VSG control method, device and equipment of optical storage grid-connected system

Abstract

The invention provides a VSG control method, device and equipment of an optical storage grid-connected system, and relates to the technical field of power systems. The method comprises the following steps: acquiring real-time operation data of the optical storage grid-connected system; the real-time operation data comprises direct-current bus voltage, active power and alternating-current frequency; based on the real-time operation data and a second-order voltage-frequency nonlinear coupling model of the direct current-alternating current side, performing nonlinear coupling analysis on the direct current bus voltage, the active power and the alternating current frequency to obtain coupling parameters; based on the real-time operation data and the coupling parameters, a pre-constructed objective function and a control equation are updated, and the control equation is a voltage-frequency coupling equation set representing the coupling relation between the direct current side and the alternating current side; based on the objective function and the control equation, performing iterative optimization to obtain an optimal control parameter; and controlling the VSG of the optical storage grid-connected system based on the optimal control parameter. The stability of the optical storage grid-connected system can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power systems, and particularly relates to a VSG control method, device and equipment of a photovoltaic-storage grid-connected system. BACKGROUND

[0002] With the continuous increase of new energy generation penetration, the inherent inertia of the power grid is continuously reduced. The Virtual Synchronous Generator (VSG) has become the core technology of the grid-connected control of the energy storage system, because it can simulate the inertia characteristics of the synchronous generator and provide frequency and voltage support for the power grid.

[0003] However, when the photovoltaic and energy storage are connected to the grid, the DC side voltage will change, the AC side power will change, and then the frequency disturbance of the system will be affected, resulting in a decrease in the stability of the photovoltaic-storage grid-connected system. SUMMARY

[0004] Embodiments of the present application provide a VSG control method, device and equipment of a photovoltaic-storage grid-connected system to solve the problem of low stability of the photovoltaic-storage grid-connected system.

[0005] In a first aspect, the embodiments of the present application provide a VSG control method of a photovoltaic-storage grid-connected system, comprising: obtaining real-time operation data of the photovoltaic-storage grid-connected system; the real-time operation data comprising a DC bus voltage, an active power and an AC frequency; performing nonlinear coupling analysis on the DC bus voltage, the active power and the AC frequency based on the real-time operation data and a second-order voltage-frequency nonlinear coupling model of the DC-AC side to obtain coupling parameters; updating a pre-constructed objective function and a control equation based on the real-time operation data and the coupling parameters, the control equation being a voltage-frequency coupling equation set representing the coupling relationship between the DC side and the AC side; iteratively optimizing based on the objective function and the control equation to obtain optimal control parameters; and controlling the VSG of the photovoltaic-storage grid-connected system based on the optimal control parameters.

[0006] In a second aspect, the embodiments of the present application provide a VSG control device of a photovoltaic-storage grid-connected system, comprising: a communication module configured to obtain real-time operation data of the photovoltaic-storage grid-connected system; the real-time operation data comprising a DC bus voltage, an active power and an AC frequency; a processing module configured to perform nonlinear coupling analysis on the DC bus voltage, the active power and the AC frequency based on the real-time operation data and a second-order voltage-frequency nonlinear coupling model of the DC-AC side to obtain coupling parameters; update a pre-constructed objective function and a control equation based on the real-time operation data and the coupling parameters, the control equation being a voltage-frequency coupling equation set representing the coupling relationship between the DC side and the AC side; iteratively optimize based on the objective function and the control equation to obtain optimal control parameters; and control the VSG of the photovoltaic-storage grid-connected system based on the optimal control parameters.

[0007] In a third aspect, an electronic device is provided, which comprises a memory and a processor, the memory storing a computer program, and the processor implementing the method according to the first aspect or any possible implementation of the first aspect when executing the computer program.

[0008] In the embodiment of the present application, the second-order voltage-frequency nonlinear coupling model on the DC-AC side is used to capture the dynamic coupling characteristics on both sides, and the coupling parameters are quantitatively extracted through nonlinear coupling analysis of the DC bus voltage, active power and AC frequency, so as to accurately depict the complex coupling effect between the DC-AC side. Further, the coupling parameters are combined with real-time operation data, the parameters are dynamically filled in a preset framework, the objective function and the voltage-frequency coupling equation set are constructed, the dynamic change requirements of the disturbances on both sides are synchronously adapted, and the real-time and adaptability of the control model are ensured. The optimal control parameters of the virtual inertia and the damping coefficient are obtained through iterative optimization with the objective function as the optimization guide. The optimal control parameters are directly applied to the active frequency modulation process, the power oscillation and frequency disturbance caused by the transmission of the DC side fluctuation to the AC side are effectively suppressed, and the dynamic response speed and stability of the grid-connected photovoltaic storage system are improved from the coupling root. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 is a schematic diagram of a photovoltaic storage grid-connected system provided by the embodiment of the present application; Figure 2 is an implementation flowchart of a VSG control method of a photovoltaic storage grid-connected system provided by the embodiment of the present application; Figure 3 is an implementation flowchart of an improved particle swarm optimization algorithm for iterative optimization of optimal control parameters provided by the embodiment of the present application; Figure 4 is a control block diagram of a virtual inertia setting method provided by the embodiment of the present application; Figure 5 is a comparison diagram of frequency recovery simulation results in a grid-connected scenario of a grid-connected photovoltaic storage system provided by the embodiment of the present application; Figure 6 is a comparison diagram of bus voltage fluctuation recovery simulation results in a grid-connected scenario of a grid-connected photovoltaic storage system provided by the embodiment of the present application; Figure 7 is a comparison diagram of active power output fluctuation simulation results of a photovoltaic storage system in a grid-connected scenario of a grid-connected photovoltaic storage system provided by the embodiment of the present application; Figure 8 is a structural schematic diagram of a VSG control device of a photovoltaic storage grid-connected system provided by the embodiment of the present application; Figure 9 is a schematic diagram of an electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0010] Existing VSG control strategies mostly focus on frequency dynamic adjustment on the AC side, generally ignoring the implicit coupling effect between the DC side voltage and the grid frequency, resulting in that when transient faults (such as photovoltaic power fluctuation and load mutation) occur, the DC side voltage is prone to large fluctuations, which in turn affects the output stability of the inverter and weakens the frequency support capability of the VSG to the grid.

[0011] In-depth analysis shows that the DC side voltage fluctuation will directly change the modulation ratio of the inverter, and the change of the modulation ratio will affect the output voltage amplitude of the VSG, ultimately causing active power fluctuation on the AC side, forming a two-way coupling cycle of "DC voltage-modulation ratio-AC power-grid frequency". The existing voltage-power coupling model is mostly a first-order linearized model, which can only describe the single influence of the DC voltage on the active power and cannot consider the second-order coupling effect caused by the dynamic change of the modulation ratio, so it cannot fully depict the nonlinear interaction between the voltage and the frequency, resulting in insufficient precision in describing the dynamic characteristics of the system, and it is difficult to support the accurate optimization of the virtual inertia and damping parameters.

[0012] At the parameter optimization level, the traditional particle swarm optimization (PSO) algorithm often uses fixed inertia weight and learning factor, which is prone to slow convergence speed and local optimum in the later search stage; at the same time, the existing optimization objective function mostly focuses on frequency deviation or voltage deviation alone, without considering the nonlinear coupling strength and energy consumption of the energy storage, resulting in that the optimized parameters can meet the standard on a single indicator, but cannot balance the system stability, dynamic response speed and energy storage life, making it difficult to realize voltage-frequency collaborative stability control and restricting the application effect of VSG in high proportion of new energy grid.

[0013] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0014] Figure 1 A schematic diagram of a grid-connected photovoltaic and energy storage system provided by the embodiments of the present application is shown in FIG. 1. Figure 1 As shown in the figure, the grid-connected photovoltaic and energy storage system includes a DC side part, an inverter side part, an AC side part and a control part. The DC side part includes a photovoltaic array, an energy storage system and a DC bus. The inverter side part includes an inverter. The AC side part includes an AC bus and a grid-connected point. The control part includes a controller.

[0015] The output power signal of the photovoltaic array, the charge / discharge power / battery state signal of the energy storage system and the voltage / current signal of the DC bus are collected by sensors and uploaded to the controller to provide basic data for coupling analysis. The three-phase voltage, current, power angle and frequency signals of the grid-connected point and the voltage / power signals of the AC bus are collected by monitoring sensors and uploaded to the controller to support the construction of control equations and the calculation of objective functions.

[0016] After receiving uplink data from various components, the controller calculates coupling parameters based on DC / AC side data using a second-order voltage-frequency nonlinear coupling model. Combining these coupling parameters with real-time data, the objective function and control equations are updated. An improved PSO algorithm is used for iterative optimization to obtain optimal virtual inertia, damping coefficient, and other control parameters. The controller then converts these optimal control parameters into VSG control commands (such as modulation ratio and switching transistor drive signals) and sends them to the inverter to adjust the inverter's output voltage and frequency, simulating synchronous generator characteristics. Simultaneously, charging and discharging power adjustment commands are sent to the energy storage system to smooth out photovoltaic output fluctuations, stabilize the DC bus voltage, and ensure the system dynamically adapts to changes in operating conditions.

[0017] See Figure 2 The document illustrates a flowchart of the VSG control method for a photovoltaic-storage grid-connected system provided in an embodiment of the present invention, which is described in detail below: Step 201: Obtain real-time operating data of the photovoltaic-storage grid-connected system; real-time operating data includes DC bus voltage, active power, and AC frequency.

[0018] In some embodiments, the DC bus voltage is the real-time voltage value of the DC side (the connection bus between the photovoltaic panels, energy storage batteries, and inverter) of the photovoltaic-storage grid-connected system, and is a core representation of the DC side energy state. The DC bus voltage is acquired in real time by a DC side voltage sensor.

[0019] In some embodiments, active power is the real-time active power transmitted or absorbed by the photovoltaic-storage grid-connected system to the grid. It is a core indicator of the system's energy conversion and transmission efficiency, containing only the active power component. It is calculated by monitoring the effective values ​​of the three-phase AC voltage, effective values ​​of the current, and the power factor at the grid connection point, and according to relevant formulas.

[0020] In some embodiments, the AC frequency is the real-time frequency of the AC voltage at the grid connection point (or grid side) of the photovoltaic-storage grid-connected system. The standard power frequency in my country is 50Hz, which is a core indicator of grid stability. It is acquired in real-time using an AC-side frequency sensor.

[0021] Step 202: Based on real-time operating data and a second-order voltage-frequency nonlinear coupling model of the DC-AC side, perform nonlinear coupling analysis on the DC bus voltage, active power and AC frequency to obtain coupling parameters.

[0022] In some embodiments, real-time operating data may also include energy storage power, photovoltaic output power, effective value of three-phase AC voltage at grid connection point, power angle value, DC side capacitor capacity, and inverter output AC active power.

[0023] For example, energy storage power is the real-time charging and discharging power of the energy storage system (charging is negative, discharging is positive), and it is a key indicator of DC-side energy balance. Photovoltaic output power is the real-time DC power output of the photovoltaic array, dynamically changing with sunlight intensity. The effective value of the three-phase AC voltage at the grid connection point is the effective value of the three-phase AC voltage at the connection point between the system and the grid, and it is the core indicator of the AC-side voltage state. The power angle value is the phase difference between the AC voltage at the grid connection point and the grid voltage, reflecting the active power balance state of the system. The DC-side capacitor capacity is the capacitance of the capacitors connected in parallel on the DC bus, and it is a key parameter for DC-side energy buffering. The inverter output AC active power is the active power output to the AC side after the inverter converts DC to AC.

[0024] In some embodiments, the second-order voltage-frequency nonlinear coupling model on the DC-AC side is a mathematical model characterizing the dynamic correlation between the DC side (e.g., DC bus voltage) and the AC side (e.g., active power, AC frequency) of a photovoltaic-storage grid-connected system. It considers both linear coupling effects (first-order terms) and nonlinear coupling effects (second-order terms), accurately describing the complex relationship between voltage fluctuations and frequency and power changes. It is used for coupling analysis of DC bus voltage, active power, and AC frequency, quantifying the mutual influence among them. The second-order voltage-frequency nonlinear coupling model on the DC-AC side is derived based on the AC / DC coupling principle of power systems and the dynamic characteristics of VSG. The model relies on real-time operating data (e.g., DC bus voltage, power angle value) and preset parameters (e.g., equivalent line reactance, AC / DC voltage coupling coefficient).

[0025] In some embodiments, the coupling parameters include a primary voltage coupling coefficient, a secondary voltage coupling coefficient, and a frequency coupling coefficient. These coupling parameters are key parameters calculated based on a second-order coupling model and are used to quantify the coupling strength and characteristics between various physical quantities on the AC and DC sides. The primary voltage coupling coefficient characterizes the linear coupling effect of DC voltage changes on active power and AC frequency. The secondary voltage coupling coefficient characterizes the nonlinear coupling effect caused by DC voltage changes (such as frequency oscillations caused by voltage fluctuation amplification). The frequency coupling coefficient characterizes the linear coupling effect of power angle deviation on inverter output power.

[0026] As one possible implementation, step 202 can be specifically implemented as steps A11-A14.

[0027] A11: The modulation coefficient is calculated based on the DC bus voltage, set voltage reference value, initial modulation ratio, modulation ratio and AC side voltage in the real-time operation data.

[0028] A12: The primary voltage coupling coefficient is calculated based on the initial modulation ratio, grid connection point voltage amplitude, power angle steady-state value, modulation coefficient, AC / DC voltage coupling coefficient, DC bus voltage steady-state value, and equivalent line reactance.

[0029] A13: The secondary voltage coupling coefficient is calculated based on the modulation coefficient, AC / DC voltage coupling coefficient, grid connection point voltage amplitude, equivalent line reactance, and steady-state power angle value; the secondary voltage coupling coefficient is used to characterize the nonlinear coupling effect caused by DC voltage changes.

[0030] A14: Based on the initial modulation ratio, steady-state value of DC bus voltage, voltage amplitude at grid connection point, equivalent line reactance, set voltage reference value, modulation coefficient, AC / DC voltage coupling coefficient, and steady-state value of power angle, the linear coupling effect characterizing the power angle deviation on the inverter output power is calculated and used as the frequency coupling coefficient.

[0031] In some embodiments, the modulation coefficient characterizes the dynamic parameter of the amplitude ratio of the modulating signal to the carrier signal during the inverter PWM modulation process, reflecting the modulation depth.

[0032] In some embodiments, the set voltage reference value is the system's preset target AC voltage value at the grid connection point, which conforms to the grid connection standard (such as 380V / 110kV in my country).

[0033] In some embodiments, the initial modulation ratio is the initial amplitude ratio of the inverter PWM modulation, which is the initial parameter of the modulation strategy.

[0034] In some embodiments, the modulation ratio is the amplitude ratio of the real-time PWM modulation of the inverter, which is dynamically adjusted according to the control command.

[0035] In some embodiments, the AC side voltage is the real-time AC voltage of the AC bus or grid connection point, which is feedback data of the modulation strategy.

[0036] In some embodiments, the AC / DC voltage coupling coefficient is an inherent coefficient characterizing the degree of mutual influence between DC voltage and AC voltage, reflecting the voltage correlation characteristics of the AC / DC side.

[0037] In some embodiments, the grid connection point voltage amplitude is the amplitude of the three-phase AC voltage at the grid connection point, which is derived from the effective value of the three-phase voltage.

[0038] In some embodiments, the steady-state value of the power angle is the steady-state component of the power angle, which is obtained by filtering the real-time power angle value.

[0039] In some embodiments, the steady-state value of the DC bus voltage is the steady-state component of the DC bus voltage, which is obtained by filtering the real-time DC bus voltage.

[0040] In some embodiments, the equivalent line reactance is the combined equivalent reactance of the AC-side filter inductance and the line inductance, reflecting the impedance characteristics of AC-side power transmission.

[0041] As one possible implementation, the embodiments of the present invention clearly define the basis for calculating the coupling parameters, the specific objects involved, and the derivation logic.

[0042] Step 203: Based on real-time running data and coupling parameters, update the pre-constructed objective function and control equations. The control equations are a set of voltage-frequency coupling equations characterizing the coupling relationship between the DC side and the AC side.

[0043] In some embodiments, the objective function is a mathematical expression that quantifies the system control optimization objective, integrating multi-dimensional indicators such as voltage stability, frequency stability, coupling suppression, energy storage adaptation, and inertia constraint, and is used to evaluate the quality of control parameters (virtual inertia, damping coefficient).

[0044] In some embodiments, the governing equations are a set of mathematical equations characterizing the dynamic characteristics of the AC and DC sides of a photovoltaic-storage grid-connected system. They primarily describe the causal relationships between DC bus voltage, active power, AC frequency, and control parameters (virtual inertia, damping coefficient), serving as a mathematical mapping of the system's dynamic behavior. The governing equations accurately describe the AC and DC side coupling mechanisms (linear and nonlinear), providing a tool for predicting system response and optimizing control parameters.

[0045] In some embodiments, the voltage-frequency coupling equations are the core representation of the control equations, specifically focusing on the coupling relationship between DC-side voltage and AC-side frequency and active power. They include the DC energy balance equations, the active power expression (containing linear / nonlinear coupling terms), and the VSG active-frequency dynamic equations, which together form a closed loop. The equations can accurately describe coupling phenomena such as DC voltage fluctuations and AC frequency oscillations.

[0046] In some embodiments, the pre-constructed objective function includes a voltage deviation term, a frequency deviation term, a nonlinear coupling strength term, an energy storage output power term, and a virtual inertia dynamic constraint term.

[0047] In some embodiments, the voltage deviation term is a mathematical term that quantifies the degree of deviation between the real-time value and the rated value of the DC bus voltage, reflecting the DC-side voltage stability. Minimizing the voltage deviation ensures that the DC bus voltage remains stable within the allowable range.

[0048] In some embodiments, the frequency deviation term is a mathematical term that quantifies the degree of deviation between the real-time AC frequency value and the grid's rated frequency (50Hz in my country), reflecting the frequency synchronization between the AC side and the grid. Minimizing the frequency deviation complies with grid connection standards.

[0049] In some embodiments, the nonlinear coupling strength is a mathematical term that quantifies the strength of the nonlinear coupling effect between the DC and AC sides, and is a core feature of the objective function of this invention. It specifically suppresses nonlinear coupling disturbances, unlike traditional objective functions that only consider linear indices.

[0050] In some embodiments, the energy storage output power term is a mathematical term that quantifies the rationality of the energy storage system's charging and discharging power, balancing energy storage peak shaving with equipment losses. This avoids overcharging and discharging of energy storage (extending battery life) and sudden power surges (impacting the DC bus).

[0051] In some embodiments, the virtual inertia dynamic constraint term is a mathematical term that quantifies the rationality of virtual inertia adjustment, constraining the range and rate of change of virtual inertia values ​​to prevent system dynamic response instability caused by excessively large or small inertia. It limits the virtual inertia to an engineering-feasible range to avoid exceeding the inverter's regulation capacity.

[0052] As one possible implementation, this embodiment of the invention solves the problem of unclear optimization direction by clearly defining the specific composition of the objective function, balances the multi-dimensional constraints of the system, and supports real-time optimization under dynamic operating conditions.

[0053] Step 204: Based on the objective function and the control equation, iteratively optimize to obtain the optimal control parameters.

[0054] As one possible implementation, step 204 can be specifically implemented as steps A21-A211.

[0055] A21: Initialize algorithm parameters, including particle swarm size, maximum number of iterations, initial inertia weights, and adaptive learning factor.

[0056] A22: Set safety constraints, which include preset ranges for voltage, frequency, energy storage power, and control parameters.

[0057] A23: Randomly initialize particle position and velocity, where particle position corresponds to the initial value of virtual inertia and damping coefficient, and particle velocity corresponds to the adjustment step size of the parameters.

[0058] A24: Based on the virtual inertia, damping coefficient, and control equations corresponding to the current particle position, combined with real-time running data and coupling parameters, the objective function value is calculated and used as the initial fitness value for each particle.

[0059] A25: Initialize the individual optimal solution and the group optimal solution based on the initial fitness value.

[0060] A26: Dynamically adjust inertia weights and adaptive learning factors.

[0061] A27: Update the particle's velocity and position based on the adjusted inertia weights, adaptive learning factor, individual optimal solution, and swarm optimal solution.

[0062] In some embodiments, safety constraints are physical / engineering constraints that must be followed during the iterative optimization process, limiting the allowable range of control parameters and system operating states.

[0063] As one possible implementation, step A27 can be specifically implemented as steps B11-B15.

[0064] B11: Based on the adjusted inertial weights and the current particle velocity, the historical velocity components are calculated.

[0065] B12: The individual optimal guidance component is calculated based on the adaptive learning factor, the difference between the particle's current position and the individual optimal solution.

[0066] B13: The optimal guidance component of the population is calculated based on the adjusted adaptive learning factor, the difference between the current position of the particle and the optimal solution of the population.

[0067] B14: Based on the coupling gain term, coupling feedback vector, and nonlinear coupling strength, the coupling correction component is calculated.

[0068] B15: Based on the historical velocity component, the individual optimal guidance component, the group optimal guidance component, and the coupling correction component, the updated velocity of the particle is obtained.

[0069] In some embodiments, the coupling gain term is a coefficient (unitless, such as 0.01-0.1) that quantifies the strength of the nonlinear coupling effect on the particle velocity correction, used to adjust the magnitude of the influence of the coupling correction component.

[0070] In some embodiments, the coupling feedback vector is a vector characterizing the DC-AC side coupling state (e.g., [DC bus voltage deviation, AC frequency deviation]). This reflects the real-time state of the coupled disturbance.

[0071] In some embodiments, nonlinear coupling strength is a physical quantity that quantifies the strength of nonlinear coupling disturbances on the AC and DC sides, and its core reflects nonlinear effects such as frequency oscillations caused by DC voltage fluctuations.

[0072] In some embodiments, the individual optimal guidance component is the velocity component that guides the particle to move toward its own historical optimal solution (individual optimal guidance component = individual learning factor × (individual optimal solution - particle current position) × random factor).

[0073] In some embodiments, the swarm optimal guidance component is the velocity component that guides the particle to move toward the global optimal solution (swarm optimal guidance component = global learning factor × (swarm optimal solution - particle current position) × random factor).

[0074] In some embodiments, the coupling correction component is a velocity correction component for the nonlinear coupling effect on the AC / DC side (coupling correction component = coupling gain term × coupling feedback vector × nonlinear coupling strength), which is the core improvement of this invention.

[0075] As one possible implementation, this embodiment of the invention designs a speed update mechanism (including targeted coupling correction) for the superposition of four-dimensional components to accurately guide the search direction of particles in the parameter space. This retains the global exploration and local convergence capabilities of the traditional PSO algorithm while adapting to the nonlinear coupling characteristics of the photovoltaic-storage grid-connected system, ultimately improving the accuracy and efficiency of finding the optimal control parameters.

[0076] As one possible implementation, step A27 can be specifically implemented as steps B21-B24.

[0077] B21: Determine the adaptive time step based on the ratio of the current iteration number to the maximum iteration number and the adjusted inertia weight.

[0078] B22: The basic adjustment amount is calculated based on the adaptive time step and the updated velocity.

[0079] B23: The trend correction amount is calculated based on the difference between the particle's current velocity and its velocity at the previous moment.

[0080] B24: Calculate the updated position of the particle based on its current position, basic adjustment amount, and trend correction amount.

[0081] In some embodiments, the base adjustment is the basic movement amplitude of the particle position calculated using an adaptive time step and the updated velocity, and is the core basic data for particle position updates.

[0082] In some embodiments, the trend correction amount is a correction parameter calculated based on the difference between the particle's current velocity and its velocity at the previous moment, used to characterize the trend of particle velocity change.

[0083] In some embodiments, the adaptive time step is a dynamic time coefficient calculated based on the optimization progress and the adjusted inertia weight, rather than a fixed value.

[0084] As one possible implementation, this embodiment of the invention designs a dual optimization mechanism of adaptive time step and trend correction to ensure that the position adjustment of particles in the parameter space is stable, accurate and adapted to the optimization process, thereby avoiding optimization distortion caused by position fluctuations and improving the quality of candidate solutions for control parameters.

[0085] A28: Based on the updated virtual inertia, damping coefficient, and control equations corresponding to the particle positions, combined with real-time running data and coupling parameters, the objective function value is calculated and used as the current fitness value of each particle.

[0086] A29: Based on the current fitness value of each particle, update the individual optimal solution and determine the population optimal solution.

[0087] A210: Repeat the steps of dynamically adjusting the inertia weights and adaptive learning factors, and the subsequent steps.

[0088] A211: If the maximum number of iterations or the fitness value converges, stop the iteration and use the inertia and damping coefficient corresponding to the population optimal solution as the optimal control parameters.

[0089] In some embodiments, the optimal control parameters are the control parameters (virtual inertia J, damping coefficient D) corresponding to the group optimal solution after the iterative optimization terminates, which are the optimal solutions adapted to the current working conditions.

[0090] In some embodiments, virtual inertia is an equivalent parameter used in the inverter to simulate the rotational inertia of a synchronous generator through a control algorithm. Essentially, it is a dynamic indicator characterizing the system's ability to resist sudden frequency changes. The rotational inertia of a synchronous generator comes from the mechanical inertia of the physical rotor, while the virtual inertia of a VSG is simulated without a physical rotor through energy storage power regulation and inverter control logic.

[0091] In some embodiments, the damping coefficient is a parameter characterizing the frequency decay oscillation and fast convergence capability of the VSG. Essentially, it is equivalent to the damping torque that suppresses frequency oscillation through active power regulation. The damping of the synchronous generator comes from physical effects such as mechanical friction and electromagnetic damping. The damping coefficient of the VSG is realized through frequency deviation-active power compensation logic.

[0092] As one possible implementation, embodiments of the present invention transform abstract optimization into concrete, safe, and efficient technical steps by clarifying the algorithm framework, setting security constraints, designing adaptive mechanisms, and defining termination conditions.

[0093] Step 205: Control the VSG of the photovoltaic-storage grid-connected system based on the optimal control parameters.

[0094] In some embodiments, the VSG in a photovoltaic-storage grid-connected system refers to the core execution unit of the system, which integrates virtual synchronous generator control logic. By simulating the inertia and damping characteristics of a synchronous generator, the photovoltaic-storage system achieves grid-friendliness similar to traditional thermal power. Optimal control parameters (virtual inertia and damping coefficients) are converted into specific control commands for the VSG, and through the coordinated action of the inverter and energy storage system, precise adjustment of the operating state of the photovoltaic-storage grid-connected system is achieved.

[0095] In this embodiment of the invention, a second-order voltage-frequency nonlinear coupling model on the DC-AC side is used to specifically capture the dynamic coupling characteristics of both sides. Through nonlinear coupling analysis of DC bus voltage, active power, and AC frequency, coupling parameters are quantified and extracted, achieving a precise characterization of the complex coupling effects on the DC-AC side. Furthermore, by combining these coupling parameters with real-time operating data, parameters are dynamically filled within a preset framework to construct an objective function and a voltage-frequency coupling equation set, synchronously adapting to the dynamic changes in disturbances on both sides and ensuring the real-time performance and adaptability of the control model. Using the objective function as an optimization guide, the optimal control parameters for virtual inertia and damping coefficient are obtained through iterative optimization. These optimal control parameters are directly applied to the active power frequency regulation process, effectively suppressing power oscillations and frequency disturbances caused by the transmission of DC-side fluctuations to the AC side, thereby improving the dynamic response speed and stability of the photovoltaic-storage grid-connected system from the root of the coupling.

[0096] As one possible implementation, steps A31-A33 can be performed before step A12.

[0097] A31: The steady-state value of the DC bus voltage is calculated based on the DC bus voltage in the real-time operating data.

[0098] A32: The steady-state value of the power angle is calculated based on the power angle value in the real-time operating data.

[0099] A33: The voltage amplitude at the grid connection point is calculated based on the effective value of the three-phase AC voltage at the grid connection point in the real-time operating data.

[0100] In some embodiments, the steady-state value of the DC bus voltage is a stable component obtained by filtering or moving average processing of real-time DC bus voltage data, thus eliminating transient voltage fluctuations.

[0101] In some embodiments, the steady-state value of the power angle is a stable component obtained by filtering or moving average processing the real-time power angle data, thus eliminating instantaneous fluctuations in the power angle.

[0102] In some embodiments, the voltage amplitude at the grid connection point is the peak AC voltage calculated by formula based on the effective value of the three-phase AC voltage at the grid connection point, and is a direct representation of the AC side voltage intensity.

[0103] As one possible implementation, this embodiment of the invention extracts the stationary component of real-time data and converts it into a voltage representation form, providing high-quality basic data for the accurate calculation of subsequent coupling parameters, thus avoiding calculation distortion caused by transient fluctuations or data form mismatch.

[0104] As one possible implementation, steps A41-A44 can be performed before step 203.

[0105] A41: Based on the historical operating data of the photovoltaic-storage grid-connected system, including DC-side capacitor capacity, DC bus voltage, photovoltaic output power, energy storage power, and inverter output active power, construct a DC energy balance equation.

[0106] A42: Based on the calculated value of AC output voltage amplitude, the effective value of three-phase AC voltage at the grid connection point, the power angle value, and the equivalent line reactance, and combined with coupling parameters, an active power expression containing linear and nonlinear coupling effects is constructed.

[0107] A43: Based on virtual inertia, damping coefficient and real-time frequency data, construct the VSG active-frequency dynamic equation.

[0108] A44: Based on the DC energy balance equation, the active power expression, and the VSG active-frequency dynamic equation, the pre-constructed control equation is obtained.

[0109] In some embodiments, historical operating data is key data recorded during the past operation of the photovoltaic-storage grid-connected system (including DC-side capacitor capacity, DC bus voltage, photovoltaic output power, etc.), which serves as the basic data source for offline equation construction.

[0110] In some embodiments, the calculated AC output voltage amplitude is the AC side voltage amplitude calculated based on system topology parameters and control commands, and is a theoretical reference value for the active power expression.

[0111] In some embodiments, the DC energy balance equation is a mathematical equation based on the principle of energy conservation that describes the relationship between energy input, output and storage on the DC side (photovoltaic, energy storage, capacitor).

[0112] In some embodiments, the active power expression is a mathematical expression that characterizes the relationship between the inverter's output active power and key AC / DC side parameters, while incorporating the influence of the primary voltage coupling coefficient and the secondary voltage coupling coefficient.

[0113] In some embodiments, the VSG active-frequency dynamic equation simulates the mathematical equation relating the active power and frequency dynamic response of a synchronous generator, with core associations including virtual inertia, damping coefficient, frequency deviation, and active power adjustment.

[0114] As one possible implementation, this invention integrates three dimensions—DC energy balance, AC / DC coupling power relationship, and VSG dynamic characteristics—to construct a complete mathematical model that can accurately characterize the energy flow, coupling effect, and dynamic response of a photovoltaic-storage grid-connected system. This provides a scientific and practical framework for subsequent dynamic updates and iterative optimization of control equations, serving as a key bridge connecting system modeling and optimized control.

[0115] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0116] The above embodiments are in Figure 2 Based on the method shown, each step will be discussed in detail. To facilitate understanding of the complete execution process, the overall method flow will be discussed below with reference to an embodiment.

[0117] The system studied in this invention consists of a photovoltaic array outputting power to a DC bus via a DC-DC converter, an energy storage system connected in parallel with the photovoltaic array on the bus, and then connected to the grid via a DC-AC inverter. In grid-connected scenarios, grid-based control provides stable voltage-frequency support for the power grid. However, during transient faults, the output power of the energy storage system is prone to rapid changes. DC-side voltage fluctuations affect the modulation ratio of the inverter, thereby causing changes in AC output power and creating a bidirectional coupling effect between frequency and voltage. Therefore, it is necessary to establish a voltage-frequency coupling model that includes the dynamics of both the DC and AC sides.

[0118] In a photovoltaic-storage grid-connected system, the inverter adopts grid-type VSG control, and its DC energy balance equation is:

[0119] In the formula, C dc For DC side capacitors, U dc This is the DC bus voltage. P pv , P bat and P e These are the active power outputs of photovoltaics, energy storage, and inverters, respectively.

[0120] Among them, the inverter outputs active power P e The expression is:

[0121] In the formula: X For the equivalent line reactance, U s The grid connection point voltage amplitude, E is the output voltage amplitude, and the power angle is... .

[0122] If modulation ratio is taken into account m The effect of this is that the output voltage amplitude is:

[0123] therefore:

[0124] This reveals that active power is simultaneously affected by DC bus voltage. U dc With the angle The dual effects of (i.e., frequency dynamics) are the source of nonlinear voltage-frequency coupling.

[0125] Now, let's establish the relationship between the active power output of the inverter. P e Small-signal linearization model, Steady-state operating point:

[0126] In the formula, U dc0 This is the steady-state value of the DC bus voltage. δ 0 represents the steady-state value of the work angle.

[0127] At the steady-state operating point ( U dc0 , δ 0) Expand the first-order linearization in the vicinity:

[0128] In the formula, This refers to the deviation in the inverter's output power. This refers to the DC bus voltage deviation. This is a small signal deviation of the power angle.

[0129] The expression for the first-order differential component is:

[0130] In the formula, m 0 represents the initial modulation ratio. U dc0 This is the steady-state value of the DC bus voltage. δ 0 represents the steady-state value of the work angle.

[0131] therefore,

[0132] In the formula, K v and K δ These are the primary voltage coupling coefficient and frequency coupling coefficient in the first-order model, respectively. K v and K δ The expression is:

[0133]

[0134] However, in order to represent the nonlinear power coupling effect caused by the change of DC voltage itself, it is necessary to consider the extended model of the modulation ratio dynamics and construct a second-order voltage-power coupling model.

[0135] Modulation ratio m It is inherently affected by the voltage outer loop control, and can be set as follows:

[0136] In the formula, K m The modulation coefficient, U ref To set the voltage reference value, U ac This is the AC side voltage.

[0137] AC side voltage U ac With DC side voltage U dc Dynamic coupling, i.e.:

[0138] In the formula, k v1 This represents the AC / DC voltage coupling coefficient.

[0139] therefore,

[0140] At this point, the expression for active power is:

[0141] After rearranging and linearizing, we get:

[0142] In the formula, K v1 This is the primary voltage coupling coefficient. K v2 The voltage coupling coefficients are for both voltage coupling operations. This refers to the deviation in the inverter's output power. This represents the power angle deviation. The expressions for the primary and secondary voltage coupling coefficients in the second-order model are:

[0143]

[0144] The expression for the angle deviation is:

[0145] In the formula, f For frequency.

[0146] Because the photovoltaic-storage grid-connected system adopts a grid-type VSG control method, the active power-frequency control link in the control structure is combined with the voltage-power coupling relationship. A nonlinear voltage-frequency coupled dynamic equation set is constructed, which fully describes the interaction between voltage fluctuations, frequency disturbances, and energy storage power dynamics.

[0147] In conjunction with the active power-frequency control loop of the VSG and the voltage-power coupling relationship:

[0148] In the formula, equation one is the active power-frequency equation of the VSG control loop, and equation two is the DC voltage equation. J Here, D is the virtual inertia, and D is the damping coefficient. P m To output mechanical power to the system's virtual synchronizer, I bat , I inv These are the energy storage current and the inverter input current, respectively.

[0149] Small perturbation linearization selects a stable point:

[0150] In the formula, P e0 This represents the steady-state value of the inverter's output power. This represents the deviation in the inverter's output power.

[0151] And in the second-order linear model Substituting into the active-frequency equation, the first-order linear model... Substituting into the DC voltage equation, we get:

[0152] In the formula, P e0 This represents the steady-state value of the inverter's output power. It is the steady-state value of the frequency.

[0153] The established nonlinear voltage-frequency coupling dynamic equations characterize in detail the dynamic coupling process between bus voltage and grid frequency.

[0154] Construct a multi-objective function based on voltage-frequency integrated performance indicators; The objective function is:

[0155] In the formula: α , β , γ, ηThese are the weighting coefficients for frequency, voltage, nonlinear coupling, and energy storage output power, respectively, and they satisfy... α + β + γ + η = λ 1; J ’ This refers to the dynamic constraint term for virtual inertia. The term directly quantizes the second-order coupling into part of the objective, and squaring it can enhance the penalty for strong nonlinearity; It is the energy storage output power, which punishes excessive energy output and frequent high-power operation to protect the energy storage and its lifespan. T The assessment window includes the period between the occurrence of the disturbance and the basic recovery period of the system.

[0156] This includes adding a virtual inertia dynamic constraint term. J ’ Its expression is:

[0157] This formula reflects the virtual inertia. J Sensitivity constraints, where, μ This is the coupling sensitivity penalty coefficient, set between 0.1 and 0.3, choosing the smallest possible value to avoid excessive computational overhead. It is the peak value of the frequency.

[0158] The constraints are: During the evaluation window T Internally, the inverter output active power and DC current constraints are as follows:

[0159] In the formula, It is the rated active power of the inverter.

[0160] Similarly, the DC-side current constraint is:

[0161] In the formula, It is the maximum threshold for DC current.

[0162] If the inverter's rated active power is set... =1MW, minimum allowable DC bus voltage U dc =1200V, then the maximum DC current is:

[0163] If a candidate in simulation or actual operation J Cause instantaneous Therefore, this result must be considered infeasible.

[0164] Battery energy constraints: Calculate energy storage consumption (discharging or charging) when evaluating disturbance scenarios, with the following requirements:

[0165] In the formula, For energy storage energy consumption, For available energy storage capacity, Figure 3 To ensure safety and avoid overcharging or over-discharging of energy storage.

[0166] For grid-connected systems, we typically need to meet the voltage and frequency limits required by the power system.

[0167] Frequency deviation extreme values:

[0168] In the formula, To limit the amplitude of frequency fluctuations, set .

[0169] The constraint condition for the rate of change of frequency (ROCOF) is:

[0170] Based on the actual values ​​that meet grid connection requirements, set... .

[0171] DC bus voltage limit:

[0172] The DC bus voltage limit value in the formula is:

[0173] If any scenario violates the above safety limits, the result should be judged as infeasible or subject to a maximum penalty in the objective.

[0174] The improved particle swarm optimization (PSO) algorithm based on nonlinear coupled dynamic feedback is adopted, and its operation steps are as follows: Figure 4 The improved particle swarm optimization algorithm flowchart is used for iterative solution. In the operation of the improved particle swarm optimization algorithm (PSO) based on nonlinear coupling dynamic feedback in this invention, initialization is performed first, and each particle is set as a set of virtual inertia and damping parameter combinations. J i , D i At the same time, configure the initial position and initial velocity of the particle swarm, as well as the parameters required by the algorithm (including relevant parameters to be set). T *The process then proceeds to the update iteration process. In each iteration, the adaptive inertia weight, adaptive learning factor, and nonlinear coupling dynamic feedback process are dynamically calculated. Next, position updates are performed, and the velocity and position of the particles are adjusted according to the standard update rules of the PSO algorithm using the updated parameters. Then, evaluation and updates are performed, and the fitness of each particle (a fitness function constructed based on comprehensive frequency deviation, voltage deviation, and other indicators) is evaluated. The individual optimal solution of the particle and the global optimal solution of the entire particle swarm are updated based on the fitness results obtained from the evaluation. Finally, a convergence judgment is performed to check whether the preset convergence conditions are met (specifically, reaching the maximum number of iterations or the fitness value reaching a preset threshold). If the convergence conditions are met, the iteration stops; otherwise, the process returns to the update iteration step to continue the next iteration.

[0175] The formula for updating the particle swarm velocity in the particle swarm algorithm is:

[0176] In the formula, For the first i The velocity vector of each particle corresponds to the virtual inertia. J i , D i The rate of change; For the first i The position vector of each particle represents the current inertia-damping combination ( J i , D i ); For the first i The particle iterates to the... t The optimal position of an individual in a generation; This represents the optimal position in the particle's history. For adaptive dynamic inertia weights; and Adaptive learning factor; and A random number in the range [0,1]; This is the voltage-frequency coupling gain term; This is the voltage-frequency coupled feedback vector; k 2 represents the second-order nonlinear correction coefficient; The term represents the squared term of the vector, enhancing the diversity of the search.

[0177] The formula for updating the particle swarm position in the particle swarm algorithm is:

[0178] In the formula, This is the dynamic time step coefficient, associated with the inertia weight, which balances the convergence speed; To weight the acceleration term, the current velocity change trend is used to prevent oscillations. The expressions are as follows:

[0179]

[0180] In the formula, T max To determine the maximum number of iterations, t This represents the current iteration number. It is the time inertia coefficient, used to control the convergence speed during the iteration process. It is the oscillation coefficient, used to control the rate of change of the current speed, and can be set by the user.

[0181] Adaptive design of dynamic inertia weights and adaptive learning factor design. Traditional PSO inertia weights. The current approach only decreases linearly and lacks group feedback. This invention defines an adaptive weighting mechanism based on group fitness feedback:

[0182] In the formula, This represents the standard deviation of the current population fitness. Current average fitness of the population; This is an adjustment coefficient used to control the sensitivity to weight changes. When the population diversity is high, i.e. big, When the group is large, it continues to explore; when the group converges, that is... Small, Smaller values ​​accelerate convergence.

[0183] The expression for the standard deviation of population fitness is:

[0184] Current population average fitness The expression is:

[0185] In the formula, It is the first i The particle in the first t The fitness of the next iteration.

[0186] To comprehensively evaluate the impact of different virtual inertia J and damping coefficient D on the system's voltage-frequency dynamic performance, a fitness function f(J,D) is defined to characterize the overall performance of the system in both steady-state and transient states. The fitness function is constructed based on the error integral of key physical quantities and constraint penalty terms, considering frequency deviation, DC-side voltage fluctuations, and also taking into account energy storage and power limitations. It is defined as follows:

[0187] In the formula, , and For the weighting coefficients, satisfying ; For power grid frequency deviation, This refers to the DC bus voltage deviation. This refers to the equivalent energy loss or fluctuation in energy storage power. For violations of constraints (such as Penalties for exceeding limits or power limits.

[0188] A smaller fitness function f(J,D) indicates a better dynamic response and better constraint satisfaction of the system. The improved particle swarm optimization algorithm uses f(J,D) as the search objective in each iteration. By updating the velocity and position of each candidate solution in the particle swarm, it minimizes the weighted sum of system frequency deviation, DC voltage fluctuation, and energy loss, thereby obtaining the optimal virtual inertia parameter Jbest and damping parameter Dbest.

[0189] Define the ratio of particle distance to the global optimal distance :

[0190] In the formula, This is the optimal position in the particle's history. For the first i The position vectors of each particle. For the first j The position vectors of each particle.

[0191] The adaptive learning factor is defined as:

[0192] In the formula, the range of optional parameters is: =0.5, =2.5; =0.5, =2.5.

[0193] When the particle moves away from the optimal point Larger size enhances individual search capabilities; when particles approach their optimal point, Small size enhances global convergence. The learning factor self-adjusts with spatial distance, avoiding getting trapped in local optima, while automatically achieving dynamic convergence through "far-to-near" search.

[0194] The nonlinear coupling dynamic feedback process is as follows: In the nonlinear voltage-frequency model of this invention, we have:

[0195] Its linearized coupling relationship can be approximated as:

[0196] In the formula, , , and These are the partial derivatives of the various parameters that are coupled together, illustrating the coupling process.

[0197] Feedback to speed updates:

[0198] In the formula, The coupling gain factor is initially set between 0.01 and 0.1. The hyperbolic tangent function tanh suppresses excessive gradients and ensures stability.

[0199] An improved particle swarm optimization algorithm is used to input optimized parameters into the VSG control module to adjust inertia injection and coordinate the tuning of the virtual synchronous machine VSG control power loop parameters, achieving voltage-frequency coordinated stable control. The optimal parameter set (Jbest, Dbest) obtained by the improved particle swarm optimization algorithm is safely, in real-time, and controllably written into the VSG's active power-frequency control loop, enabling the inverter to inject / withdraw energy according to the new virtual inertia. This improves the voltage-frequency coupling dynamic performance without violating power, current, DC voltage, safety, and stability constraints.

[0200] In this invention, the optimal parameters obtained based on the improved particle swarm optimization algorithm are... J best , D best The control equations are dynamically introduced. To prevent control oscillations caused by sudden parameter changes, a smooth transition function is defined:

[0201] In the formula, and Initial inertia and damping settings and For inertia and damping smoothing factor, both are in the range (0,1); and The typical selection of the adjustment function that reflects the coupling between DC voltage and frequency deviation is as follows:

[0202] In the formula, , and This is the adjustment coefficient.

[0203] The VSG active power-frequency control loop control equations are updated as follows:

[0204] After the virtual inertia and damping coefficient are injected in the control stage, the above equation is combined with the DC-side dynamic equation (and... U dc (Related) to form a voltage-frequency coupled equation set, realizing the linkage adjustment of inertia and damping with the system state.

[0205] Figure 1 This is a block diagram illustrating the control principle of the method of the present invention. The diagram shows the virtual inertia of the active-frequency element in VSG control, based on a multi-objective function and improved by using a particle swarm optimization algorithm with nonlinear coupled dynamic feedback. J and damping coefficient D The process involves iterative optimization. Finally, the optimized parameters are input into the VSG control module to achieve inertia injection between energy storage and the grid, and voltage-frequency coordinated stability control, effectively suppressing the bidirectional coupling effects between voltage fluctuations, power fluctuations, and frequency oscillations. The simulation model of the photovoltaic-storage grid-connected system of this invention is built on the Matlab / Simulink simulation platform. Figure 5 A simulation model was built to establish the theoretical framework. The rated voltage of the energy storage device is 1200V. The simulation mimics the grid connection process. The simulation time is set to 20s. The experimental procedure is as follows: during the grid connection process, the initial large-capacity load is connected at 5s, and the remaining load is gradually connected at 10s, 13s, and 16s to simulate the entire grid connection process. The grid connection stability is judged by sudden changes in grid-side load.

[0206] like Figure 6 As shown, after tuning the virtual inertia and damping parameters of the photovoltaic-storage grid-connected system using the parameter optimization method proposed in this invention, the system exhibits superior dynamic characteristics during grid-connection transients and load step disturbances. The maximum frequency drop of the optimized system is significantly reduced, the frequency fluctuation curve becomes smoother, and it can quickly converge and stably recover to its rated value after the disturbance. These results demonstrate that the method described in this invention effectively improves the dynamic response capability and active support level of the photovoltaic-storage grid-connected system to frequency deviations, avoiding system oscillation and stability risks caused by insufficient inertia response or delayed frequency recovery under traditional control. The particle swarm optimization algorithm based on nonlinear coupled dynamic feedback can accurately tune key parameters in the control of the virtual synchronous generator (VSG), including the virtual inertia. J With damping coefficient D This method achieves dynamic decoupling of the voltage-frequency coupling link, enabling the system to maintain good frequency stability and robustness under complex disturbances. Overall simulation results fully verify the significant advantages of the proposed method in improving the inertial response quality of photovoltaic-storage systems, shortening frequency recovery time, and suppressing overshoot.

[0207] like Figure 7As shown, the optimized photovoltaic-storage system, after this invention, exhibits significantly reduced DC bus voltage fluctuations and improved voltage stability during grid connection and load switching. At the moment of large-capacity load connection at 6 seconds, the energy storage unit rapidly releases power to provide active voltage support, allowing the DC bus voltage to quickly return to a steady state without significant deviation from the rated value. When other loads are connected sequentially at 10 seconds, 13 seconds, and 16 seconds, bus voltage fluctuations are effectively suppressed, with amplitudes far lower than before optimization. Full-process monitoring results demonstrate that the method described in this invention effectively decouples DC voltage from grid frequency, significantly reduces the sensitivity of the DC side voltage to load disturbances, and ensures stable system power output without oscillations or overshoot. These results fully illustrate that the optimized VSG control strategy can achieve coordinated and stable operation under strong voltage-frequency coupling conditions, providing reliable control support for the safe and efficient grid connection of the photovoltaic-storage system.

[0208] like Figure 8 As shown, the optimized photovoltaic-storage system of this invention, during grid connection and load switching, to better demonstrate the power point tracking (PPT) effect and fluctuation suppression effect, assumes a certain level of illuminance in the initial state of the photovoltaic array, similar to constant power control. During the load switching phase, the photovoltaic-storage system switches from constant power control to maximum power point tracking (MPPT) control, highlighting the change in output active power. When a large-capacity load is connected at 6 seconds, power fluctuations are significantly suppressed. When the remaining loads are connected sequentially at 10 seconds, 13 seconds, and 16 seconds, active power fluctuations are effectively suppressed, with amplitudes far lower than before optimization. The method described in this invention effectively decouples the DC voltage from the grid frequency, significantly reducing the sensitivity of the DC voltage to output power fluctuations. The system power output gradually stabilizes without oscillations or overshoot.

[0209] The core objective of this invention is to improve upon a series of technical shortcomings in existing grid-connected virtual synchronous generator (VSG) control systems. These shortcomings include: insufficient incorporation of the dynamic coupling effect between DC-side voltage and grid frequency; insufficient accuracy of the coupling model; limitations in the search accuracy and convergence performance of the parameter optimization algorithm; and the failure of the optimization objective function to fully reflect the multi-dimensional performance requirements of the system. In existing control systems, when the DC bus voltage shifts due to load disturbances or energy fluctuations, it leads to changes in the inverter modulation ratio, causing transient fluctuations in active power output, which in turn triggers grid frequency oscillations and power instability. To address these issues, this invention proposes a virtual inertia setting scheme with a clear modeling approach and highly feasible algorithm structure. This scheme establishes a second-order voltage-frequency coupling model by accurately characterizing the transmission mechanism of DC-side voltage changes to AC-side power and frequency dynamics, and constructs a multi-objective optimization function that considers voltage stability, frequency response characteristics, and energy storage power constraints. Furthermore, it employs an improved particle swarm optimization algorithm based on nonlinear dynamic feedback to perform global optimization and real-time adaptive tuning of the virtual inertia and damping parameters. This strategy effectively suppresses the transmission effect of DC-side voltage fluctuations to the power channel, weakening their impact on system frequency and power stability at the source. It enables the energy storage system to actively support voltage disturbances and precisely inject inertia, thus achieving the dual goals of voltage-frequency decoupling and coordinated stable control. Ultimately, this scheme not only significantly reduces the power oscillation amplitude caused by DC-side voltage fluctuations and improves the dynamic response speed and operational stability of the grid-connected system, but also considers the energy utilization efficiency and lifespan of energy storage devices, providing reliable technical support for the safe, stable, and efficient operation of the power grid under conditions of high-proportion renewable energy grid integration.

[0210] This invention breaks through the limitations of existing first-order linear models at the modeling level, establishing a second-order voltage-power coupling model that incorporates the dynamics of the inverter modulation ratio. By introducing the coupling relationship between the modulation ratio and AC / DC voltage, it fully characterizes the second-order influence of DC voltage on active power. This model is then combined with the VSG inertia loop equation to form a voltage-frequency coupling state equation that accurately describes the dynamic interaction between voltage fluctuations, frequency disturbances, and energy storage power, significantly improving the model's accuracy in describing the system's coupling characteristics. In constructing the optimization objective function, this invention abandons the traditional single-index approach, constructing a multi-objective function that integrates frequency deviation, voltage deviation, nonlinear coupling strength, and energy storage energy consumption. It adds a virtual inertia dynamic constraint term and a safety boundary penalty mechanism to achieve multi-objective collaborative optimization of system stability, dynamic response, and energy storage lifetime. In terms of optimization algorithms, this invention proposes an improved PSO algorithm based on nonlinear coupling dynamic feedback. It designs an adaptive inertia weight based on population fitness feedback and an adaptive learning factor that adjusts with the global optimal distance of particles. Simultaneously, it introduces the voltage-frequency coupling gain term into the speed update formula, effectively solving the problem of traditional PSO easily getting trapped in local optima and improving the accuracy and convergence speed of optimal parameter search.

[0211] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0212] Figure 8 A schematic diagram of the VSG control device for a photovoltaic-storage grid-connected system provided in an embodiment of the present invention is shown. For ease of explanation, only the parts relevant to the embodiment of the present invention are shown, and are described in detail below: like Figure 9 As shown, the VSG control device 8 of the photovoltaic-storage grid-connected system includes: a communication module 81 for acquiring real-time operating data of the photovoltaic-storage grid-connected system; the real-time operating data includes DC bus voltage, active power, and AC frequency; a processing module 82 for performing nonlinear coupling analysis on DC bus voltage, active power, and AC frequency based on the real-time operating data and a second-order voltage-frequency nonlinear coupling model of the DC-AC side to obtain coupling parameters; updating the pre-constructed objective function and control equations based on the real-time operating data and coupling parameters, the control equations being a set of voltage-frequency coupling equations characterizing the coupling relationship between the DC side and the AC side; iteratively optimizing based on the objective function and control equations to obtain the optimal control parameters; and controlling the VSG of the photovoltaic-storage grid-connected system based on the optimal control parameters.

[0213] Figure 9 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. For example... Figure 9 As shown, the electronic device 9 of this embodiment includes a processor 90 and a memory 91. The memory 91 stores a computer program 92. When the processor 90 executes the computer program 92, it implements the steps in the various method embodiments described above. Alternatively, when the processor 90 executes the computer program 92, it implements the functions of each module / unit in the various device embodiments described above.

[0214] For example, computer program 92 may be divided into one or more modules / units, which are stored in memory 91 and executed by processor 90 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 92 in electronic device 9.

[0215] Electronic device 9 may include, but is not limited to, processor 90 and memory 91. Those skilled in the art will understand that... ​ This is merely an example of electronic device 9 and does not constitute a limitation on electronic device 9. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 9 may also include input / output devices, network access devices, buses, etc.

[0216] For the sake of simplicity and clarity, only the above-described functional modules / units are used as examples. In practical applications, the functions described above can be assigned to different functional modules / units as needed. These modules / units can be implemented in hardware, software, or a combination of both.

[0217] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0218] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A VSG control method for a photovoltaic-storage grid-connected system, characterized in that, include: Acquire real-time operating data of the photovoltaic-storage grid-connected system; the real-time operating data includes DC bus voltage, active power, and AC frequency; Based on real-time operating data and a second-order voltage-frequency nonlinear coupling model of the DC-AC side, nonlinear coupling analysis is performed on DC bus voltage, active power and AC frequency to obtain coupling parameters; Based on the real-time operating data and the coupling parameters, the pre-constructed objective function and control equations are updated. The control equations are a set of voltage-frequency coupling equations characterizing the coupling relationship between the DC side and the AC side. Based on the objective function and the control equation, the optimal control parameters are obtained through iterative optimization. Based on the optimal control parameters, the VSG of the photovoltaic-storage grid-connected system is controlled.

2. The VSG control method for a photovoltaic-storage grid-connected system according to claim 1, characterized in that, The real-time operating data also includes energy storage power, photovoltaic output power, effective value of three-phase AC voltage at the grid connection point, power angle value, DC side capacitor capacity, and inverter output AC active power; the coupling parameters include primary voltage coupling coefficient, secondary voltage coupling coefficient, and frequency coupling coefficient. Based on real-time operating data and a second-order voltage-frequency nonlinear coupling model of the DC-AC side, nonlinear coupling analysis is performed on the DC bus voltage, active power, and AC frequency to obtain coupling parameters, including: The modulation coefficient is calculated based on the DC bus voltage, set voltage reference value, initial modulation ratio, modulation ratio and AC side voltage from real-time operating data. The primary voltage coupling coefficient is calculated based on the initial modulation ratio, grid connection point voltage amplitude, power angle steady-state value, modulation coefficient, AC / DC voltage coupling coefficient, DC bus voltage steady-state value, and equivalent line reactance. The secondary voltage coupling coefficient is calculated based on the modulation coefficient, AC / DC voltage coupling coefficient, grid connection point voltage amplitude, equivalent line reactance, and steady-state power angle value; the secondary voltage coupling coefficient is used to characterize the nonlinear coupling effect caused by DC voltage changes. Based on the initial modulation ratio, the steady-state value of the DC bus voltage, the voltage amplitude at the grid connection point, the equivalent line reactance, the set voltage reference value, the modulation coefficient, the AC / DC voltage coupling coefficient, and the steady-state value of the power angle, the linear coupling effect characterizing the power angle deviation on the inverter output power is calculated and used as the frequency coupling coefficient.

3. The VSG control method for a photovoltaic-storage grid-connected system according to claim 2, characterized in that, Before calculating the primary voltage coupling coefficient based on the initial modulation ratio, grid connection point voltage amplitude, steady-state power angle value, the modulation coefficient, AC / DC voltage coupling coefficient, steady-state DC bus voltage value, and equivalent line reactance, the following steps are included: The steady-state value of the DC bus voltage is calculated based on the DC bus voltage in real-time operating data. The steady-state value of the power angle is calculated based on the power angle value in the real-time operation data; The voltage amplitude at the grid connection point is calculated based on the effective value of the three-phase AC voltage at the grid connection point from the real-time operating data.

4. The VSG control method for a photovoltaic-storage grid-connected system according to claim 1, characterized in that, Before updating the pre-built objective function and governing equations based on the real-time running data and the coupling parameters, the process further includes: Based on the historical operating data of the photovoltaic-storage grid-connected system, including DC-side capacitor capacity, DC bus voltage, photovoltaic output power, energy storage power, and inverter output active power, a DC energy balance equation is constructed. Based on the calculated value of AC output voltage amplitude, the effective value of three-phase AC voltage at the grid connection point, the power angle value, and the equivalent line reactance, combined with the coupling parameters, an active power expression containing linear and nonlinear coupling effects is constructed. Based on virtual inertia, damping coefficient and real-time frequency data, the active-frequency dynamic equation of VSG is constructed; Based on the DC energy balance equation, the active power expression, and the VSG active-frequency dynamic equation, the pre-constructed control equation is obtained.

5. The VSG control method for a photovoltaic-storage grid-connected system according to claim 1, characterized in that, The pre-built objective function includes: Voltage deviation term, frequency deviation term, nonlinear coupling strength, energy storage output power term, and virtual inertia dynamic constraint term.

6. The VSG control method for a photovoltaic-storage grid-connected system according to claim 1, characterized in that, The process of iteratively optimizing based on the objective function and the control equation to obtain the optimal control parameters includes: Initialize the algorithm parameters, including particle swarm size, maximum number of iterations, initial inertia weight, and adaptive learning factor; Set safety constraints, which include preset ranges for voltage, frequency, energy storage power, and control parameters; Randomly initialize the particle position and velocity, where the particle position corresponds to the initial value of the virtual inertia and damping coefficient, and the particle velocity corresponds to the adjustment step size of the parameters; Based on the virtual inertia, damping coefficient, and control equation corresponding to the current particle position, combined with real-time running data and coupling parameters, the objective function value is calculated and used as the initial fitness value of each particle. Based on the initial fitness value, initialize the individual optimal solution and the group optimal solution; Dynamically adjust inertia weights and adaptive learning factors; The velocity and position of the particles are updated based on the adjusted inertia weights, adaptive learning factors, individual optimal solutions, and swarm optimal solutions. Based on the updated virtual inertia, damping coefficient, and control equations corresponding to the particle positions, combined with real-time running data and coupling parameters, the objective function value is calculated and used as the current fitness value of each particle. Based on the current fitness value of each particle, update the individual optimal solution and determine the population optimal solution; Repeat the steps of dynamically adjusting the inertia weights and adaptive learning factors, and then repeat the subsequent steps. If the maximum number of iterations or the fitness value converges, the iteration stops, and the inertia and damping coefficient corresponding to the population optimal solution are used as the optimal control parameters.

7. The VSG control method for a photovoltaic-storage grid-connected system according to claim 6, characterized in that, The process of updating the particle velocity based on the adjusted inertia weight, adaptive learning factor, individual optimal solution, and swarm optimal solution includes: Based on the adjusted inertial weights and the current particle velocity, the historical velocity components are calculated. The individual optimal guidance component is calculated based on the adaptive learning factor, the difference between the particle's current position and the individual optimal solution; Based on the adjusted adaptive learning factor and the difference between the particle's current position and the swarm optimal solution, the swarm optimal guidance component is calculated. The coupling correction component is calculated based on the coupling gain term, coupling feedback vector, and nonlinear coupling strength. Based on the historical velocity component, the individual optimal guidance component, the group optimal guidance component, and the coupling correction component, the updated velocity of the particle is obtained.

8. The VSG control method for a photovoltaic-storage grid-connected system according to claim 6, characterized in that, The process of updating the particle's position based on the adjusted inertia weight, adaptive learning factor, individual optimal solution, and swarm optimal solution includes: The adaptive time step is determined based on the ratio of the current iteration number to the maximum iteration number and the adjusted inertia weight. Based on the adaptive time step and the updated speed, the basic adjustment amount is calculated; The trend correction amount is calculated based on the difference between the particle's current velocity and its velocity at the previous moment. The updated position of the particle is calculated based on its current position, the base adjustment amount, and the trend correction amount.

9. A VSG control device for a photovoltaic-storage grid-connected system, characterized in that, include: The communication module is used to acquire real-time operating data of the photovoltaic-storage grid-connected system; the real-time operating data includes DC bus voltage, active power, and AC frequency. The processing module is used to perform nonlinear coupling analysis on DC bus voltage, active power and AC frequency based on real-time operating data and a second-order voltage-frequency nonlinear coupling model of DC-AC side, and obtain coupling parameters. Based on the real-time operating data and the coupling parameters, the pre-constructed objective function and control equations are updated. The control equations are a set of voltage-frequency coupling equations characterizing the coupling relationship between the DC side and the AC side. Based on the objective function and the control equations, the optimal control parameters are obtained through iterative optimization. Based on the optimal control parameters, the VSG of the photovoltaic-storage grid-connected system is controlled.

10. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 8.

Images