ΔH_s acts as the core output. This architecture enables precise calibration of the initial inertia estimates for each unit [33,34]. The corresponding implementation is illustrated in Figure 3.

The coordinated control based on fuzzy logic operates through the following steps. First, the frequency deviation and the RoCoF are fuzzified using quantization factors k_f and k_df, respectively. Then, fuzzy rules are applied alongside a scaling factor k_H for defuzzification, yielding the inertia variation value. Finally, this variation value undergoes secondary allocation, and the inertia change amount for each unit is added to its initial value, thus obtaining the real-time inertia value during the transient period.

The fuzzy-logic-based coordinated control adaptively regulates system inertia by tracking real-time frequency variations. The total inertia output is collectively provided by wind, solar, thermal, and storage units, with the synthesis calculation method given by Equation (19). The frequency stability controller of thermal power units can automatically perform frequency regulation following changes in output from wind, solar, and storage systems, without requiring separate parameter tuning. The allocation rules for compensating parameters of the renewable and storage inertia controllers in the fuzzy-logic-based coordinated control are as follows:

$$\Delta H_s=\Delta H_{vw}/{\eta }_{vw}=\Delta H_{pv}/{\eta }_{pv}=\Delta H_{vb}/{\eta }_s$$

(24)

The instantaneous penetration rates of wind and PV power are, in reality, time-varying functions of meteorological conditions. However, for the purpose of this study, they are modeled as constant parameters. This simplification is justified as the analysis focuses on the inertial response immediately following a disturbance—a transient period typically lasting 10–30 s, during which natural resource fluctuations are negligible. The constant penetration rates, determined by the installed capacity of each resource as defined in Equation (19), provide a stable basis for evaluating the system’s inertial characteristics and the performance of the control strategies.

The fuzzy rules are defined as follows:

The input and output variables characterize the fluctuation range of system parameters [35]. The physical domain of the system frequency deviation is (−0.3 to 0.3 Hz), and that of the system rate of change in frequency (RoCoF) is (−0.4 to 0.4 Hz/s). Based on the calculated range of required system inertia variation, the physical domain of the output variable is set as (0 to 4 s). The fuzzy subsets for the input Δf are {NB (negative big), MB (negative medium), B (negative small), M (zero), S (positive small), MS (positive medium), NS (positive big)}. The fuzzy subsets for the input df/dt are {NB (negative big), NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), PB (positive big)}. The fuzzy subsets for the output H_s are {VB (very big), MB (medium big), B (big), M (medium), S (small), MS (medium small), vs. (very small)}. The corresponding fuzzy inference rules are shown in

Table 2. Fuzzy rules for system inertia ΔH_s.

ΔHs		Gf
		NB	NM	NS	ZE	PS	PM	PB
Gdf	NB	PB	PM	PM	PB	NM	NM	NB
	NM	PB	PM	PS	PM	NS	NM	NB
	NS	NS	PM	PS	ZE	NS	NM	NM
	ZE	NB	NM	ZE	ZE	ZE	NM	NB
	PS	NM	NM	NS	ZE	PS	PM	PM
	PM	NB	NM	NM	PB	PM	PM	PB
	PB	NB	NM	NM	PB	PM	PM	PB

, and the membership functions of the input and output variables are presented in Figure 4, respectively. A common fuzzy domain of [−3, 3] is adopted, with the quantization factors k_f and k_df set to 10 and 7.5, respectively, and the scaling factor k_H set to 0.667.

The revised inertia assessment enables a more accurate comparison to determine whether the current system inertia reserve meets the minimum inertia requirement, thereby facilitating the formulation of appropriate coordinated control strategies.

4. Coordinated Inertia Support Strategy for Wind-Solar-Storage Systems Under Security Constraints

4.1. Inertia Allocation Method

4.1.1. Disturbance Reference Interval Partitioning

When a disturbance occurs, inertia support control should adapt to the interval range of the power disturbance value, meet the system inertia requirements under different operating conditions, and ensure the normal operation of the units.

With the goal of satisfying the minimum system inertia requirement and constrained by the maximum frequency response value, the output power of each unit can be adjusted by modifying the activation status of inertia control, thereby maximizing the utilization of inertia resources from all units. Based on Equation (2) and the fundamental principles of system frequency response in inertia assessment, there exists a close relationship among disturbance power, inertia, and frequency response values. This allows for the definition of action intervals for frequency regulation commands according to the range of the disturbance power.

This paper calculates the reference value of balanced disturbance power that meets the system inertia requirement and matches the system inertia reserve capacity. This reference value takes into account the mechanism of system frequency response, the inertia resources reserved in the system model under initial conditions, the maximum disturbance tolerance of the system, and the penetration rates of each unit. Based on this reference value, three disturbance intervals are defined. By constraining the action periods of frequency regulation control for each unit according to these disturbance reference intervals, a zonal and segmented control strategy is implemented.

Based on the range of disturbance variation, the following three scenarios can be derived:

• Small disturbance: Inertia control of renewable energy units is not required. When the synchronous generator inertia equals the minimum system inertia requirement, the disturbance magnitude is ΔP_d1. That is, when ΔP_d ≤ ΔP_d1, H_smin ≤ H_g.
• Moderate disturbance: Wind turbines and photovoltaic systems are required to provide inertia support. When the sum of the maximum inertia reserves from wind, photovoltaic, and thermal power equals the minimum system inertia requirement, the disturbance magnitude is ΔP_d2. Combining with Equation (6), when ΔP_d1 < ΔP_d ≤ ΔP_d2, H_g < H_smin ≤ F(H_{vw_max}, H_{pv_max}, H_g).
• Large disturbance: Wind, photovoltaic, thermal, and storage systems collectively provide inertia support. Energy storage inertia control is activated, and the system’s reserved inertia meets the minimum system inertia requirement. That is, when ΔP_d2 < ΔP_d, H_smin ≤ H_s.

To satisfy both frequency constraints, the disturbance interval thresholds are parameterized from the perspectives of frequency deviation and rate of change in frequency. Supportive control actions are triggered when the disturbance exceeds any of these thresholds. The calculation of the reference disturbance power values ΔP_d1 and ΔP_d2 under frequency constraints is given by:

$$\left\{\begin{array}{c}\Delta P_{d11}=F\left[f , H_{\mathrm{g}}(1-{\eta }_s)\right]\\ \Delta P_{d21}=2{\dot{f}}_{\max }{H}_{\mathrm{g}}(1-{\eta }_s)\\ \Delta P_{d12}=F\left[f , H_{\mathrm{vw}\text{\_}\max }{\eta }_{vw}+H_{\mathrm{pv}\text{\_}\max }{\eta }_{pv}+H_{\mathrm{g}}(1-{\eta }_s)\right]\\ \Delta P_{d22}=2{\dot{f}}_{\max }\left[H_{\mathrm{vw}\text{\_}\max }{\eta }_{vw}+H_{\mathrm{pv}\text{\_}\max }{\eta }_{pv}+H_{\mathrm{g}}(1-{\eta }_s)\right]\\ \Delta P_{d1}=\min (\Delta P_{d11},\Delta P_{d21})\\ \Delta P_{d2}=\min (\Delta P_{d12},\Delta P_{d22})\end{array}\right.$$

(25)

where $\Delta P_{d11}$ and $\Delta P_{d12}$ are the calculated reference values under the frequency deviation constraint, while $\Delta P_{d21}$ and $\Delta P_{d22}$ are the calculated reference values under the RoCoF constraint.

Based on the allocation rules, this paper divides the disturbance power into three intervals: [0, ΔP_d1], [ΔP_d1, ΔP_d2], and [ΔP_d2, ΔP_de), where $\Delta P_{de}$ represents the maximum disturbance power that the system can withstand under secure grid operation.

4.1.2. Inertia Allocation Scheme

Considering the time required for online calculation of wind-solar-storage power requirements and the inherent response lag of frequency support control in practical scenarios, the control strategy should adopt the maximum value of the disturbance interval as a conservative assessment. The inertia allocation scheme is shown in

Table 3. Wind-PV-Storage inertia allocation scheme.

Disturbance Scenarios	ΔPd/p.u.	Hg/s	Hvw_set/s	Hpv_set/s	Hvb_set/s
Scenario 1	[0, ΔPd1]	Hg	0	0	0
Scenario 2	[ΔPd1, ΔPd2]	Hg	Hvw	Hpv	0
Scenario 3	[ΔPd2, ΔPde)	Hg	Hvw_max	Hpv_max	Hvb

, where the inertia response provided by wind, solar, and storage systems corresponds to the maximum inertia time constant of each energy unit.

Based on the information in Table 3, under different frequency regulation conditions, control commands can be issued to selectively activate the frequency regulation control of wind, solar, and storage units. This optimizes the control logic and achieves full utilization of their inertia capabilities. For example:

• In Scenario 1, the system generates frequency regulation command 1, where wind, solar, and storage units only activate frequency regulation control without providing inertia support.
• In Scenario 2, the system generates frequency regulation command 2, where wind turbine inertia control is activated, while energy storage inertia control remains inactive.
• In Scenario 3, the system generates frequency regulation command 3, where both wind and storage inertia controls are activated.

4.2. Coordinated Control Strategy

Wind turbines adopt inertia control based on rotor speed tracking, photovoltaic systems utilize droop control with reserved power margins, and battery energy storage systems implement inertia control based on frequency tracking. The proposed strategy in this study consists of an inertia assessment module, a frequency regulation command generation module, a parameter tuning module for wind-photovoltaic-storage controllers, and a frequency support control module for wind-photovoltaic-storage systems, as shown in Figure 5.

In the frequency regulation command generation module, disturbance reference intervals are established based on predicted system parameters and operational security requirements. The overall disturbance power variation on the load side is collected as the real-time disturbance magnitude. After evaluating the system’s disturbance tolerance limit and determining the interval in which the current disturbance magnitude falls, frequency regulation commands are generated to activate the corresponding wind, photovoltaic, and storage frequency regulation controllers.

In the parameter tuning module for wind-photovoltaic-storage controllers, system parameters are sampled in real time to calculate the minimum system inertia requirement and the maximum available inertia from wind, photovoltaic, and storage resources. The evaluation results are optimized using fuzzy-logic-based coordinated control, and the controller parameters for wind, photovoltaic, and storage systems are derived under constraints including wind speed, solar irradiance, and battery energy levels. The output of these controller parameters is then regulated according to the predefined disturbance intervals.

The coordinated wind-solar-storage control structure proposed in this paper is shown in Figure 6:

When determining the frequency regulation command, the reference disturbance power under frequency variation constraints is obtained from Equation (25), and the disturbance intervals are divided into [0, ΔP_d1], [ΔP_d1, ΔP_d2], and [ΔP_d2, ΔP_de). The corresponding control actions for each interval are as follows: no inertia control activation for wind-solar-storage systems, activation of wind-solar control without energy storage inertia control, and full activation of wind-solar-storage control. The decision module determines the disturbance interval and outputs the corresponding command.

Next, the generated frequency regulation command is sent to the parameter tuning module for wind-solar-storage controllers. In this module, constraints related to wind speed and solar irradiance are established. Based on conversion relationships, wind speed constraints are equivalent to wind turbine rotor speed constraints, and solar irradiance constraints are equivalent to voltage constraints. Under satisfied constraint conditions, the controller coefficients for wind turbines, photovoltaic systems, and energy storage are generated in combination with system parameter values.

Finally, the tuned controller parameters for wind, solar, and storage systems are transmitted to the frequency support control module. Combining the virtual inertia from wind-solar-storage systems, the system inertia time constant is calculated based on initial simulation results and delivered to the action command module. To accurately respond to frequency variation scenarios, a decision module is incorporated within the support control module to compare the frequency regulation inertia with the system inertia requirement, ensuring timely deactivation of wind and storage inertia control during the frequency recovery phase. Additionally, decision modules are added for the rotor speed and voltage constraints in the controller parameter tuning process. These modules compare the sampled variations with the preset operational variation ranges to determine whether the wind and solar units are operating normally during control actions. Specifically, when ω_{r_set} − Δω₀ < 0, ΔP_w is output; when ω_{r_set} − Δω₀ > 0, no correction value is output, and normal operation is maintained. Similarly, when U_{pv_set} − ΔU₀ < 0, ΔP_pv is output; when U_{pv_set} − ΔU₀ > 0, no correction value is output, and normal operation is maintained.

5. Simulation Analysis

To validate the effectiveness of the proposed scheme, this study modifies the classic 3-machine 9-bus test system. One of the conventional generators is replaced with an integrated wind-solar-thermal-storage power generation base. The wind power section consists of 250 aggregated 2 MW doubly fed induction generators (DFIGs), with a total installed capacity of 500 MW. The photovoltaic (PV) section comprises 300 aggregated 0.5 MW PV arrays, with a total installed capacity of 150 MW. The energy storage system (ESS) adopts a battery energy storage system (BESS) configuration, with a total capacity of 80 MW. Synchronous generator G3 is modeled as a steam turbine unit with an installed capacity of 200 MW.

Additionally, synchronous generators G1 and G2 are also modeled as steam turbine units, each with an installed capacity of 200 MW. In the simulation experiments, G1 and G2 are responsible for absorbing the residual power fluctuations. The system supplies three load units with a total demand of 600 MW, and each load unit is set to 200 MW. The grid’s rated frequency is 50 Hz. The renewable energy penetration rate, defined as the ratio of the total rated capacity of wind and PV units to the total installed system capacity, is calculated to be 52%. The simulation model established for the system is shown in Figure 7.

In the model, wind turbines, photovoltaic systems, energy storage, and thermal power unit G1 collectively constitute the generation sector. Based on existing frequency security regulations [30,31,32], this study sets ${\dot{f}}_{\max }$ = 0.7 Hz/s and $\Delta f_{\max }$ = 1 Hz. A load variation event is simulated at t = 3 s, where all three load units change equally, resulting in a sudden change in total system load.

Taking the 9-bus grid model constructed in this study as an example, the wind and solar power generation at the initial simulation moment meet the load demand, with specific parameter settings listed in

Table 4. Simulation model parameter settings.

Component Type	Total Fixed Capacity or Load Level/MW	Configuration Situation
Total load: Pload	600	3 × 200
Thermal power unit: Pst	600	3 × 200
Wind turbine unit: Pvw	500	250 × 2
Photovoltaic unit: Ppv	150	300 × 0.5
Energy storage unit: Pvb	70	1 × 70

. Within the disturbance range tolerable by the system, the thermal power unit provides reliable inertia support. By simulating a load step event with a variation ratio of 10%, the system frequency dynamic response curve supported solely by synchronous generators is obtained, as shown in Figure 8.

Limited by their generation characteristics, although thermal power units possess strong mechanical inertia, their power output changes slowly during inertia response. In contrast, renewable energy units employing inertia control not only possess certain inertia reserves but can also rapidly adjust power output. Therefore, compared to directly integrating thermal power and distributed generation, the proposed support control strategy significantly improves frequency oscillations.

Based on the experimental results, the system parameters when only synchronous generators provide inertia can be analyzed and are presented in

Table 5. Key parameters in the frequency response model.

Simulation Parameters	Measured Value (a/b)
Inertial response time: th/s	3.562/7.044
RoCoF: $\dot{f}$/Hz/s	−0.655/0.558
Frequency variation: Δf/Hz	−0.377/0.425

. The parameter values obtained from the simulation will serve as input conditions for the calculation and tuning of the wind-solar-storage controller parameters. The rate of change in frequency is taken at the moment when synchronous generator inertia is activated, and the frequency deviation is taken as its maximum value. Scenarios of sudden frequency decrease and increase are categorized as Case a and Case b, respectively.

5.1. Experimental Design

To verify the superiority of the proposed strategy, simulation experiments will compare the following two control methods:

• Traditional Control: Wind and storage systems adopt conventional inertia control with fixed controller parameters, while photovoltaic systems utilize traditional power tracking control.
• Proposed Coordinated Strategy: Fuzzy-logic-based coordinated control is applied to adjust the parameters of wind-storage inertia controllers and photovoltaic droop controllers. Based on the predefined disturbance intervals, a coordinated allocation strategy is implemented to achieve stepwise activation of wind, photovoltaic, and storage controllers.

Taking ΔP_d > 0 as an example, according to the simulation model parameters and Equation (5), the initial rotor speed of the wind turbine is ω_r0 = 0.84 p.u. Based on Equations (9) and (17), the maximum virtual inertia values of the wind turbine and the energy storage system, H_{vw_max} and H_{vb_max}, are calculated as 6.3 s and 4.2 s, respectively. Using Equation (25), the reference interval values of the disturbance power, ΔP_d1 and ΔP_d2, are determined to be 0.09 p.u. and 0.12 p.u., respectively.

By varying the disturbance power level, the following three scenarios are designed:

• Scenario 1 (Synchronous Generators Solely Providing Support): When the step change in disturbance load is 0.07 p.u., the minimum system inertia requirement H_smin is determined to be 3.5 s based on the frequency constraint. The inertia provided by the synchronous generators is 7.3 s. Since 0 < ΔP_d < ΔP_d1, the system generates frequency regulation command 1. Renewable sources (wind and PV) and energy storage only participate in power regulation and do not provide inertia support.
• Scenario 2 (Renewables Participating in Inertia Support): When the step change in disturbance load is 0.10 p.u., the minimum system inertia requirement H_smin is 6.3 s. Since ΔP_d1 < ΔP_d < ΔP_d2, the system generates frequency regulation command 2. The parameter setting module outputs the parameters for the wind turbine inertia controller and the PV power regulation coefficient, while the parameters for the energy storage inertia controller are set to zero. Wind turbines and synchronous generators jointly provide inertia support, and wind, PV, thermal, and energy storage work coordinatively to accomplish frequency regulation.
• Scenario 3 (Coordinated Inertia Support from Wind, PV, Thermal, and Storage): When the step change in disturbance load is 0.15 p.u., the minimum system inertia requirement H_smin is 7.2 s. Since ΔP_d > ΔP_d2, the system generates frequency regulation command 3. The parameter setting module outputs the parameters for the wind and energy storage inertia controllers and the PV power regulation coefficient. Wind turbines, energy storage, and synchronous generators jointly provide inertia support, achieving coordinated frequency support from wind, PV, thermal, and energy storage.
• According to Figure 8, in Scenario 1 under the proposed strategy, a sudden load increase leads to a frequency drop, with a maximum RoCoF of −0.558 Hz/s and a maximum frequency deviation of −0.425 Hz. Conversely, a sudden load decrease results in a frequency rise, with a maximum RoCoF of 0.655 Hz/s and a maximum frequency deviation of 0.377 Hz. These values satisfy the frequency security constraints, with wind and storage providing no inertia support. The controller parameters for the three scenarios are listed in

  Table 6. Controller parameters for scenarios 1–3.

  Simulation Parameters	Scenario 1	Scenario 2	Scenario 3
  Fan controller parameters: Kvw	0	3.55/2.45	4.41/3.63
  Photovoltaic controller parameters: Kpv	0.7/0	2.31/1.56	2.75/1.92
  Energy storage controller parameters: Kvb	0	0	4.74/4.88

  .

5.2. Simulation Experiment for Sudden Load Increase

The corresponding simulation results under this scenario are presented in Figure 9.

• In Scenario 2, under the proposed strategy, the inertial response time is 2.084 s, the maximum RoCoF is −0.307 Hz/s, and the frequency nadir is 0.215 Hz, all of which satisfy the frequency constraints. During the initial stage of the disturbance, the wind turbine’s inertial control is activated, increasing its output power by 0.102 p.u. before recovering to 1.01 p.u. The photovoltaic system increases its output by 0.1 p.u. when reserve capacity is available, while the energy storage system remains inactive.
• In Scenario 3, under the proposed strategy, the inertial response time is 1.86 s, the maximum RoCoF is −0.34 Hz/s, and the frequency nadir is 0.2 Hz, also meeting the frequency constraints. During the initial stage of the disturbance, the wind turbine’s inertial control is activated, increasing its output power by 0.147 p.u. before recovering to 1.01 p.u. The photovoltaic system increases its output by 0.151 p.u. when reserve capacity is available, while the energy storage system’s inertial control is activated, briefly delivering 0.51 p.u. of power and further increasing to 0.72 p.u. according to frequency regulation requirements.

Compared to traditional control methods, the proposed strategy demonstrates a faster inertial response from wind turbines, releasing more power in the short term to provide inertia support. The photovoltaic system flexibly utilizes its reserve capacity to increase output based on power demand through droop control. The coordinated action of wind and storage ensures effective inertia support, enabling renewables to actively participate in frequency regulation and reducing frequent actions by energy storage. As a result, the system frequency nadir is significantly improved, the maximum RoCoF is reduced, and the system transitions more quickly to the frequency recovery stage.

5.3. Simulation Experiment for Sudden Load Decrease

The corresponding simulation results under this scenario are presented in Figure 10.

• In Scenario 2, under the proposed strategy, the maximum RoCoF is 0.312 Hz/s, the frequency peak is 50.383 Hz, and the inertial response time is 6.76 s, all of which satisfy the frequency constraints. During the initial stage of the disturbance, the wind turbine’s inertial control is activated, reducing its output power by 0.153 p.u. before recovering to 0.863 p.u. The photovoltaic system reduces its output by 0.057 p.u. under droop control, while the energy storage system remains inactive.
• In Scenario 3, under the proposed strategy, the maximum RoCoF is 0.411 Hz/s, the frequency peak is 50.47 Hz, and the inertial response time is 7.77 s, also meeting the frequency constraints. During the initial stage of the disturbance, the wind turbine’s inertial control is activated, reducing its output power by 0.093 p.u. and further decreasing to 0.894 p.u. The photovoltaic system reduces its output by 0.107 p.u. under droop control, while the energy storage system’s inertial control is activated, absorbing 0.6 p.u. of power.

Compared to traditional control methods, the proposed strategy enables wind turbines to provide inertia support earlier and reduce power output promptly under deloading control, offering the main support and frequency regulation capacity. Photovoltaic systems cooperate with wind and storage to participate in primary frequency regulation during sudden frequency increases, with the coordinated control providing inertia support to prevent excessive RoCoF from triggering protection actions and generator tripping. Energy storage absorbs excess power to maintain system power balance, while wind turbines assume the primary role to avoid frequent actions by energy storage. As a result, the system frequency peak is significantly reduced, security and stability are improved, and the system becomes capable of handling more severe load variation scenarios.

The superior performance of the coordinated control strategy for the wind-PV-thermal-ESS is evidenced by the simulation results, with a comparative summary provided in

Table 7. Results and Analysis of Simulation Experiments.

Aspect	Sudden Load Increase	Sudden Load Decrease
Key Metrics	Scenario 2: Max RoCoF: −0.307 Hz/s Frequency Nadir: 49.785 Hz Scenario 3: Max RoCoF: −0.340 Hz/s Frequency Nadir: 49.800 Hz	Scenario 2: Max RoCoF: +0.312 Hz/s Frequency Peak: 50.383 Hz Scenario 3: Max RoCoF: +0.411 Hz/s Frequency Peak: 50.470 Hz
Control Actions	Wind: Instantaneous inertial power boost. PV: Reserved power injection via droop control. BESS (S3): Activated for critical power support	Wind: Primary power reduction via deloading. PV: Power curtailment via droop control. BESS (S3): Activated for excess power absorption.
Advantages & Mechanism	1. Coordinated Inertia: Wind provides instant response, PV supplements with reserves, and BESS engages only for severe deficits. 2. Enhanced System Inertia: This staged activation effectively elevates the overall system inertia support. 3. Optimized Resource Use: Minimizes BESS dependency, reserving it for critical needs.	1. Hierarchical Response: Wind leads the power reduction, PV cooperates, and BESS acts as a final sink for excess power. 2. Staged Inertia Support: This coordinated strategy provides structured and robust inertia against over-frequency. 3. Improved Stability: Prevents excessive RoCoF and generator tripping, enhancing system resilience.

.

6. Conclusions

This paper proposes a coordinated inertia support strategy for wind-PV-thermal-storage systems that explicitly considers system inertia demand. The main conclusions and contributions are summarized as follows:

• A virtual inertia evaluation method incorporating the inertia time constant was developed to quantify system inertia demand and the inertia support capability of individual units. This method provides an intuitive assessment of system inertia reserve and enables effective analysis of the impact of parameters such as wind turbine speed and energy storage capacity on overall system inertia. Simulation results verify that the method accurately assesses system frequency security requirements, establishing a critical foundation for subsequent coordinated control.
• A fuzzy-logic-based coordinated control with the inertia time constant as its output was designed to optimize the inertia evaluation results and achieve precise allocation of virtual inertia across units. To address the challenge of multi-source inertia fitting and distribution, this approach innovatively employs the inertia time constant as the fuzzy-logic-based coordinated control output, replacing traditional fixed-parameter allocation methods. Utilizing adaptive rules to dynamically optimize the control parameters for wind, PV, and storage units, this method significantly enhances allocation accuracy and system dynamic response performance, ensuring the evaluation results align closely with actual system needs.
• An integrated “evaluation-allocation-control” framework was established, where inertia evaluation results directly inform the parameter tuning of wind, PV, and storage controllers. The core innovation lies in the seamless integration of inertia assessment with controller setpoints. The evaluated system inertia demand is no longer an isolated metric but serves as a direct input for adaptively adjusting the control commands of renewable units and storage, thereby translating system-level inertia requirements into precise, coordinated unit-level actions.
• A scenario-based unit commitment strategy was proposed, which selectively activates resources according to disturbance type and severity, effectively enhancing the system’s inertia support capability under various operating conditions. By categorizing power disturbances into different intervals and types (e.g., sudden increase/decrease), this strategy enables the intelligent switching and role assignment of wind, PV, and storage units. Simulation results demonstrate its effectiveness: under a sudden load increase scenario (Scenario 2), the strategy coordinated a wind power increase of 0.102 p.u. and a PV output increase of 0.1 p.u., maintaining the frequency nadir at 49.785 Hz without requiring energy storage intervention. Under a sudden load decrease scenario (Scenario 3), the strategy directed wind power to rapidly reduce output by 0.093 p.u. and activated the energy storage system to absorb 0.6 p.u. of excess power, successfully suppressing the frequency peak to 50.47 Hz. This fully proves the strategy’s capability to flexibly and efficiently utilize system resources to handle bidirectional power disturbances while avoiding frequent actions by any single unit.