Flexible polymer virtual inertia frequency modulation control method and electronic device

Abstract

The application provides a flexible polymer virtual inertia frequency modulation control method and electronic equipment, and relates to the technical field of new energy auxiliary frequency modulation. The method comprises the following steps: establishing a heterogeneous energy collaborative frequency modulation system model considering system frequency deviation, operation state and dynamic constraints of each heterogeneous energy unit in the flexible polymer, and linearizing the model into a state space model suitable for model predictive control; constructing a system frequency prediction function for minimizing frequency deviation and control cost, and solving system frequency information at the next moment through model predictive control rolling optimization; according to the solved system frequency information, fusing system frequency change rate and unit state information, and dynamically adjusting the virtual inertia control coefficient of each unit in the flexible polymer; in the frequency recovery stage, taking the pre-generated virtual inertia response stage switching time as the control locking condition, and closing the virtual inertia controller of part or all units. The application can improve the dynamic response and steady-state recovery performance of the power system frequency.

Description

Technical Field

[0001] This application relates to the field of new energy auxiliary frequency modulation technology, and in particular to a method and electronic device for virtual inertia frequency modulation control of flexible polymers. Background Technology

[0002] Driven by energy conservation and emission reduction goals, renewable energy sources, represented by wind power and photovoltaics, are rapidly replacing traditional synchronous generators and gradually becoming the main power source of the power system. However, renewable energy power generation units are usually connected to the grid through power electronic converters, whose inherent "weak inertia" and "weak damping" characteristics have replaced the rotational inertia of traditional synchronous generators, severing the natural connection between system frequency and rotor mechanical kinetic energy. This leads to a sharp decrease in the total equivalent rotational inertia of the power system, and the frequency stability problem becomes increasingly serious. The system's frequency disturbance immunity is weakened, making it more prone to sharp frequency fluctuations and even exceeding limits, posing a serious challenge to the safe and stable operation of the power system. To address the inertia loss problem caused by the high proportion of renewable energy grid connection, Virtual Inertia Control (VIC) technology has emerged. This technology simulates the inertial response of synchronous generators through control algorithms, providing instantaneous power support when the system frequency changes, thereby suppressing the rate of change of frequency (RoCoF).

[0003] Existing research primarily focuses on using the reserved power of Energy Storage Systems (ESS) or photovoltaic (PV) units to achieve virtual inertia control. However, these studies often treat the research object as a single unit, failing to fully explore the synergistic optimization potential of "wind-solar-storage-charging" as an organic whole. In reality, the aggregate contains complex energy flows and constraints: PV output is intermittent and uncertain, electric vehicle charging stations exhibit random fluctuations, and the frequency regulation capability of energy storage units is strictly limited by their State of Charge (SOC). Under extreme operating conditions or continuous disturbances, if traditional fixed-parameter virtual inertia control is used, it can easily lead to the frequency regulation resources (such as wind turbines) releasing energy too quickly, resulting in short support time, and potentially causing secondary frequency drops or overshoot problems during the frequency recovery phase due to reverse power absorption.

[0004] Furthermore, to address the limitations of fixed-parameter control, some scholars have proposed the idea of ​​adaptively adjusting virtual inertia. For example, patent CN113904386B discloses a photovoltaic frequency regulation control parameter optimization method that considers equivalent inertia and damping requirements. This method identifies the equivalent inertia and damping of the photovoltaic grid-connected system using the prediction error method, and optimizes the virtual inertia coefficient and primary frequency regulation coefficient of the photovoltaic power station based on this, thereby improving the system's frequency response characteristics. Although such methods offer some improvement, they are essentially still "local optima" and "post-event responses," lacking the ability to predict the future state of the system and making it difficult to achieve global optimization under multiple constraints. Summary of the Invention

[0005] This application provides a virtual inertia frequency modulation control method and electronic device for flexible polymers, which solves the problem that existing virtual inertia adjustment schemes are difficult to achieve global optimization under multiple constraints. It fully explores the synergistic frequency modulation potential of heterogeneous energy in flexible polymers, achieves accurate prediction and adaptive matching of inertia demand during system disturbances, and avoids overshoot and excessive operation of energy storage devices during the frequency recovery phase.

[0006] In a first aspect, embodiments of this application provide a method for frequency modulation control of virtual inertia in a flexible polymer, comprising: A model of heterogeneous energy cooperative frequency regulation system considering system frequency deviation, operating status of each heterogeneous energy unit in flexible polymer and dynamic constraints is established, and it is linearized into a state-space model suitable for model predictive control. Construct a system frequency prediction function that minimizes frequency deviation and control cost, and solve for the system frequency information at the next moment through model predictive control rolling optimization. Based on the solved system frequency information, the virtual inertia control coefficients of each unit in the flexible polymer body are dynamically adjusted by integrating the system frequency change rate and the unit state information of each heterogeneous energy unit. During the frequency recovery phase, the virtual inertia controller of some or all units is turned off by using the pre-generated virtual inertia response phase switching time as the control lockout condition to avoid frequency overshoot.

[0007] In one possible implementation, establishing a heterogeneous energy coordinated frequency modulation system model that considers system frequency deviation, the operating states of each heterogeneous energy unit within the flexible polymer, and dynamic constraints includes: Wind turbines, photovoltaic units, energy storage units, and electric vehicle charging stations are taken as components of a flexible aggregate. Mathematical models of the virtual rotational inertia and virtual inertial time constant of each unit are established, and the control law for virtual inertia control of each unit through differential control is determined.

[0008] In one possible implementation, the establishment of the heterogeneous energy coordinated frequency regulation system model further includes: The model integrates safety constraints on the rate of change of frequency and establishes a calculation expression for the equivalent virtual inertia time constant of the system, thus characterizing the overall inertia response capability of the flexible polymer as a comprehensive weighted sum of the virtual inertia contributions of its internal units.

[0009] In one possible implementation, linearizing it into a state-space model suitable for model predictive control includes: The system frequency dynamic response model is discretized using the Euler method, and state variables, control inputs and disturbance variables are defined to construct a discrete state-space model. By defining the prediction time domain and the control time domain, the discrete state-space model is extended into a prediction model.

[0010] In one possible implementation, constructing the system frequency prediction function that minimizes frequency deviation and control cost includes: The system frequency prediction function is transformed into a quadratic programming (QP) form using the Lyapunov equation, and a weighted diagonal matrix is ​​set to balance the weights between minimizing frequency deviation and minimizing control cost.

[0011] In one possible implementation, constructing the system frequency prediction function that minimizes frequency deviation and control cost further includes: At least one of the following constraints—active power output of thermal power units, kinetic energy storage and active power output of wind power units, active power output of photovoltaic units, charging and discharging power and state of charge of energy storage units, and charging power of electric vehicle charging stations—is integrated into a QP-form system frequency prediction function.

[0012] In one possible implementation, the step of dynamically adjusting the virtual inertia control coefficients of each unit within the flexible polymer body, based on the solved system frequency information and fusing the system frequency change rate and the unit state information of each heterogeneous energy unit, includes: Based on the magnitude of the system load disturbance and the reassessed minimum inertia requirement of the system, the inertia requirement that each unit in the flexible polymer body needs to provide is determined, and the virtual inertia control coefficient of each unit is solved based on the inertia requirement.

[0013] In one possible implementation, determining the inertia requirement that each unit within the flexible polymer needs to provide includes: When the load disturbance is less than the first disturbance value, only the wind turbine generators are used to provide virtual inertia support; When the load disturbance is greater than or equal to the first disturbance value and less than the second disturbance value, wind turbine units and energy storage units are coordinated to provide virtual inertia support. When the load disturbance is greater than or equal to the second disturbance value, wind turbine units, energy storage units and electric vehicle charging stations are coordinated to provide virtual inertia support.

[0014] In one possible implementation, the dynamic adjustment of the virtual inertia control coefficients of each unit within the flexible polymer further includes: During frequency regulation, the state of charge of the energy storage unit is managed to keep it within the preset safe operating range to avoid overcharging and discharging.

[0015] In one possible implementation, the step of using the pre-generated virtual inertia response stage switching time as a control latching condition to shut down the virtual inertia controller of some or all units includes: Based on the system frequency dynamic response model, the moment when the system frequency change rate is zero is solved by Laplace transform and inverse Laplace transform, and this moment is taken as the switching moment of the virtual inertia response stage; When the system running time is less than the switching time, enable the virtual inertia controller of each unit and tune the parameters according to the solved control coefficients; When the system runtime is greater than or equal to the switching time, the virtual inertia controllers of some or all units are turned off.

[0016] In one possible implementation, when the virtual inertia controller of some or all units is turned off, it further includes: A first-order inertial element is introduced to control the smooth exit of frequency modulation power, avoiding secondary frequency abrupt changes caused by the sudden exit of the controller.

[0017] Secondly, embodiments of this application provide an electronic device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect or any possible implementation of the first aspect.

[0018] In this embodiment, by establishing a heterogeneous energy collaborative frequency regulation system model that fits the actual operating conditions of the system and completing the linearization processing of the adaptive model predictive control, the correlation characteristics between frequency deviation and the operating state of the flexible aggregate can be accurately reflected, providing a reliable foundation for subsequent prediction and control. Then, by using rolling optimization to solve the frequency prediction function to obtain future frequency information, the frequency change trend can be predicted in advance. By dynamically adjusting the virtual inertia control coefficient in combination with the frequency change rate and the unit state, the inertia support can be accurately matched with the system requirements. At the same time, the virtual inertia controller is locked at a preset switching time during the frequency recovery phase, which avoids the frequency overshoot problem from the root and comprehensively improves the frequency dynamic response and steady-state recovery performance of the power system. Attached Figure Description

[0019] Figure 1This is an application scenario diagram of the flexible polymer virtual inertia frequency modulation control method provided in one embodiment of this application; Figure 2 This is a structural diagram of a virtual inertia cooperative support controller provided in an embodiment of this application; Figure 3 This is a flowchart illustrating the implementation of a virtual inertia frequency modulation control method for flexible polymers according to an embodiment of this application. Figure 4 This is a schematic diagram of the frequency response model of a heterogeneous energy collaborative frequency modulation system provided in an embodiment of this application; Figure 5 This is a schematic diagram of the frequency response of a system without additional control according to an embodiment of this application; Figure 6 This is a schematic diagram of the response process of an auxiliary frequency modulation active support system provided in an embodiment of this application; Figure 7 This is a schematic diagram illustrating the relationship between system minimum inertia assessment, frequency change rate limitation, and load disturbance provided in an embodiment of this application; Figure 8 This is a schematic diagram of the system dynamic response simulation results for disturbance scenario 1 in a specific embodiment of this application; Figure 9 This is a schematic diagram of the system dynamic response simulation results in disturbance scenario 2 of a specific embodiment of this application; Figure 10 This is a schematic diagram of the system dynamic response simulation results for disturbance scenario 3 in a specific embodiment of this application. Detailed Implementation

[0020] Model Predictive Control (MPC), as an advanced control method, has significant advantages over traditional control methods. It can handle multivariable and complex constraint optimization problems and can perform forward-looking optimization of future dynamics based on the system model. In recent years, it has been widely used in power system optimization scheduling and control.

[0021] Compared with traditional fixed-parameter control strategies and fixed-inertia control strategies, the method proposed in this application can fully exploit the virtual inertia response potential of flexible aggregates when the frequency regulation power of synchronous generators is insufficient, achieving coordinated assisted frequency regulation of heterogeneous energy sources. By dynamically adjusting control parameters to achieve precise matching of inertia support, it effectively reduces the operating frequency of energy storage units and suppresses overshoot during the frequency recovery phase, significantly improving the dynamic response characteristics and steady-state recovery stability of the system frequency. Furthermore, the method proposed in this application not only provides more precise and rapid power support during disturbances, effectively suppressing the rate of frequency change and reducing steady-state deviation, but also avoids overcharging and discharging by managing the energy storage SOC, thereby maintaining the system's continuous frequency regulation capability.

[0022] To facilitate understanding of the technical solution of this application, the embodiments of this application will be described below in conjunction with the accompanying drawings.

[0023] Figure 1 This diagram illustrates an application scenario of the virtual inertia frequency modulation control method for flexible polymers provided in an embodiment of this application. Figure 1 As shown, the Flexible Aggregator (FA) constructed in this embodiment includes wind turbines, photovoltaic (PV) generators, energy storage units, and electric vehicle charging stations. Under normal operating conditions, the wind turbines and PV generators employ Maximum Power Point Tracking (MPPT) control to maximize renewable energy absorption; the energy storage units monitor and mitigate fluctuations in wind and solar power; and the electric vehicle charging stations charge at a set power.

[0024] To ensure that the heterogeneous energy collaborative frequency modulation system meets the minimum inertia requirement, this application provides a virtual inertia collaborative support controller, the structure of which is as follows: Figure 2 As shown, the system mainly consists of two layers: system inertia requirement assessment and FA virtual inertia control. Layer 1 includes the MPC controller, a system minimum inertia requirement assessment module, and inertia requirement assessment modules for each FA unit. Layer 2 receives the requirement assessment results from Layer 1 and calculates the virtual inertia control coefficients for each FA unit. Without additional control, both wind power and photovoltaic systems use MPPT control, and the energy storage unit monitors wind turbine power fluctuations to smooth them out. When the system experiences disturbances, the frequency signal is introduced into the FA controller to generate dynamic power compensation commands, adjusting the active power of each unit to achieve rapid inertia support. When the system encounters a positive load disturbance, except for the synchronous motor, the FA frequency regulation output priority is, in order, wind turbine, energy storage unit, and electric vehicle charging station; while when the system encounters a negative disturbance, the priority is, in order, wind turbine, energy storage unit, and photovoltaic unit.

[0025] FA virtual inertia control uses the virtual inertia response stage switching time. As a control latch-up condition, it is used to avoid frequency overshoot caused by the effect of virtual inertia during the frequency recovery phase. When, the system calculates and tunes the parameters; when At this time, the virtual inertia controller of some units in the FA is turned off and completely exits after the inertia response ends. The virtual inertia control of the FA uses the system frequency as the input signal and realizes the inertia response through frequency detection and differentiation. In order to ensure the smooth exit of the supporting power of some units in the FA, a first-order inertial element is introduced into the controller to avoid secondary frequency abrupt changes caused by the sudden exit of the controller.

[0026] When the system experiences load disturbances (such as sudden increases or decreases in load), the system frequency changes. The flexible aggregate virtual inertia controller receives the frequency signal from the power grid as input. Internally, the controller first uses the MPC module to perform rolling optimization prediction of the frequency change trend and outputs the predicted frequency deviation for the next moment. Simultaneously, the system inertia demand assessment module calculates the minimum inertia support required by the system in real time based on the frequency change rate constraint and the predicted frequency deviation. Subsequently, the inertia demand allocation module decomposes this inertia deficit into each unit of the flexible aggregate, solves for the virtual inertia control coefficient of each unit, and generates its respective power response command.

[0027] Power response commands are executed by the converters of each unit to achieve rapid power injection or absorption. Through the aforementioned hierarchical and adaptive energy flow and control signal interaction, the flexible aggregate can quickly fill in the gaps in the early stages of system disturbances, share the frequency regulation pressure of the synchronous generator, and orderly exit during the frequency recovery phase according to the calculated virtual inertia response switching time, thereby avoiding frequency overshoot.

[0028] The frequency modulation contribution weight of each unit is affected by, for example Figure 1 The energy flow architecture constraints shown are dynamically allocated through priority logic. When the system encounters a positive load disturbance, the flexible aggregate provides virtual inertia support based on the synchronous unit response, prioritizing wind power, energy storage, and electric vehicle charging stations. Conversely, when the system encounters a negative load disturbance, the frequency regulation output priority dynamically switches to wind power, energy storage, and photovoltaic units. This mechanism optimizes the equivalent virtual inertia characteristic distribution of the flexible aggregate by adjusting the priority sequence under different disturbance conditions in real time, thereby improving the flexibility of multi-source coordinated frequency regulation.

[0029] The above provides a general description of the scenarios and overall process of the embodiments of this application. The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0030] See Figure 3 The diagram illustrates the implementation flowchart of the virtual inertia frequency modulation control method for flexible polymers provided in this application embodiment, including the following steps: S301. Establish a heterogeneous energy coordinated frequency regulation system model that considers the system frequency deviation, the operating state of each heterogeneous energy unit in the flexible polymer, and dynamic constraints, and linearize it into a state-space model suitable for model predictive control.

[0031] The execution subject of each embodiment of this application can be a server, processor, microprocessor, or other device with data processing capabilities. In actual implementation, the specific implementation method of the execution subject can be selected according to actual needs. This embodiment does not impose any special restrictions on this, as long as it is a device with data processing capabilities.

[0032] In one possible implementation, establishing a heterogeneous energy coordinated frequency modulation system model that considers system frequency deviation, the operating states of each heterogeneous energy unit within the flexible polymer, and dynamic constraints includes: Wind turbines, photovoltaic units, energy storage units, and electric vehicle charging stations are taken as components of a flexible aggregate. Mathematical models of the virtual rotational inertia and virtual inertial time constant of each unit are established, and the control law for virtual inertia control of each unit through differential control is determined.

[0033] Specifically, for wind turbines, their virtual inertia originates from the kinetic energy stored in the rotor's rotation. When the system frequency changes, wind turbines can simulate the inertial response of a synchronous generator by releasing or absorbing rotor kinetic energy. Their virtual moment of inertia is related to rotor speed, rated speed, and mechanical parameters, while the virtual inertia time constant characterizes their ability to provide power support using kinetic energy for a duration. For energy storage units, their virtual inertia originates from the electrical energy stored in the battery. The rapid power throughput characteristics of the power electronic converter simulate rotational inertia. Their virtual moment of inertia is related to the terminal voltage, rated capacity, and the rate of change of SOC (State of Charge). The virtual inertia time constant reflects their ability to sustainably provide inertial support at the current SOC level. For photovoltaic (PV) generators and electric vehicle (EV) charging stations, although they lack rotating parts, inertial response can be simulated by reserving power or adjusting charging power. The virtual inertia of PV generators is related to the voltage, current, and energy storage capacity of their PV array; the virtual inertia of EV charging stations is related to the range of charging power adjustment and battery status. Each of the aforementioned units achieves virtual inertia control through a differential control loop, meaning that the increment of the frequency modulation power output by each unit is proportional to the derivative of the system frequency deviation. This control law enables each unit to act rapidly during instantaneous frequency changes, providing inertial support similar to that of a synchronous machine rotor, effectively suppressing the initial rate of frequency change.

[0034] Optionally, the establishment of the heterogeneous energy coordinated frequency regulation system model further includes: The model integrates safety constraints on the rate of change of frequency and establishes a calculation expression for the equivalent virtual inertia time constant of the system, thus characterizing the overall inertia response capability of the flexible polymer as a comprehensive weighted sum of the virtual inertia contributions of its internal units.

[0035] Specifically, the rate of frequency change is a key indicator for measuring the frequency stability of a system. An excessively high rate of frequency change may trigger protective devices or even lead to system disconnection. Therefore, this embodiment integrates a safety constraint on the rate of frequency change into the model, treating it as a boundary condition that the system must meet for operation. By setting the maximum permissible rate of frequency change, the minimum inertia level required by the system can be derived, thus providing a safety benchmark for subsequent optimized control. To quantify the overall inertia support capacity of the flexible aggregate, this embodiment establishes a calculation expression for the system's equivalent virtual inertia time constant. This expression is not a simple arithmetic average, but rather uses a weighted summation method, combining the virtual inertia time constants of each unit according to their capacity proportion or frequency regulation contribution weight. This weighted summation method accurately reflects the actual contribution of different capacities and types of energy units to the overall system inertia. For example, in systems with a high proportion of large-capacity wind power, the virtual inertia contribution weight of wind power is larger, and its effect on improving the system's equivalent inertia is more significant. This expression allows for the unified mapping of dispersed, heterogeneous unit inertia capabilities to the overall system inertia index, providing an intuitive quantitative basis for the MPC controller to assess the system's inertia deficit and formulate frequency regulation strategies. This design not only achieves unified modeling of heterogeneous energy units but also ensures the safety and effectiveness of the frequency regulation process through the integration of safety constraints and the quantification of equivalent inertia.

[0036] This application uses a provincial regional power grid as a typical scenario for simulation analysis. The system has a rated power of 800MW for thermal power units, 200MW for wind power units, 100MW for photovoltaic units, 20MW / 10MWh for energy storage units, and 20MW for electric vehicle charging stations. The total user-side load is set at 1020MW (including the initial 20MW charging power of the electric vehicle charging stations). The total user-side load is less than the grid-side power of the power system to ensure that the grid side has sufficient frequency regulation resources. Other parameters of the simulation system are detailed in Table 1. The system is set in... A load disturbance occurs at 2s. The frequency response characteristics and virtual inertia of each element of the flexible polymer are analyzed as follows: Table 1. Installed capacity and unit parameters of a power grid in a certain region

[0037] First, we analyze the frequency response characteristics and virtual inertia of wind turbine units.

[0038] When the rotor speed changes, the rotational kinetic energy stored in the rotor of the wind turbine unit for: (1) In the formula: This indicates the mechanical angular velocity of the synchronous generator unit; This represents the inherent rotational inertia of the wind turbine. This represents the rotor angular velocity of the wind turbine. Referring to the inertial response of a synchronous generator, the virtual moment of inertia of the wind turbine is defined. for: (2) In the formula: This indicates the change in the speed of the synchronous generator unit. This indicates its initial angular velocity; This indicates the change in the wind turbine's rotational speed. This indicates its initial rotational speed.

[0039] The virtual inertial time constant of the wind turbine can be calculated from equation (2). The expression is: (3) In the formula: Indicates the rated capacity of the wind turbine unit; This indicates the energy reserves of the wind turbine.

[0040] Secondly, the frequency response characteristics and virtual inertia of the energy storage unit are analyzed.

[0041] According to the definition of state of charge, lead-acid batteries The SOC parameter at time t can be expressed as: (4) In the formula: Indicates that the battery is in The state of charge at any given moment; Indicates that the battery is in The discharge current at any given moment; This indicates the rated capacity of the battery when fully charged; Indicates that the battery is in The remaining capacity at any given time.

[0042] Combining equation (4), and referring to the definition of the moment of inertia of a synchronous generator, the energy stored in the battery... It can be represented as: (5) In the formula: This indicates the rated voltage of the battery; This indicates the discharge current of the battery; This indicates the initial state of charge of the battery.

[0043] According to the definition of the rotational kinetic energy of a synchronous motor, the energy stored in a battery can be further expressed as: (6) In the formula: This represents the virtual moment of inertia of the battery.

[0044] Combining equation (5), and further expanding equation (6), the virtual rotational inertia of the battery during the system inertia response stage can be expressed as: (7) In the formula: This represents the inherent moment of inertia of a synchronous generator; This represents the ratio of the rate of change of the battery's state of charge to the rate of change of the synchronous generator's speed. ; This represents the rotor kinetic energy of a synchronous generator.

[0045] The virtual inertial time constant of the energy storage unit is obtained from equation (7). for: (8) In the formula: Indicates the rated capacity of the battery; This indicates the energy reserves in the battery.

[0046] Then, the frequency response characteristics and virtual inertia of electric vehicle charging stations are analyzed.

[0047] Power variation at electric vehicle charging stations The relationship between the voltage and current of the charging station is expressed as follows: (9) In the formula: This indicates the voltage change at the charging station; This indicates the change in current at the charging station.

[0048] Referring to the definition of the moment of inertia of a synchronous generator, the energy stored in an electric vehicle charging station It can be represented as: (10) Referring to the energy storage unit, and combining with equation (10), it can be seen that, during the system inertia response stage, the virtual rotational inertia of the electric vehicle charging station is defined. for: (11) The virtual inertial time constant of the electric vehicle charging station can be obtained from equation (11). The expression is: (12) In the formula: This indicates the rated capacity of the electric vehicle charging station.

[0049] Next, the frequency response characteristics and virtual inertia of the photovoltaic unit are analyzed.

[0050] Power variation of photovoltaic units The relationship between voltage and current is expressed as follows: (13) In the formula: This indicates the voltage change of the photovoltaic unit; This indicates the change in current of the photovoltaic unit.

[0051] Referring to the definition of the moment of inertia of a synchronous generator, the energy stored by a photovoltaic unit It can be represented as: (14) Referring to the energy storage unit, and combining with equation (14), it can be seen that, in the system inertia response stage, the virtual rotational inertia of the photovoltaic unit is defined. for: (15) From equation (15), the virtual inertial time constant of the photovoltaic unit can be obtained. The expression is: (16) In the formula: Indicates the rated capacity of the photovoltaic unit; Indicates the energy reserves of the photovoltaic unit; This represents the inherent moment of inertia of a synchronous generator with the same capacity as a photovoltaic unit. This represents the rotor kinetic energy of a synchronous generator with the same capacity as a photovoltaic unit.

[0052] To enable the flexible polymer to rapidly support frequency changes in the early stages of disturbance, all its internal units implement virtual inertia control through differential control elements, namely: (17) In the formula: This represents the power response signal of each unit in the polymer; This indicates that energy storage absorbs power from the grid. Indicates the released power; This indicates a decrease in photovoltaic power, which is equivalent to a decrease in released power, or a reduction in output during the frequency rise phase. This indicates that the power absorbed by the electric vehicle charging station decreases, which is equivalent to increasing the output during the frequency decrease phase. This represents the virtual inertia control coefficient of each unit in the aggregate.

[0053] Finally, the equivalent virtual inertia characteristics are analyzed.

[0054] From a system equivalence perspective, the overall inertial response capability of this flexible aggregate can be characterized as the comprehensive weighted sum of the virtual inertial contributions of its internal units, and its system equivalent virtual inertial time constant. for: (18) In the formula: Indicates the inertia sensitivity factor of each unit in the polymer ( ); This indicates its currently adjustable energy; This represents the inertial time constant of a thermal power unit; This indicates the adjustable kinetic energy of the thermal power unit at the current operating point; This indicates the system's baseline capacity.

[0055] In this embodiment, wind power, photovoltaic energy storage, and electric vehicle charging stations are treated as flexible aggregate units, and virtual inertia-related mathematical models are constructed for each unit. The rules for realizing virtual inertia control by differential control are clarified, and the inertia contribution and control logic of each unit can be clearly defined. This provides accurate unit-level theoretical support for heterogeneous energy collaborative frequency regulation and ensures that the virtual inertia output of each unit meets the overall frequency regulation requirements of the system.

[0056] In one possible implementation, linearizing it into a state-space model suitable for model predictive control includes: The system frequency dynamic response model is discretized using the Euler method, and state variables, control inputs and disturbance variables are defined to construct a discrete state-space model. By defining the prediction time domain and the control time domain, the discrete state-space model is extended into a prediction model.

[0057] Specifically, the core of Model Predictive Control (MPC) lies in "prediction" and "rolling optimization." To predict future frequency trends, a prediction time domain P and a control time domain M need to be defined. The prediction time domain P refers to the number of time steps the controller predicts forward, determining how far the controller can "see." The control time domain M refers to the number of time steps the controller optimizes the control sequence; typically, M is less than or equal to P. Outside the control time domain M, the control variables are usually assumed to remain constant or be zero. By using the defined prediction time domain P and control time domain M, and recursively applying the aforementioned discrete state-space equations, the state equation at a single moment can be extended into a predictive model. This predictive model can predict the system frequency deviation sequence at P future moments based on the current system state. This extended predictive model forms the basis for subsequent rolling optimization solutions by the MPC controller, enabling the controller to make decisions in advance based on future frequency trends, rather than simply reacting passively to the current state. In this way, this embodiment successfully transforms a complex continuous nonlinear system model into a standard form suitable for the MPC algorithm, laying a solid foundation for subsequent optimization solutions.

[0058] For reference Figure 4 The heterogeneous energy coordinated frequency regulation system model shown is based on the power balance principle. It characterizes the dynamic characteristics of the system frequency evolution over time by algebraically summing the output responses of each heterogeneous energy unit within the system and external disturbances. Specifically, the total output increment, composed of steam turbines, wind power, photovoltaics, energy storage, and electric vehicle charging stations, after offsetting load disturbances and its own damping, forms the system's net power deviation. This deviation further acts on the equivalent generator transfer function model with rotational inertia, thereby driving a dynamic shift in the system frequency and establishing a dynamic mapping relationship from multi-source power input at the source end to the frequency response at the end. Based on the above physical logic, the mathematical model of the frequency dynamic response under load disturbances is derived as follows: (19) In the formula: Indicates frequency deviation; Represents the system's inertial time constant; This represents the power response signal of the synchronous generator; Indicates load disturbance; This represents the damping coefficient. Wherein, It depends on the droop coefficient of the synchronous generator and the turbine parameters: (20) System frequency response characteristics under different wind power penetration conditions, such as Figure 5 As shown, without additional control of the wind turbine, increasing the proportion of wind power accelerates the system frequency descent rate, increases the frequency descent amplitude, and increases the steady-state frequency deviation, but the system inertial response time does not show significant changes in observation. Figure 6 It can be seen that during the response process of the assisted frequency modulation active support system, when the system frequency reaches its lowest point, the rate of change of the system frequency is zero, and the system inertial response stage time can be calculated. .

[0059] Substituting equations (17) and (20) into equation (19) and performing a Laplace transform, we obtain: (twenty one) in (twenty two) Performing an inverse Laplace transform on equation (21) yields the time-domain expression for the frequency response: (twenty three) in (twenty four) Substituting the initial conditions and differentiating equation (23), we can obtain the switching time of the virtual inertia response stage of the flexible polymer. for: (25) In the initial stage of the disturbance, the synchronous generator speed governor and load frequency regulation have not yet been activated. The system frequency change rate mainly depends on the inertial response of the synchronous generator and the equivalent virtual inertia provided by the flexible aggregate. Therefore, the frequency change rate is the largest at this time. From the frequency dynamic response mathematical model (17), we can obtain: (26) To ensure the safe and stable operation of the new energy system, this application sets a constraint limit for the maximum frequency change rate, referencing relevant power grid standards. This constraint value can be adaptively adjusted according to the actual system conditions to ensure safe operation; therefore, the minimum inertia based on system safety constraints is: (27) In the formula: This represents the system's minimum inertia requirement; This indicates the system's rated frequency.

[0060] Equation (27) achieves a quantitative assessment of the minimum equivalent inertia requirement of the power system by constructing an analytical model that takes into account the rate of frequency change and frequency security constraints. Its mapping relationship with the frequency security threshold and load disturbance is as follows: Figure 7 As shown. Among them, Figure 7 Figures (a) and (b) respectively depict the evolution of the system's inertia demand under positive and negative load disturbance conditions.

[0061] Based on the fundamental principles of MPC, the Euler method is used to discretize the frequency response model expression (19) of the heterogeneous energy coordinated frequency modulation system. The system discretization simulation step size is... We can obtain: (28) make , , , Then the discretized model simplifies to: (29) make The discretized model then simplifies further to: (30) Define state variables Control input Disturbance variables The simplified discretized model is then constructed into a state-space model: (31) Let the prediction time domain be Control time domain is ( ), extending the state-space model into a prediction model: (32) In the formula: , , .

[0062] In this embodiment, the Euler method is used to discretize the frequency dynamic response model and define relevant variables to construct a discrete state-space model. Then, by combining the prediction time domain and the control time domain, it is extended into a prediction model. This enables the model predictive control algorithm to efficiently adapt to the dynamic characteristics of the system and stably achieve accurate prediction of future frequency states, providing a standardized and feasible model carrier for rolling optimization control.

[0063] S302, construct a system frequency prediction function that minimizes frequency deviation and control cost, and solve for the system frequency information at the next moment through model predictive control rolling optimization.

[0064] In one possible implementation, constructing the system frequency prediction function that minimizes frequency deviation and control cost includes: The system frequency prediction function is transformed into QP form using Lyapunov equations, and a weighted diagonal matrix is ​​set to balance the weights between minimizing frequency deviation and minimizing control cost.

[0065] In practice, MPC minimizes frequency deviation and control cost by optimizing control inputs in the future time domain: (33) The system frequency prediction function (33) is transformed into QP form using the Lyapunov equation: (34) In the formula: and This represents the weight diagonal matrix. , , and Both represent the weighting coefficients (constants) of the single-step frequency deviation. and They represent peacekeeping An identity matrix of dimension 1.

[0066] The goal of Model Predictive Control (MMCC) is to find an optimal set of control input sequences in the prediction time domain that minimizes system frequency deviation while keeping control costs (such as the magnitude and frequency of regulation power) as low as possible. To this end, a two-part objective function is constructed: the first part is the weighted sum of squares of system frequency deviation in the prediction time domain, reflecting the system's frequency stability; the second part is the weighted sum of squares of control increments in the control time domain, reflecting the control's adjustment strength and economic cost. For ease of computer solution, this embodiment transforms the objective function into the standard QP form using Lyapunov equations. It should be understood that QP problems have mature solution algorithms (such as the interior-point method), offering high computational efficiency and meeting the needs of real-time power system control. During the transformation, weighted diagonal matrices Q and R are introduced. Matrix Q corresponds to the weight of the frequency deviation term; a larger value indicates a higher requirement for frequency recovery accuracy and a faster frequency recovery speed, but may require greater regulation power. Matrix R corresponds to the weight of the control cost term; a larger value indicates a stronger constraint on regulation power, effectively suppressing drastic fluctuations in control quantities and protecting equipment lifespan. By adjusting the values ​​of Q and R, the contradiction between frequency recovery speed and frequency modulation resource consumption can be flexibly balanced, achieving multi-objective optimization control.

[0067] In this embodiment, the frequency prediction function is transformed into QP form using the Lyapunov equation and a weighted diagonal matrix is ​​configured. This effectively balances the dual objectives of frequency deviation suppression and control cost reduction, allowing the frequency control effect to achieve optimal matching with system operating costs, thereby improving the economy and effectiveness of the overall control strategy.

[0068] In one possible implementation, constructing the system frequency prediction function that minimizes frequency deviation and control cost further includes: At least one of the following constraints—active power output of thermal power units, kinetic energy reserve and active power output of wind power units, active power output of photovoltaic units, charging and discharging power and state of charge of energy storage units, and charging power of electric vehicle charging stations—is integrated into the system frequency prediction function of the QP form.

[0069] The following constraints apply during implementation: (1) Active power output constraints of thermal power units (35) In the formula: Indicates that thermal power units are in Contributing effort at all times; This indicates the minimum active power output after being limited by factors such as boiler combustion stability and turbine vibration. This indicates the maximum output of the nameplate or the maximum output that can be achieved at present.

[0070] (2) Energy reserve constraints of wind turbine units (36) In the formula: The fan speed at the moment the inertial response begins; This indicates the minimum permissible speed for stable operation of the fan; This indicates the maximum permissible speed of the fan during stable operation.

[0071] (3) Active power output constraints of wind turbine units (37) In the formula: Indicates wind turbine units at Contributing effort at all times; This indicates the maximum output that the fan can achieve at its highest permissible speed; This indicates the minimum output that the fan can achieve at the lowest permissible speed.

[0072] (4) Active power output constraints of photovoltaic units (38) In the formula: Indicates that the photovoltaic unit is in Contributing effort at all times; This represents the minimum active power output of the photovoltaic system, determined by the minimum operating power of the inverter, and is typically taken as 0.05. Or 0.1 (To avoid frequent start-stop of the inverter); This indicates the MPPT power corresponding to photovoltaic power.

[0073] (5) Charge and discharge power constraints of energy storage units (39) In the formula: Indicates energy storage Contributing effort at all times; , These represent the upper and lower limits of charging power, respectively. , These represent the upper and lower limits of the discharge power, respectively.

[0074] (6) SOC constraints of energy storage units (40) In the formula: Take 0.2, Taking 0.8, the premise for energy storage to participate in frequency regulation is that the SOC is within a safe range.

[0075] (7) Charging power constraints of electric vehicle charging stations (41) In the formula: Indicates electric vehicle charging station The charging power at any given moment; , These represent the upper and lower limits of charging power, respectively.

[0076] Combining the above constraints, substituting the state-space model into the system frequency prediction function yields the QP problem: (42) In the formula: , .

[0077] The optimal control sequence is obtained using a QP solver (such as the interior-point method). Only the first step is executed. The next time step will be re-optimized. The predicted frequency trajectory is: (43) In this embodiment, the operational constraints of various units and equipment are integrated into the QP prediction function, which ensures that the control optimization process always follows the physical operating limits and safety boundaries of the equipment, avoids equipment operating beyond its limits, ensures that each unit of the flexible polymer participates in frequency modulation under the premise of safety and stability, and improves the practicality and reliability of the control strategy.

[0078] S303, based on the solved system frequency information, integrates the system frequency change rate and the unit state information of each heterogeneous energy unit, and dynamically adjusts the virtual inertia control coefficient of each unit in the flexible polymer body.

[0079] This step embodies the key implementation logic of the "adaptive" characteristic of this invention. Traditional virtual inertia control often uses fixed parameters, which cannot adapt to the system requirements under different disturbance levels. This step dynamically calculates and allocates the virtual inertia control coefficients of each unit based on the frequency information predicted by MPC, combined with real-time monitoring of the system frequency change rate and the real-time status of each unit (such as energy storage SOC, wind turbine rotor speed, etc.). For example, when a small frequency drop is predicted, only the control coefficient of the wind turbine needs to be adjusted to meet the support requirements; when a large frequency drop is predicted, the control coefficients of the energy storage unit and even the electric vehicle charging station need to be adjusted in a coordinated manner. This dynamic adjustment mechanism realizes the "on-demand allocation" of frequency regulation resources, which not only ensures the accuracy of frequency support, but also avoids ineffective actions of units such as energy storage, and extends the service life of the equipment.

[0080] Specifically, after the MPC controller obtains the system frequency information for the next moment through rolling optimization, it needs to dynamically adjust the virtual inertia control coefficients of each unit based on this information. This adjustment process consists of two key steps: In one possible implementation, the step of dynamically adjusting the virtual inertia control coefficients of each unit within the flexible polymer body based on the obtained system frequency information, integrating the system frequency change rate and the unit state information of each heterogeneous energy unit, includes: Based on the magnitude of the system load disturbance and the reassessed minimum inertia requirement of the system, the inertia requirement that each unit in the flexible polymer body needs to provide is determined, and the virtual inertia control coefficient of each unit is solved based on the inertia requirement.

[0081] The determination of the inertia requirements to be provided by each unit within the flexible polymer includes: When the load disturbance is less than the first disturbance value, only the wind turbine generators are used to provide virtual inertia support; When the load disturbance is greater than or equal to the first disturbance value and less than the second disturbance value, wind turbine units and energy storage units are coordinated to provide virtual inertia support. When the load disturbance is greater than or equal to the second disturbance value, wind turbine units, energy storage units and electric vehicle charging stations are coordinated to provide virtual inertia support.

[0082] This embodiment employs a tiered allocation strategy, dynamically selecting the combination of frequency regulation resources based on the degree of load disturbance. The core idea of ​​this strategy is "on-demand allocation," which means precisely matching frequency regulation resources according to the actual inertia requirements of the system, avoiding ineffective actions of energy storage and other units, and extending the service life of equipment.

[0083] A first disturbance value and a second disturbance value are set, where the second disturbance value is greater than the first disturbance value. When the load disturbance is of the first level (i.e., less than the first disturbance value, a small disturbance), the system's minimum inertia requirement is low, and the system requirements can be met solely through the virtual inertia control of the wind turbine units. When the load disturbance is of the second level (i.e., greater than or equal to the first disturbance value and less than the second disturbance value, a medium disturbance), the system's minimum inertia requirement increases, and the virtual inertia control of the wind turbine units alone is insufficient to meet the system requirements, necessitating the collaborative provision of virtual inertia support by energy storage units. When the load disturbance is of the third level (i.e., greater than or equal to the second disturbance value, a large disturbance), the system's minimum inertia requirement further increases, and even with both wind turbine units and energy storage units operating at full power, the system requirements still cannot be met, requiring the collaborative provision of virtual inertia support by electric vehicle charging stations.

[0084] In this embodiment, flexible aggregate units are deployed in a tiered manner according to the degree of load disturbance to provide inertia support. This approach can meet the system's inertia requirements while avoiding redundant resource investment, achieving precise matching of frequency regulation resources: only wind power is deployed in low-disturbance scenarios, wind power and energy storage are coordinated in medium-disturbance scenarios, and multiple units work together in high-disturbance scenarios. This strategy not only ensures system frequency security but also effectively avoids ineffective operations of energy storage and electric vehicle charging stations, reduces the frequency of energy storage charging and discharging, and helps extend equipment lifespan.

[0085] Furthermore, the method for dynamically adjusting the virtual inertia control coefficients of each unit within the flexible polymer also includes: During frequency regulation, the state of charge of the energy storage unit is managed to keep it within the preset safe operating range to avoid overcharging and discharging.

[0086] Specifically, the frequency regulation capability of energy storage units is strictly limited by their State of Charge (SOC). If the SOC is too high (close to 1), the energy storage unit cannot continue charging and cannot absorb power during the frequency rise phase; if the SOC is too low (close to 0), the energy storage unit cannot continue discharging and cannot release power during the frequency fall phase. Therefore, the SOC of the energy storage unit must be managed in real time during frequency regulation to keep it within a safe operating range. Optionally, the safe operating range is set to 0.2 to 0.8.

[0087] In this embodiment, the State of Charge (SOC) of the energy storage unit is used as a key constraint when dynamically adjusting the virtual inertia control coefficient. When the SOC of the energy storage unit approaches the upper or lower limit of the safe range, the controller automatically reduces the virtual inertia control coefficient of the energy storage unit, decreasing its frequency regulation output and avoiding overcharging and discharging. Simultaneously, the controller correspondingly increases the virtual inertia control coefficients of other units (such as wind turbines or electric vehicle charging stations) to compensate for the decrease in the frequency regulation capability of the energy storage unit. This SOC management mechanism ensures that the energy storage unit can continuously and stably participate in system frequency regulation, guaranteeing the system's continuous frequency regulation capability.

[0088] In the specific implementation process, taking the frequency drop process caused by a sudden increase in load as an example, three points are randomly selected for illustration, such as... Figure 7 As shown in (a): (1) When the load disturbance is small =0.07(pu), the maximum frequency change rate is relatively small. =0.35Hz / s (i.e.) Figure 7Point 1 in (a) shows that the minimum inertia requirement of the system is calculated to be 5.4s, while the equivalent inertia of the synchronous unit is only 4s, indicating an inertia gap. In this case, virtual inertia control needs to be added to the wind turbine to provide inertia support. If the virtual inertia time constant of the wind turbine is set to 10s, the equivalent inertia time constant of the system can be increased to 5.454s. At this point, energy storage can meet the minimum inertia requirement of the system without additional control.

[0089] (2) When the load disturbance is large =0.12(pu), the maximum frequency change rate is relatively small. =0.4Hz / s (i.e. Figure 7 Point 2 in (a) shows the calculated minimum inertia requirement of the system, which is 7.5s. If the maximum value of the virtual inertia time constant of the wind turbine is 15s, then the equivalent inertia time constant of the system is only 6.363s. That is, even if the virtual inertia time constant of the wind turbine reaches its maximum value, it still cannot meet the minimum inertia requirement of the system. At this time, the energy storage needs to supplement the virtual inertia to meet the minimum inertia requirement of the system.

[0090] (3) When the load disturbance is large =0.21(pu), the maximum frequency change rate is relatively large. =0.5Hz / s (i.e.) Figure 7 Point 3 in (a) shows the calculated minimum inertia requirement of the system, which is 10.5 s. If the maximum virtual inertia time constant of the wind turbine is 15 s and the maximum virtual inertia time constant of the energy storage is 24 s, then the equivalent inertia time constant of the system is only 6.799 s. That is, both the virtual inertia time constants of the wind turbine and the energy storage have reached their maximum values, which still cannot meet the minimum inertia requirement of the system. At this time, the electric vehicle charging station needs to supplement the virtual inertia to meet the minimum inertia requirement of the system.

[0091] During the frequency response period, the rotor speed of a synchronous generator typically ranges from 0.96 (pu). 1(pu), corresponding to the maximum synchronous speed deviation =0.04 (pu). The operating range for wind turbine rotor speeds is generally set at 0.7 (pu). 1.2(pu), substituting the above boundary conditions into equation (3), the maximum virtual inertial time constant that the wind turbine can provide within this operating range can be derived as: (44) In the formula: This represents the inherent inertial time constant of the wind turbine.

[0092] Similarly, when the safe range of the state of charge (SOC) of the energy storage unit is 0.2... At 0.8, the maximum virtual inertial time constant that the energy storage unit can provide within this operating range can be derived from equation (8) as follows: (45) according to Figure 7 The results show that, under low-power disturbances, the system can meet its inertia requirements during the inertia response phase solely through virtual inertia control of the wind turbine and coordination with the synchronous generator. At this point, the system's total equivalent inertia time constant is... for: (46) According to equation (27), at the maximum frequency change rate With power disturbance Under the given conditions, in order to meet the minimum inertia requirement constraint, the system's equivalent inertia time constant should satisfy the following: Combining equation (46), when the synchronous turbine and the wind turbine jointly undertake the task of inertia support, the minimum virtual inertia that the wind turbine needs to provide is: (47) As shown in equation (27), with the increase of load disturbance, the system's ability to resist disturbance decreases. At this time, the aggregate needs to provide corresponding inertia support according to the system requirements. If the maximum virtual inertia that the wind turbine can provide is lower than its minimum requirement, the energy storage unit needs to provide virtual inertia support. In this case, the total equivalent inertia time constant of the system is... for: (48) Combining equation (48) and the system's minimum inertia requirement constraint, it can be seen that the virtual inertia required by the energy storage unit is: (49) However, as load disturbances increase again, the system's ability to resist disturbances decreases, at which point the flexible aggregate needs to provide further inertial support. If the maximum virtual inertia that the wind turbine and energy storage units can provide is lower than their minimum required value, then the electric vehicle charging station needs to collaboratively provide virtual inertial support. In this case, the system's total equivalent inertial time constant... for: (50) Combining equation (50) and the system's minimum inertia requirement constraint, the virtual inertia that the electric vehicle charging station needs to provide is: (51) In this embodiment, the inertia requirements of each unit are determined based on the load disturbance level and the minimum inertia requirement of the system, and the corresponding control coefficients are solved. This enables the on-demand allocation and precise control of frequency modulation resources, allowing the virtual inertia support strength to fully match the real-time requirements of the system, thereby maximizing the improvement of frequency support efficiency and system stability.

[0093] S304, during the frequency recovery phase, uses the pre-generated virtual inertia response phase switching time as the control interlocking condition to shut down the virtual inertia controllers of some or all units to avoid frequency overshoot.

[0094] like Figure 2 As shown, the virtual inertia control module of the flexible polymer uses equation (25) to calculate the virtual inertia response stage switching time. This serves as a control interlock condition to prevent frequency overshoot caused by virtual inertia during the frequency recovery phase. The control module includes a decision-making component. ,when At that time, the system adjusts the parameters based on the calculation results of equations (47), (49) and (51). Otherwise, the virtual inertia controllers of some units in the flexible polymer will be shut down and will completely exit after the inertia response ends. The virtual inertia control of the flexible polymer uses the system frequency as the input signal and realizes the inertia response through frequency detection and differentiation.

[0095] This step is a safety measure for the frequency recovery process. During the recovery phase after the frequency drops to its lowest point, if the virtual inertia controller continues to operate, it may inject excessive power into the system, causing the frequency to rise in the opposite direction and exceed the rated value, resulting in frequency overshoot or even secondary oscillation. This step pre-calculates a switching moment of the virtual inertia response phase (usually corresponding to the moment when the frequency change rate is zero) and uses it as the trigger condition for control interlocking. When the system runtime reaches this switching moment, the controller automatically shuts down the virtual inertia control function of some or all units, stopping inertia support. This interlocking mechanism effectively prevents over-adjustment during the frequency recovery phase, ensuring a smooth return of the frequency to the rated value and significantly improving the stability of the system's frequency recovery.

[0096] In one possible implementation, the step of using the pre-generated virtual inertia response stage switching time as a control latching condition to shut down the virtual inertia controller of some or all units includes: Based on the system frequency dynamic response model, the moment when the system frequency change rate is zero is solved by Laplace transform and inverse Laplace transform, and this moment is taken as the switching moment of the virtual inertia response stage; When the system running time is less than the switching time, enable the virtual inertia controller of each unit and tune the parameters according to the solved control coefficients; When the system runtime is greater than or equal to the switching time, the virtual inertia controllers of some or all units are turned off.

[0097] Specifically, during the recovery phase after the frequency drops to its lowest point, if the virtual inertia controller continues to operate, it may inject excessive power into the system, causing the frequency to rise in the opposite direction and exceed the rated value, resulting in frequency overshoot or even secondary oscillation. To avoid this problem, this embodiment pre-calculates the switching time of the virtual inertia response phase and uses it as the trigger condition for control latch-up.

[0098] The calculation of the switching time is based on the system frequency dynamic response model. As described above, this embodiment first establishes the transfer function model of the system frequency dynamic response, and then transforms it into a frequency domain expression through Laplace transform. In the frequency domain, the system frequency response can be represented as the superposition of load disturbance, synchronous unit response, and flexible aggregate response. By performing an inverse Laplace transform on this frequency domain expression, the time domain expression of the system frequency response can be obtained. Differentiating the time domain expression yields the time domain expression of the system frequency change rate. Setting the frequency change rate to zero, the moment when the system frequency reaches its lowest point can be solved, which is the switching time of the virtual inertia response stage. .

[0099] When the system runtime t is less than the switching time At this time, the system is in a frequency drop phase, requiring power support from the virtual inertia controller. The controller enables the virtual inertia controllers of each unit and sets the tuning parameters according to the control coefficients obtained in the aforementioned embodiment. Each unit outputs frequency-modulated power according to the differential control law to suppress the frequency drop.

[0100] When the system runtime t is greater than or equal to the switching time When the system enters the frequency recovery phase, the virtual inertia controllers of some or all units should be shut down to stop inertia support. This embodiment uses a control interlocking mechanism to automatically shut down the virtual inertia controllers at the switching point, preventing over-adjustment during the frequency recovery phase and ensuring a smooth return of the frequency to its rated value. This mechanism significantly improves the stability of the system's frequency recovery and effectively avoids frequency overshoot and secondary oscillations.

[0101] In one possible implementation, when the virtual inertia controller of some or all units is turned off, it further includes: A first-order inertial element is introduced to control the smooth exit of frequency modulation power, avoiding secondary frequency abrupt changes caused by the sudden exit of the controller.

[0102] If the virtual inertia controller is suddenly shut down during the switching process, the frequency modulation power will drop to zero instantaneously. This abrupt change may cause secondary frequency fluctuations. To avoid this problem, this embodiment introduces a first-order inertial element to smooth the frequency modulation power when the virtual inertia controller is shut down.

[0103] like Figure 2As shown, the transfer function of the first-order inertial element is: (52) in, It is a time constant. These are Laplace transform variables. The gain of this transfer function is 1, indicating that the output equals the input in steady state. This is achieved by adjusting the time constant. This allows control over the attenuation rate of the frequency modulation power. When When the power is low, the power decays rapidly, almost abruptly; when... When the value is large, the power decay rate is slower, and the exit process is smoother. This embodiment selects an appropriate time constant based on the actual needs of the system. This allows the frequency modulation power to gradually decay to zero after the switching moment, rather than dropping to zero instantaneously. This smooth exit mechanism avoids secondary frequency abrupt changes caused by sudden controller exit, further improving the stability and safety of frequency recovery.

[0104] In this embodiment, by establishing a heterogeneous energy collaborative frequency regulation system model that fits the actual operating conditions of the system and completing the linearization processing of the adaptive model predictive control, the correlation characteristics between frequency deviation and the operating state of the flexible aggregate can be accurately reflected, providing a reliable foundation for subsequent prediction and control. Then, by using rolling optimization to solve the frequency prediction function to obtain future frequency information, the frequency change trend can be predicted in advance. By dynamically adjusting the virtual inertia control coefficient in combination with the frequency change rate and the unit state, the inertia support can be accurately matched with the system requirements. At the same time, the virtual inertia controller is locked at a preset switching time during the frequency recovery phase, which avoids the frequency overshoot problem from the root and comprehensively improves the frequency dynamic response and steady-state recovery performance of the power system.

[0105] To verify that the control strategy proposed in this application can improve the frequency dynamic response and steady-state recovery of the power system, the following simulation experiments were conducted using a simulation system: At the initial rotor speed of the fan =0.8 (pu), energy storage SOC is 0.8, inertia sensitivity factor 1. Maximum frequency change rate In a simulation system with a frequency of 0.5 Hz / s, the following three disturbance scenarios were designed for analysis.

[0106] Disturbance Scenario 1: The load disturbance suddenly increases by 0.08 (pu). According to Equation (27), the system requires 4s of inertia support. If only the synchronous generator unit responds, the system frequency security cannot be guaranteed. As shown in Equation (47), the virtual inertia that the wind power system needs to provide is 2s. Under the strategy proposed in this application, the system requirements can be met by the virtual inertia output of the wind turbine unit alone, without the participation of other units in the flexible aggregate.

[0107] Disturbance Scenario 2: The load disturbance suddenly increases by 0.102 (pu), at which point the system inertia requirement increases to 5.1s. The response of the synchronous generator and the wind turbine is still insufficient to ensure frequency stability. According to Equation (44), the maximum virtual inertia that the wind turbine can provide at the current wind speed is 8s. However, according to the analysis of Equation (47), if the frequency regulation of the energy storage unit is not introduced, the actual equivalent inertia of the system is 5.091s, which is still lower than the minimum inertia requirement. Therefore, in this scenario, an energy storage unit needs to be configured to provide additional inertia support.

[0108] Disturbance Scenario 3: The load disturbance suddenly increases by 0.11 (pu), at which point the system inertia requirement increases to 6s. The coordinated response of the synchronous generator, wind turbine, and energy storage unit is insufficient to ensure frequency stability. According to Equation (45), the maximum virtual inertia that the energy storage unit can provide under the current SOC is 0.9375s. However, according to Equation (48), if electric vehicle charging stations are not introduced to participate in frequency regulation, the actual equivalent inertia of the system is 5.108s, which is lower than the minimum inertia requirement of the system. Therefore, in this scenario, an electric vehicle charging station needs to be configured to provide additional inertia support. The specific values ​​of system inertia requirement, virtual inertia that the synchronous generator can provide, and inertia support that the flexible aggregate needs to supplement under the above three scenarios are summarized in Table 2.

[0109] Table 2 System inertia requirements under different disturbance scenarios

[0110] In Scenario 1, the calculated minimum inertia support required by the system is 4s. Besides the thermal power unit, the wind turbine alone can provide 2s of virtual inertia to meet the system requirements. Energy storage and electric vehicle charging stations do not need to participate in frequency regulation. The simulation results are as follows: Figure 8 As shown. Figure 8 Tables (a)-(c) show the frequency change comparison curves, frequency deviation change curves, and deviation change rate comparison curves for the three control algorithms, respectively. Some detailed data are shown in Table 3. Table 3 shows the minimum frequency point of each control algorithm and the RoCoF comparison results. In Algorithm 1, the synchronous generator unit responded to the disturbance alone without providing additional inertia support, causing the system frequency to drop to a minimum of 49.81Hz, with a maximum RoCoF of -0.423Hz / s. The frequency drop rate was fast and deep, resulting in poor system frequency security. After introducing MPC frequency prediction in Algorithm 2, the frequency regulation performance was significantly improved, with the minimum frequency rising to 49.85Hz and the RoCoF optimized to -0.343Hz / s, verifying the improvement effect of MPC prediction on frequency regulation. Based on MPC prediction, Algorithm 3 combined the system inertia requirements with dynamic coordination of the wind turbine's virtual inertia output to further optimize the control effect, achieving the best performance among the three algorithms: the minimum frequency rose to 49.86Hz, the highest among the three algorithms, and the absolute value of RoCoF dropped to the minimum, only 0.304Hz / s, effectively suppressing the frequency drop rate.

[0111] Table 3. Detailed data for disturbance scenario 1

[0112] Figure 8 Figure (d) shows the power increment change comparison curves of Algorithm 2 and Algorithm 3. Although the final power change is basically the same as the load disturbance, Algorithm 3 has a significantly faster power response speed than Algorithm 2, and can provide power support to the system more quickly. At the same time, Algorithm 3 effectively suppresses overshoot during the frequency recovery process through parameter adaptive control, while Algorithm 2, due to its fixed control parameters, ... The inability to adjust in a timely manner can lead to unnecessary overshooting during the recovery phase. Figure 8 In Figure (e), the power increment curves of the synchronous generator and the flexible aggregate are compared. Power 1 is the power response curve of the synchronous generator and power 2 is the power response curve of the flexible aggregate. By comparison, the power response speed of the flexible aggregate is much faster than that of the synchronous generator, and it can quickly fill the gap in the early stage of disturbance and share the frequency regulation pressure of the synchronous generator. Figure 8 Figure (f) shows the comparison curves of power increment changes in each unit of the flexible polymer, where power changes 1, 2, and 3 represent the power changes of wind power, energy storage, and electric vehicle charging stations, respectively. It can be seen that during the switching time in the virtual inertia response phase... As the wind turbines gradually exit the frequency response, the wind power output alone can meet the system's minimum inertia requirement, eliminating the need for additional energy storage output. This reduces the number of energy storage charge-discharge cycles and helps extend the lifespan of the energy storage units.

[0113] Table 4. Detailed data for disturbance scenario 2

[0114] In Scenario 2, the minimum inertia requirement of the system increases to 5.1s. The coordinated response of the synchronous turbine and the wind turbine still cannot meet the frequency stability requirements. Calculations show that the energy storage unit needs to be activated to provide additional inertia support. Simulation results are as follows: Figure 9As shown in Table 4, some detailed data are presented. Table 4 shows the comparison results of the lowest control frequency and RoCoF for each algorithm: Algorithm 1 did not provide additional inertia support, and its insufficient frequency modulation capability caused the lowest frequency to drop to 49.75Hz, with a maximum RoCoF of -0.542Hz / s, significantly increasing the system frequency security risk. After introducing virtual inertia control, the frequency modulation indicators of Algorithm 2 and Algorithm 3 were significantly improved. Algorithm 2 raised the lowest frequency to 49.79Hz and optimized the RoCoF to -0.440Hz / s. Algorithm 3, on the other hand, relied on parameter adaptive control to achieve efficient allocation of inertia resources, resulting in better control performance. The lowest frequency was further raised to 49.81Hz, and the absolute value of RoCoF dropped to 0.389Hz / s, verifying the effectiveness of the proposed strategy in multi-resource collaborative frequency modulation scenarios.

[0115] Figure 9 In the figure, (a)-(c) represent the comparison curves of frequency change, frequency deviation change and deviation change rate of the three control algorithms, respectively. Figure 9 The power increment curves for Algorithm 2 and Algorithm 3 provided in (d) show that Algorithm 3 has a significantly faster power response speed than Algorithm 2, and through adaptive parameter adjustment, in Algorithm 1 achieves precise matching of power output at all times, effectively suppressing overshoot during the frequency recovery process. Algorithm 2, due to its fixed parameters, cannot adapt to changes in system state in a timely manner, resulting in significant overshoot during the recovery phase. Figure 9 In Figure (e), the curves show the power increment changes of the synchronous unit and the flexible aggregate under Algorithm 3 in Scenario 2. The purple curve represents the power increment change curve of the flexible aggregate, while the green curve represents the power increment change curve of the synchronous unit. The power response speed of the flexible aggregate is much faster than that of the synchronous unit, and Algorithm 3 can provide stronger power support for the system in the early stage of frequency drop, thus rapidly slowing down the frequency drop rate. Figure 9 Figure (f) shows the power increment curves of each unit in the flexible polymer, indicating the switching time during the virtual inertia response phase. As the wind turbines gradually exit the frequency response, the combined output of the wind turbines and energy storage units in this scenario can meet the minimum inertia requirement of the system, and the electric vehicle charging station does not need to output additional power. At the same time, Algorithm 3 makes the frequency recovery process after the system reaches steady state smoother and improves the stability of frequency recovery.

[0116] Table 5. Detailed data for disturbance scenario 3

[0117] Scenario 3 represents the maximum gradient disturbance, further increasing the system's minimum inertia requirement to 6 seconds. Calculations show that even with full-power utilization of wind power and energy storage resources, the system's inertia requirement cannot be met. Therefore, it is necessary to introduce electric vehicle charging stations to participate in frequency regulation, achieving coordinated support between the wind-storage-charging units. Simulation results are as follows: Figure 10 As shown in Table 5, the results of comparing the minimum control frequency and RoCoF for each algorithm are as follows: Under Algorithm 1, the system is subjected to greater disturbance pressure and has no additional inertia support, resulting in deteriorated frequency modulation performance. The minimum frequency drops to 49.73Hz, and the RoCoF reaches -0.584Hz / s, far exceeding the conventional safety limit, indicating extremely poor system frequency safety. After introducing MPC frequency prediction, Algorithm 2 improves the frequency modulation performance to some extent, raising the minimum frequency to 49.77Hz and optimizing the RoCoF to -0.475Hz / s. However, due to the fixed control parameters, the inertia support cannot be precisely matched with the system requirements, resulting in limited improvement. Algorithm 3, relying on multi-source collaborative adaptive control, exhibits the best frequency modulation performance. By dynamically adjusting the virtual inertia control parameters of each unit of the flexible aggregate, it achieves coordinated cooperation among the units, controlling the minimum frequency at 49.79Hz and further optimizing the RoCoF to -0.420Hz / s, effectively reducing the depth and rate of frequency drop.

[0118] Figure 10 In the figure, (a)-(c) represent the comparison curves of frequency change, frequency deviation change and deviation change rate of the three control algorithms, respectively. Figure 10 The power increment curves for Algorithm 2 and Algorithm 3 provided in (d) show that Algorithm 3's power response speed is still significantly faster than Algorithm 2's, and under adaptive parameter control, it suppresses overshoot during frequency recovery. Algorithm 2, due to its fixed parameters, exhibits... The inability to adjust in a timely manner leads to significant overshooting during the recovery phase, posing a risk of secondary oscillations. Figure 10 In the middle (e), the power increment change curves of the synchronous unit and the flexible aggregate under Algorithm 3 are compared. This once again verifies the fast response advantage of the flexible aggregate compared with the synchronous unit. It can provide emergency power support for the system in the early stage of disturbance and effectively alleviate the frequency drop trend. Figure 10 Figure (f) shows the power increment curves of each unit in the flexible polymer, indicating the switching time during the virtual inertia response phase. As the wind turbines gradually exit the frequency response, Algorithm 3 achieves precise matching of system inertia requirements through three-unit coordinated frequency regulation. Compared with Algorithm 2, it not only significantly reduces the maximum frequency drop, but also effectively reduces the maximum frequency change rate, ensuring that the system meets frequency safety constraints. At the same time, the overshoot is lower, reducing the risk of secondary oscillation during the frequency recovery process.

[0119] Simulation results for three gradient perturbation scenarios fully validate the effectiveness, robustness, and multi-scenario adaptability of the proposed MPC-based flexible aggregate virtual inertia adaptive assisted frequency modulation control strategy. Compared with traditional parameter-fixed virtual inertia control (Algorithm 1) and MPC-based parameter-fixed virtual inertia control (Algorithm 2), the proposed strategy (Algorithm 3) exhibits significant performance advantages: (1) Stronger frequency modulation support: It can effectively increase the lowest frequency point and reduce the maximum frequency change rate under various disturbance scenarios. Even under a large disturbance scenario of 0.11 (pu), it can still control the frequency drop depth and rate within a safe range, solving the frequency safety problem caused by insufficient inertia in high-proportion new energy systems. (2) Faster power response: Relying on the frequency trend prediction and parameter adaptive adjustment of MPC, it can quickly provide power support to the system in the early stage of disturbance, and the response speed is significantly better than the comparison algorithm; (3) Better overshoot suppression effect: with The virtual inertial response stage switching moment enables the orderly withdrawal of frequency modulation power of each unit, effectively suppressing overshoot and secondary oscillation during the frequency recovery process and improving the stability of frequency recovery. (4) Higher resource utilization efficiency: The frequency modulation resource call combination is accurately matched according to the real-time inertia demand of the system. Low disturbance only calls the wind turbine, medium disturbance calls the wind turbine and energy storage, and large disturbance realizes multi-unit collaboration, avoiding the ineffective action of energy storage and electric vehicle charging station, reducing the frequency of energy storage charging and discharging, which is conducive to extending the service life of equipment. (5) Better adaptability to multiple scenarios: It can achieve accurate matching between inertial support and system requirements under different gradient load disturbances, showing good robustness and is suitable for complex operating conditions of high-proportion new energy systems.

[0120] In summary, the control strategy proposed in this application can fully exploit the virtual inertia response potential of flexible aggregates when the frequency regulation power of synchronous units is insufficient, achieving coordinated frequency regulation of heterogeneous energy sources. By dynamically adjusting control parameters to achieve precise matching of inertia support, it effectively reduces the operating frequency of energy storage units and suppresses overshoot during the frequency recovery phase, significantly improving the dynamic response characteristics and steady-state recovery stability of the system frequency. Simulation results show that the proposed strategy outperforms traditional frequency support control methods in terms of frequency drop suppression, rate of change control, and overshoot avoidance, providing an effective solution for frequency stability control of high-proportion renewable energy grid-connected power systems.

[0121] 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 this application.

[0122] This application also provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the above method embodiments. Exemplarily, the electronic device may be a smartphone, tablet computer, laptop computer, desktop computer, smart speaker, smartwatch, etc., and is not limited thereto.

[0123] 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.

[0124] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application 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 this application, and should all be included within the protection scope of this application.

Claims

1. A method for frequency modulation control of virtual inertia in a flexible polymer, characterized in that, include: A model of heterogeneous energy cooperative frequency regulation system considering system frequency deviation, operating status of each heterogeneous energy unit in flexible polymer and dynamic constraints is established, and it is linearized into a state-space model suitable for model predictive control. Construct a system frequency prediction function that minimizes frequency deviation and control cost, and solve for the system frequency information at the next moment through model predictive control rolling optimization. Based on the solved system frequency information, the virtual inertia control coefficients of each unit in the flexible polymer body are dynamically adjusted by integrating the system frequency change rate and the unit state information of each heterogeneous energy unit. During the frequency recovery phase, the virtual inertia controller of some or all units is turned off by using the pre-generated virtual inertia response phase switching time as the control interlocking condition.

2. The method for frequency modulation control of virtual inertia of flexible polymers according to claim 1, characterized in that, The establishment of a heterogeneous energy coordinated frequency regulation system model that considers system frequency deviation, the operating status of each heterogeneous energy unit within the flexible polymer, and dynamic constraints includes: Wind turbines, photovoltaic units, energy storage units, and electric vehicle charging stations are taken as components of a flexible aggregate. Mathematical models of the virtual rotational inertia and virtual inertial time constant of each unit are established, and the control law for virtual inertia control of each unit through differential control is determined.

3. The method for frequency modulation control of virtual inertia of flexible polymers according to claim 2, characterized in that, Establishing a model for a heterogeneous energy coordinated frequency regulation system also includes: The model integrates safety constraints on the rate of change of frequency and establishes a calculation expression for the equivalent virtual inertia time constant of the system, thus characterizing the overall inertia response capability of the flexible polymer as a comprehensive weighted sum of the virtual inertia contributions of its internal units.

4. The method for frequency modulation control of virtual inertia of flexible polymers according to claim 1, characterized in that, The linearization of this model into a state-space model suitable for model predictive control includes: The system frequency dynamic response model is discretized using the Euler method, and state variables, control inputs and disturbance variables are defined to construct a discrete state-space model. By defining the prediction time domain and the control time domain, the discrete state-space model is extended into a prediction model.

5. The virtual inertia frequency modulation control method for flexible polymers according to claim 1, characterized in that, The construction of the system frequency prediction function that minimizes frequency deviation and control cost includes: The system frequency prediction function is transformed into a quadratic programming form using the Lyapunov equation, and a weighted diagonal matrix is ​​set to balance the weights between minimizing frequency deviation and minimizing control cost.

6. The method for frequency modulation control of virtual inertia of flexible polymers according to claim 5, characterized in that, The construction of the system frequency prediction function that minimizes frequency deviation and control cost also includes: At least one of the following constraints—active power output of thermal power units, kinetic energy reserve and active power output of wind power units, active power output of photovoltaic units, charging and discharging power and state of charge of energy storage units, and charging power of electric vehicle charging stations—is integrated into the system frequency prediction function of the quadratic programming form.

7. The method for frequency modulation control of virtual inertia of flexible polymers according to claim 1, characterized in that, The process of dynamically adjusting the virtual inertia control coefficients of each unit within the flexible polymer body, based on the solved system frequency information and integrating the system frequency change rate and the unit state information of each heterogeneous energy unit, includes: Based on the magnitude of the system load disturbance and the reassessed minimum inertia requirement of the system, the inertia requirement that each unit in the flexible polymer body needs to provide is determined, and the virtual inertia control coefficient of each unit is solved based on the inertia requirement.

8. The method for frequency modulation control of virtual inertia of flexible polymers according to claim 7, characterized in that, The determination of the inertia requirements to be provided by each unit within the flexible polymer includes: When the load disturbance is less than the first disturbance value, only the wind turbine generators are used to provide virtual inertia support; When the load disturbance is greater than or equal to the first disturbance value and less than the second disturbance value, wind turbine units and energy storage units are coordinated to provide virtual inertia support. When the load disturbance is greater than or equal to the second disturbance value, wind turbine units, energy storage units and electric vehicle charging stations are coordinated to provide virtual inertia support.

9. The virtual inertia frequency modulation control method for flexible polymers according to claim 1, characterized in that, The step of using the pre-generated virtual inertia response stage switching time as a control interlocking condition to shut down the virtual inertia controllers of some or all units includes: Based on the system frequency dynamic response model, the moment when the system frequency change rate is zero is solved by Laplace transform and inverse Laplace transform, and this moment is taken as the switching moment of the virtual inertia response stage; When the system running time is less than the switching time, enable the virtual inertia controller of each unit and tune the parameters according to the solved control coefficients; When the system runtime is greater than or equal to the switching time, the virtual inertia controllers of some or all units are turned off.

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 9.

Images