A network-constructed wind and storage system frequency regulation optimization method based on explicit model predictive control

Abstract

The application provides a networked wind and storage system frequency regulation optimization method based on explicit model predictive control, comprising: obtaining real-time state data of the wind and storage system, the real-time state data comprising storage working state, system demand power, wind power working state and wind speed; determining state variables, control inputs and disturbance inputs according to the real-time state data; obtaining an initial state space equation according to the action relationship of the state variables, the control inputs and the disturbance inputs on the wind and storage system; performing discretization processing on the initial state space equation to obtain a discrete state space equation; obtaining a power tracking error prediction model according to a coefficient matrix of the discrete state space equation; obtaining an optimization control model according to the power tracking error prediction model and a control quantity variation range; and obtaining wind power control strategies and storage control strategies of the wind and storage system according to the real-time state data and the optimization control model. The application can improve the response efficiency, continuity and accuracy of the wind and storage system frequency regulation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of new energy power generation control technology, in particular to a grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control. BACKGROUND

[0002] Under the background of grid-connected new energy power generation gradually replacing traditional synchronous power supply, wind power and energy storage system need to have stronger active frequency support capability to cope with the increasingly prominent inertia deficiency and frequency fluctuation problems in the power system. The existing control strategy has the following bottlenecks when facing actual working conditions: first, the charging and discharging efficiency of the energy storage system is significantly affected by the SOC state, which is difficult to accurately control; second, the speed of the wind turbine generator is coupled with the inertia, and it is difficult to achieve effective frequency regulation in the full wind speed range by relying on a single rotor regulation mechanism; third, the frequency response demand is variable in multiple time scales, and traditional control methods are difficult to balance dynamic adaptability and control accuracy. SUMMARY

[0003] The technical problem to be solved by the present application is to provide a grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control. The response efficiency, persistence and accuracy of the wind storage system frequency regulation can be improved.

[0004] To solve the above technical problems, the technical solutions of the present application are as follows:

[0005] A grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control, comprising:

[0006] Obtaining real-time state data of the wind storage system, the real-time state data including energy storage working state, system demand power, wind power working state and wind speed;

[0007] According to the real-time state data, determining state variables, control inputs and disturbance inputs;

[0008] According to the action relationship of the state variables, control inputs and disturbance inputs on the wind storage system, obtaining an initial state space equation;

[0009] Discretizing the initial state space equation to obtain a discrete state space equation;

[0010] According to the coefficient matrix of the discrete state space equation, obtaining a power tracking error prediction model;

[0011] According to the power tracking error prediction model, combining the control variable range, obtaining an optimal control model;

[0012] According to the real-time state data and the optimal control model, obtaining wind power control strategy and energy storage control strategy of the wind storage system.

[0013] Optionally, according to the real-time state data, state variables, control inputs and disturbance inputs are determined, including:

[0014] The energy storage state of charge in the real-time state data and system demand power are taken as state variables; the wind power, energy storage power and frequency modulation power are taken as control inputs; and the system frequency and wind speed are taken as disturbance inputs.

[0015] Optionally, according to the action relationship of the state variables, control inputs and disturbance inputs on the wind storage system, an initial state space equation is obtained, including:

[0016] According to the state variables, control inputs, disturbance inputs and wind storage system device coefficients, an updated state is obtained.

[0017] According to the state variables, a wind storage system output state is obtained.

[0018] According to the updated state and the wind storage system output state, an initial state space equation is obtained.

[0019] Optionally, the initial state space equation is discretized to obtain a discrete state space equation, including:

[0020] According to a preset discrete time domain step, the coefficient matrix of the initial state space equation is discretized to obtain a discrete state space equation.

[0021] Optionally, according to the coefficient matrix of the discrete state space equation, a power tracking error prediction model is obtained, including:

[0022] According to the coefficient matrix of the discrete state space equation, a state deviation matrix, a control deviation matrix and a disturbance deviation matrix are obtained.

[0023] According to the state deviation matrix, the control deviation matrix and the disturbance deviation matrix, in combination with a state increment, a control increment and a disturbance increment, a power tracking error prediction model is obtained.

[0024] Optionally, according to the power tracking error prediction model, in combination with a control amount variation range, an optimal control model is obtained, including:

[0025] According to the power tracking error prediction model, a power deviation vector is obtained.

[0026] According to the initial state space equation, a control constraint matrix and a control cost matrix are obtained.

[0027] According to the power deviation vector, the control constraint matrix and the control cost matrix, in combination with a control amount variation range, an optimal control model is obtained.

[0028] Optionally, according to the real-time state data and the optimization control model, a wind power control strategy and an energy storage control strategy are obtained, comprising:

[0029] According to the discrete state space equation, each discrete point state variable is obtained.

[0030] The real-time state data and each discrete point state variable are input into the optimization control model, so as to obtain the wind power control strategy and the energy storage control strategy of the wind storage system.

[0031] The above technical scheme of the present application has at least the following technical effects:

[0032] The above explicit model predictive control-based frequency regulation optimization method of the grid-connected wind storage system of the present application can improve the response efficiency, continuity and accuracy of the frequency regulation of the wind storage system by obtaining real-time state data of the wind storage system, determining state variables, control inputs and disturbance inputs according to the real-time state data, obtaining an initial state space equation according to the action relationship of the state variables, control inputs and disturbance inputs on the wind storage system, performing discretization processing on the initial state space equation to obtain a discrete state space equation, obtaining a power tracking error prediction model according to the coefficient matrix of the discrete state space equation, obtaining an optimization control model according to the power tracking error prediction model and the control amount variation range, and obtaining a wind power control strategy and an energy storage control strategy of the wind storage system according to the real-time state data and the optimization control model. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 is a flowchart of the explicit model predictive control-based frequency regulation optimization method of the grid-connected wind storage system of the present application;

[0034] Figure 2 is a control framework diagram of the explicit model predictive control-based frequency regulation optimization method of the grid-connected wind storage system of the present application;

[0035] Figure 3 is a wind power control strategy diagram of the explicit model predictive control-based frequency regulation optimization method of the grid-connected wind storage system of the present application;

[0036] Figure 4 is an energy storage control strategy diagram of the explicit model predictive control-based frequency regulation optimization method of the grid-connected wind storage system of the present application;

[0037] Figure 5 is a simulation system diagram of the explicit model predictive control-based frequency regulation optimization method of the grid-connected wind storage system of the present application;

[0038] Figure 6is a frequency response comparison chart of the grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control of the present application and an existing control method;

[0039] Figure 7 is a wind power output power comparison chart of the grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control of the present application and an existing control method;

[0040] Figure 8 is a storage output power comparison chart of the grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control of the present application and an existing control method;

[0041] Figure 9 is a wind turbine speed comparison chart of the grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control of the present application and an existing control method;

[0042] Figure 10 is a storage SOC comparison chart of the grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control of the present application and an existing control method;

[0043] Figure 11 is a schematic diagram of the grid-connected wind storage system frequency regulation optimization device based on explicit model predictive control of the present application. DETAILED DESCRIPTION

[0044] Exemplary embodiments of the present application will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present application can be more thoroughly understood, and the scope of the present application can be accurately conveyed to those skilled in the art.

[0045] As shown in Figure 1 , an embodiment of the present application proposes a grid-connected wind storage system frequency regulation optimization method based on explicit model predictive control, comprising:

[0046] Step S1, obtaining real-time state data of the wind storage system, the real-time state data including storage working state, system demand power, wind power working state and wind speed;

[0047] Step S2, determining state variables, control inputs and disturbance inputs according to the real-time state data;

[0048] Step S3, obtaining an initial state space equation according to the action relationship of the state variables, control inputs and disturbance inputs on the wind storage system;

[0049] Step S4, discretizing the initial state space equation to obtain a discrete state space equation;

[0050] Step S5, obtaining a power tracking error prediction model according to a coefficient matrix of the discrete state space equation;

[0051] Step S6, obtaining an optimal control model according to the power tracking error prediction model and in combination with a control quantity variation range;

[0052] Step S7, obtaining a wind power control strategy and an energy storage control strategy of the wind storage system according to the real-time state data and the optimal control model.

[0053] In the embodiment, as shown in the figure, Figure 1 In the grid-connected wind storage system frequency modulation optimization method based on explicit model predictive control, first, real-time state data of the wind storage system is obtained, the wind storage system including a wind power system and an energy storage system, the real-time state data including an energy storage working state, system demand power, wind power working state and wind speed; specifically, energy storage state of charge (SOC), system demand power, wind power, energy storage power, VSG (virtual synchronous generator) primary frequency modulation power, system frequency and wind speed are obtained, wind power frequency modulation output power P add , energy storage SOC adjustment coefficient k soc , current adjustable energy of the wind turbine relative to its maximum adjustable energy ratio k r , current variable output power of the wind turbine relative to its maximum variable output power ratio k p , energy storage reference output droop coefficient k bess ; then, a discrete state space equation of the wind storage system, a power tracking error prediction model and an optimal control model are established; finally, a wind power control strategy and an energy storage control strategy of the wind storage system are obtained according to the real-time state data and the optimal control model.

[0054] As shown in the figure, Figure 2 By modeling based on the state space of the wind storage system, dynamically considering system disturbance and nonlinear characteristics, introducing an explicit rolling optimization solution structure, realizing fast table lookup of the control law and real-time response of the system, and optimizing wind power reserve power and energy storage SOC (state of charge of energy storage) trajectory, the system frequency regulation continuity and economy are improved. The present application can significantly reduce the maximum frequency deviation and steady-state error, and improve the safety margin and energy efficiency of system operation.

[0055] The application proposes a comprehensive control strategy fusing SOC dynamic management, rotor kinetic energy constraint and energy storage energy feedback, and improves the frequency modulation capability of the grid-connected wind storage system through multi-time scale collaborative optimization. On the basis of considering frequency modulation accuracy, equipment operation constraint and economy, a frequency modulation control framework based on multi-objective optimization predictive control is constructed. Through online model updating and disturbance estimation, the control accuracy and robustness of the nonlinear, multivariable and multi-time scale coupled system can be improved. The grid-connected wind storage system frequency modulation control strategy optimizes the dynamic response and stability of frequency regulation through rolling optimization and feedback correction mechanism, and improves the robustness and adaptability of the system. The multi-objective optimization framework comprehensively considers the accurate tracking of frequency, the safety of wind turbine power reserve and the economy of energy storage power. The frequency modulation output power of the wind power system is proportional to the reserve capacity, and optimizing the wind power frequency modulation output can reduce power loss and improve wind energy utilization. The energy storage system optimizes the battery discharge trajectory and introduces SOC constraint to balance economy and frequency modulation capability, and avoid rapid decline of energy storage SOC.

[0056] In an optional embodiment of the application, in step S2, the state variables, control inputs and disturbance inputs are determined according to the real-time state data, including:

[0057] In step S21, the energy storage state of charge in the real-time state data and the system demand power are taken as the state variables; the wind power, the energy storage power and the frequency modulation power are taken as the control inputs, and the system frequency and the wind speed are taken as the disturbance inputs.

[0058] In this embodiment, the energy storage SOC and the system demand power P are taken as the state variables ; the wind power, the energy storage power and the VSG primary frequency modulation power are taken as the control inputs , wherein is the wind power, is the energy storage power, is the VSG primary frequency modulation power; the system frequency and the wind speed change are taken as the disturbance inputs , wherein f is the system frequency and v is the wind speed.

[0059] In an optional embodiment of the application, in step S3, the initial state space equation is obtained according to the action relationship of the state variables, the control inputs and the disturbance inputs on the wind storage system, including:

[0060] In step S31, the updated state is obtained according to the state variables, the control inputs, the disturbance inputs and the wind storage system device coefficients.

[0061] In step S32, the wind storage system output state is obtained according to the state variables.

[0062] Step S33, according to the updated state and the wind storage system output state, the initial state space equation is established.

[0063] In this embodiment, the updated state is obtained according to the state variable, the control input, the disturbance input and the wind storage system device coefficient; the system output is obtained according to the state variable; the updated state and the system output are combined to establish the state space equation of the grid-connected wind storage system as:

[0064]

[0065] Wherein, is the state variable at t, is the updated state at t, is the control input, is the disturbance input, is the output vector,

[0066] , k soc indicates the energy storage SOC adjustment coefficient, k r indicates the ratio of the current adjustable energy of the wind turbine to the maximum adjustable energy thereof, k p indicates the ratio of the current variable output power of the wind turbine to the maximum variable output power thereof, indicates the VSG inertia coefficient, k bess indicates the energy storage reference output droop coefficient; indicates the correlation coefficient of the variable rotor kinetic energy ratio and the variable output power ratio, indicates the VSG primary frequency modulation coefficient, indicates the wind power reserve power proportion coefficient, indicates the compensation coefficient, indicates the current available charge and discharge capacity of the energy storage, indicates the system electric angular frequency coefficient, indicates the inertia constant of the wind turbine rotor, indicates the discharge efficiency of the energy storage, indicates the charge efficiency of the energy storage.

[0067]

[0068] , indicates a diagonal matrix;

[0069]

[0070] In an optional embodiment of the present application, the initial state space equation is discretized in step S4 to obtain a discrete state space equation, comprising:

[0071] Step S41, according to the preset discrete time domain step, the coefficient matrix of the initial state space equation is discretized to obtain a discrete state space equation.

[0072] In this embodiment, in order to obtain the optimal control sequence of the wind storage system in the prediction time domain, the initial state space equation is discretized as follows:

[0073]

[0074] Wherein , denotes the discretized matrix of C, is the discrete time domain step.

[0075] In an optional embodiment of the present application, in step S5, the power tracking error prediction model is obtained according to the coefficient matrix of the discrete state space equation, comprising:

[0076] Step S51, according to the coefficient matrix of the discrete state space equation, a state deviation matrix, a control deviation matrix and a disturbance deviation matrix are obtained;

[0077] Step S52, according to the state deviation matrix, the control deviation matrix and the disturbance deviation matrix, combined with the state increment, the control increment and the disturbance increment, the power tracking error prediction model is obtained.

[0078] In this embodiment, the power tracking error prediction model in the prediction time domain is:

[0079]

[0080] Wherein,

[0081]

[0082]

[0083]

[0084]

[0085] In the formula, denotes the deviation between the actual power and the target power at the k+1 time in the future based on the current state at the k time, , is the state increment, is the control increment, that is, the change value of the energy storage power, the wind power and the VSG primary frequency power control amount; is the disturbance increment, that is, the change value of the system frequency at the reference point and the wind speed; is a state deviation matrix, is a control deviation matrix, is an interference deviation matrix, p is a prediction time domain, and q is a control time domain.

[0086] In an optional embodiment of the present application, in step S6, an optimal control model is obtained according to the power tracking error prediction model and in combination with a control variable change range, including:

[0087] In step S61, a power deviation vector is obtained according to the power tracking error prediction model.

[0088] In step S62, a control constraint matrix and a control cost matrix are obtained according to the initial state space equation.

[0089] In step S63, an optimal control model is obtained according to the power deviation vector, the control constraint matrix and the control cost matrix and in combination with the control variable change range.

[0090] In this embodiment, the power tracking error is kept at a minimum value according to the power tracking error prediction model, and a power deviation vector at this time is obtained , wherein is a storage charge reference value, is a system demand power reference value, the control constraint matrix controls the importance of input change over time, and the control cost matrix weights the output at future time, which is helpful to adjust the minimization of future output error; specifically, the control constraint matrix , the control cost matrix , in combination with the storage output power, the economy of storage charge and discharge power, the wind power operation stability and the primary frequency effectiveness in the prediction time domain are considered, and a multi-objective optimal control problem is expressed as:

[0091]

[0092] wherein, represents an optimal target, represents a storage charge optimization target, represents a power response optimization target, represents a control cost optimization target, p is a prediction time domain, and q is a control time domain. represents a 2-norm of a value L, represents a control variable at the k+i discrete time, represents a transpose matrix of u, represents a control increment at the k+i-1 discrete time, This represents the minimum value of the control increment. This indicates the maximum value of the control increment. This represents the minimum charge of the energy storage device. E represents the maximum energy storage charge, and E represents the system power demand. This represents the minimum power requirement of the system. This indicates the maximum power required by the system. This indicates a constraint condition.

[0093] In an optional embodiment of the present invention, step S7, based on the real-time state data and the optimized control model, yields a wind power control strategy and an energy storage control strategy, including:

[0094] Step S71: Obtain the state variables of each discrete point according to the discrete state space equation;

[0095] Step S72: Input the real-time state data and the state variables of each discrete point into the optimization control model to obtain the wind power control strategy and energy storage control strategy of the wind-storage system.

[0096] In this embodiment, given the system state variable x(k), the explicit optimal control law is solved offline using a rolling time-domain optimization strategy. The steps are as follows:

[0097] (1) Obtain the system state x(k) at time k.

[0098] (2) Considering the discrete linear time-invariant system with discrete state-space equations, we can obtain the following using its linear transitivity property:

[0099]

[0100] in, Let k be the control quantity at the (k+ij-1)th discrete time. Let j be the disturbance quantity at the (k+ij-1)th discrete time, where j = 0, ..., i-1. This represents the matrix after A is discretized. This represents the matrix after B is discretized. Represents the matrix after D is discretized;

[0101] By inputting real-time state data and discrete-point state variables into the optimization control model, and adding constraints on the linear transfer property of the offline system, the expression is obtained:

[0102]

[0103] Where F* is the optimal objective, x(k) is the initial state at time k, and U(k) is the control sequence at time k. U(k) is the discrete matrix of U(k). F*, U(k), x(k) are regarded as optimal value function, decision variable and parameter vector respectively, so that the optimization problem is converted into a multi-parameter programming problem in standard form.

[0104] (3) The first control quantity calculated is applied to the multi-objective optimization control of the frequency tracking of the wind storage system.

[0105] (4) The prediction time domain is advanced by one step, and the above process is repeated at k+1 to obtain the optimal control law in the form of a segmented affine function:

[0106]

[0107] wherein, is the state feedback gain coefficient matrix, which embodies the system state Linear relationship of direct drive control quantity; after completing the calculation of a control region, the system switches to the next region until the entire feasible solution space is covered. Finally, the SOC ref and E ref are discretized into 8 intervals to obtain 64 groups of solutions corresponding to different control strategies.

[0108] In this embodiment, the wind power control strategy is a control strategy under a preset energy threshold, and the energy storage control strategy is a control strategy under a preset energy storage charge threshold. Each control strategy is divided into four regions for analysis to provide support for subsequent optimization.

[0109] (1) E>E ref and SOC>SOC ref : The system demand power exceeds the reference value, and the frequency rises. The wind turbine accelerates to absorb the excess power, and the energy storage system does not need to be charged because SOC>SOC ref .

[0110] (2) E>E ref and SOC<SOC ref : The system frequency rises, the wind turbine absorbs excess power by accelerating the rotor, and the energy storage system charges to adjust the frequency, but needs to avoid overcharging.

[0111] (3) E<E ref and SOC>SOC ref : The system frequency drops, the wind rotor responds to frequency modulation through inertia, and the energy storage system (SOC>SOC ref ) provides power supplement in the primary frequency modulation stage.

[0112] (4) E<E ref and SOC<SOC ref: system frequency rises and the energy storage SOC is lower than the reference value, the system power shortage is mainly provided by the grid-connected wind power, and the system frequency regulation is responsible.

[0113] As shown in Figure 3 , the control strategy of the wind power when the SOC ref is 65% and the E ref is 0 is shown; as shown in Figure 4 , the control strategy of the energy storage power when the SOC ref is 65% and the E ref is 0 is shown.

[0114] The effectiveness of the scheme is verified as follows:

[0115] First, the simulation system is constructed, and the simulation system is composed of grid-connected wind turbines, energy storage systems, synchronous generators and alternating current loads, as shown in Figure 5 . Among them, the wind farm is composed of 15 2MW doubly-fed wind turbines, the unit output voltage is 690V, and the voltage is raised to 35kV through a step-up transformer, and the supporting energy storage system has a rated power of 10% of the wind power rated power, and the capacity is configured as 3MW / 3MWh, and the simulation model also includes a 60MW rated capacity synchronous machine and a random fluctuating alternating current load. The upper and lower limits of the system primary frequency regulation dead zone are set to ±0.033Hz. The prediction period of the optimization strategy is set to 40s, and the sampling time interval Δt is 1s.

[0116] In order to verify the effectiveness of the proposed strategy under variable wind speed conditions, the following three kinds of collaborative control strategies of the grid-connected wind storage system are compared and analyzed under the step load disturbance condition: ① improved droop control; ② traditional MPC control; ③ multi-objective optimization control of the scheme.

[0117] The effectiveness of the strategy in this paper under the condition of high wind speed 12m / s is verified. The initial SOC is set to 50%, and a step load disturbance with an amplitude of 30MW is added at 5s. The power grid frequency deviation, wind turbine output, wind turbine speed, energy storage system output and energy storage SOC change curve are shown in Figure 6-10 .

[0118] Under the condition of high wind speed, the multi-objective optimization control of the wind storage system in the application adjusts the power output adaptively. Compared with the other two control modes, the maximum frequency deviation Δf max is reduced by 16.88% and 10.29% respectively; at the same time, compared with the traditional MPC control, the steady-state frequency deviation Δf s is reduced by 5.52%, and the frequency regulation effect is continuously improved.

[0119] Table 1 Frequency regulation index results under high wind speed

[0120] Control strategy Delta F max / Hz Delta F s / Hz SOC / % v SOC ]]> r min ]]> Droop control 0.443 0.252 0.326 1.01 1.117 Conventional MPC control 0.418 0.229 0.343 1.06 1.110 Multi-objective optimization control 0.379 0.217 0.375 1.15 1.108

[0121] At the initial stage of frequency disturbance, the energy storage system provides large power support rapidly and then gradually reduces the output. At the same time, the grid-connected wind turbine actively participates in the grid frequency regulation by fully releasing the rotor kinetic energy. At 21.07 seconds after the disturbance, the rotor speed of the wind turbine drops to the minimum of 1.108 pu. Finally, the state of charge (SOC) of the energy storage system drops to 0.375 pu. Compared with the traditional MPC control strategy, the maximum output depth of the energy storage system is reduced by about 9.32%. The synergistic effect of the energy storage system and the grid-connected wind turbine effectively slows down the frequency drop rate of the system.

[0122] Therefore, the multi-objective optimization control controls the output ratio of the grid-connected wind storage by dynamic optimization, fully utilizes the frequency regulation advantages of the wind power and the energy storage, realizes the performance improvement of frequency regulation in high wind speed conditions and stable operation in low wind speed conditions, significantly enhances the operation stability and economy of the wind storage combined system, and improves the continuity and safety of the wind storage combined system in the frequency regulation process.

[0123] The application provides a multi-objective rolling frequency regulation control method for a grid-connected wind storage system, which effectively solves the problems of response delay, regulation rigidity and insufficient economy of the existing wind storage system in frequency support. The control strategy fully integrates the kinetic energy characteristics of the wind turbine, the state of charge management of the energy storage and the multi-time scale frequency disturbance mechanism on the basis of system state modeling, and realizes the dynamic synergistic optimization of the frequency response capability, operation stability and economy of the wind storage system.

[0124] Simulation experiments prove that the application significantly reduces the maximum frequency deviation and steady-state error of the system in high wind speed disturbance conditions, improves the frequency support accuracy and system robustness. Compared with the traditional control strategy, the method has better dynamic response performance and adaptability, and can still maintain efficient and stable frequency regulation control effect in complex scenes such as severe system power fluctuation or limited energy storage boundary.

[0125] The application has the advantages of clear structure, efficient control and strong adaptability, and can be widely applied to the grid-connected wind power and energy storage combined operation system, has good engineering implementability and industrial promotion value, and can provide core technical support for stable operation of large-scale new energy power systems.

[0126] As shown in Figure 11 The embodiment of the application also provides a grid-connected wind storage system frequency regulation optimization device 110 based on an explicit model predictive control, which comprises:

[0127] An acquisition module 111 is configured to acquire real-time state data of the wind storage system, wherein the real-time state data comprises an energy storage working state, system demand power, a wind power working state and a wind speed.

[0128] The processing module 112 is configured to determine state variables, control inputs and disturbance inputs according to the real-time state data; obtain an initial state space equation according to an action relationship of the state variables, the control inputs and the disturbance inputs on the wind storage system; perform discretization processing on the initial state space equation to obtain a discrete state space equation; obtain a power tracking error prediction model according to a coefficient matrix of the discrete state space equation; obtain an optimal control model according to the power tracking error prediction model and a control quantity variation range; and obtain a wind power control strategy and an energy storage control strategy of the wind storage system according to the real-time state data and the optimal control model.

[0129] Optionally, the determining of the state variables, the control inputs and the disturbance inputs according to the real-time state data comprises:

[0130] The state variables include an energy storage state of charge in the real-time state data and a system demand power; the control inputs include wind power, energy storage power and frequency modulation power; and the disturbance inputs include a system frequency and a wind speed.

[0131] Optionally, the obtaining of the initial state space equation according to the action relationship of the state variables, the control inputs and the disturbance inputs on the wind storage system comprises:

[0132] The updated state is obtained according to the state variables, the control inputs, the disturbance inputs and a device coefficient of the wind storage system.

[0133] The wind storage system output state is obtained according to the state variables.

[0134] The initial state space equation is obtained according to the updated state and the wind storage system output state.

[0135] Optionally, the discretization processing of the initial state space equation to obtain the discrete state space equation comprises:

[0136] The discrete state space equation is obtained by performing discretization processing on the coefficient matrix of the initial state space equation according to a preset discretization time domain step.

[0137] Optionally, the obtaining of the power tracking error prediction model according to the coefficient matrix of the discrete state space equation comprises:

[0138] The state deviation matrix, the control deviation matrix and the disturbance deviation matrix are obtained according to the coefficient matrix of the discrete state space equation.

[0139] The power tracking error prediction model is obtained according to the state deviation matrix, the control deviation matrix and the disturbance deviation matrix, and a state increment, a control increment and a disturbance increment.

[0140] Optionally, according to the power tracking error prediction model, the optimized control model is obtained in combination with the control variable change range, including:

[0141] According to the power tracking error prediction model, the power deviation vector is obtained.

[0142] According to the initial state space equation, the control constraint matrix and the control cost matrix are obtained.

[0143] According to the power deviation vector, the control constraint matrix and the control cost matrix, the optimized control model is obtained in combination with the control variable change range.

[0144] Optionally, according to the real-time state data and the optimized control model, the wind power control strategy and the energy storage control strategy are obtained, including:

[0145] According to the discrete state space equation, the state variables of each discrete point are obtained.

[0146] The real-time state data and the state variables of each discrete point are input into the optimized control model, and the wind power control strategy and the energy storage control strategy of the wind storage system are obtained.

[0147] It should be noted that all the implementation manners in the above method embodiments are applicable to the embodiments of the device and can achieve the same technical effects.

[0148] Embodiments of the present application also provide a computing device, including: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the explicit model predictive control based networked wind storage system frequency regulation optimization method described in the present application. All implementation manners in the above method embodiments are applicable to the embodiments of the computing device and can achieve the same technical effects.

[0149] Embodiments of the present application also provide a computer readable storage medium, the computer readable storage medium stores a program, the program is executed by a processor to implement the explicit model predictive control based networked wind storage system frequency regulation optimization method described in the present application. All implementation manners in the above method embodiments are applicable to the embodiments of the computer readable storage medium and can achieve the same technical effects.

[0150] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0151] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0152] In the embodiments provided by the present application, it should be understood that the disclosed device and method can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the units is only a logical function division. There can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0153] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., they can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0154] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit.

[0155] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the embodiments of the method of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk, and various storage media that can store program codes.

[0156] Furthermore, it is pointed out that in the device and method of the present application, obviously, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations are to be considered as equivalent solutions of the present application. Moreover, the steps of performing the above series of processes can naturally be executed in time sequence according to the order of description, but do not necessarily have to be executed in time sequence. Some steps can be executed in parallel or independently of each other. It is understood by those skilled in the art that all or any of the steps or components of the method and device of the present application can be implemented in hardware, firmware, software, or a combination thereof, in any computing device (including a processor, a storage medium, etc.) or network of computing devices, using the basic programming skills of those skilled in the art upon reading the description of the present application.

[0157] Therefore, the object of the present application can also be achieved by running a program or a set of programs on any computing device. The computing device can be a commonly known general-purpose device. Therefore, the object of the present application can also be achieved only by providing a program product containing program code for implementing the method or device. That is, such a program product also constitutes the present application, and a storage medium storing such a program product also constitutes the present application. Obviously, the storage medium can be any commonly known storage medium or any storage medium developed in the future. It is also pointed out that in the device and method of the present application, obviously, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations are to be considered as equivalent solutions of the present application. Moreover, the steps of performing the above series of processes can naturally be executed in time sequence according to the order of description, but do not necessarily have to be executed in time sequence. Some steps can be executed in parallel or independently of each other.

[0158] The above is the preferred embodiment of the present application. It should be pointed out that for those skilled in the art, several improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements are also to be considered as the protection scope of the present application.

Claims

1. A method for frequency regulation optimization of a grid-connected wind storage system based on explicit model predictive control, characterized in that, The method comprises the following steps: acquiring real-time state data of the wind storage system, the real-time state data comprising an energy storage working state, system demand power, wind power working state and wind speed; determining state variables, control inputs and disturbance inputs according to the real-time state data; obtaining an initial state space equation according to the action relationship of the state variables, control inputs and disturbance inputs on the wind storage system; discretizing the initial state space equation to obtain a discrete state space equation; obtaining a power tracking error prediction model according to the coefficient matrix of the discrete state space equation; obtaining an optimal control model according to the power tracking error prediction model and the control variable change range; obtaining a wind power control strategy and an energy storage control strategy of the wind storage system according to the real-time state data and the optimal control model; wherein, obtaining the power tracking error prediction model according to the coefficient matrix of the discrete state space equation comprises: obtaining a state deviation matrix, a control deviation matrix and a disturbance deviation matrix according to the coefficient matrix of the discrete state space equation; obtaining the power tracking error prediction model according to the state deviation matrix, the control deviation matrix and the disturbance deviation matrix and combining state increments, control increments and disturbance increments; specifically, the power tracking error prediction model is: ， wherein, obtaining the optimal control model according to the power tracking error prediction model and the control variable change range comprises: ， ， ， ， In the formula, represents the deviation between the actual power and the target power at the k+1 moment in the future based on the current state at the k moment, , SOC is the energy storage adjustment coefficient, and P is the system demand power, is the state variable at time t, is the control input, is the disturbance input, A is the state variable coefficient, B is the control input coefficient, C is the diagonal matrix, and D is the disturbance input coefficient, is the state increment, is the control increment, that is, the change value of the energy storage power, the wind power and the VSG primary frequency modulation power control amount; is the disturbance increment, that is, the change value of the system frequency at the reference point and the wind speed; is the state deviation matrix, is the control deviation matrix, is the disturbance deviation matrix, p is the prediction time domain, q is the control time domain; obtaining a power deviation vector according to the power tracking error prediction model; obtaining a control constraint matrix and a control cost matrix according to the initial state space equation; obtaining the optimal control model according to the power deviation vector, the control constraint matrix and the control cost matrix and combining the control variable change range; specifically, the optimal control model is: determining the state variables, control inputs and disturbance inputs according to the real-time state data comprises: ， wherein denotes the optimal target, denotes the energy storage state of charge optimization target, denotes the power response optimization target, denotes the control cost optimization target, p is the prediction horizon, q is the control horizon, is the power deviation vector, Q is the control constraint matrix, is the control cost matrix, denotes the 2-norm of a vector L, denotes the k + i control at discrete time, denotes the transpose of u, denotes the k + i-1 control increment at discrete time, denotes the minimum value of the control increment, denotes the maximum value of the control increment, denotes the minimum value of the energy storage state of charge, denotes the maximum value of the energy storage state of charge, E denotes the system demand power, denotes the minimum value of the system demand power, denotes the maximum value of the system demand power, denotes the constraint condition.

2. The explicit model predictive control based frequency regulation optimization method for a networking wind storage system according to claim 1, wherein, taking the energy storage state of charge and the system demand power in the real-time state data as the state variables, taking the wind power, energy storage power and frequency modulation power as the control inputs, and taking the system frequency and wind speed as the disturbance inputs. obtaining the initial state space equation according to the action relationship of the state variables, control inputs and disturbance inputs on the wind storage system comprises:

3. The explicit model predictive control based networked wind storage system frequency regulation optimization method of claim 2, wherein, obtaining an updated state according to the state variables, control inputs, disturbance inputs and wind storage system device coefficients; obtaining a wind storage system output state according to the state variables; establishing the initial state space equation according to the updated state and the wind storage system output state. discretizing the initial state space equation to obtain a discrete state space equation comprises:

4. The explicit model predictive control based frequency regulation optimization method for networking wind storage system according to claim 1, wherein, discretizing the coefficient matrix of the initial state space equation according to a preset discrete time domain step to obtain the discrete state space equation. obtaining the wind power control strategy and the energy storage control strategy according to the real-time state data and the optimal control model comprises:

5. The explicit model predictive control based frequency regulation optimization method for networking wind storage system according to claim 1, wherein, obtaining state variables at each discrete point according to the discrete state space equation; inputting the real-time state data and the state variables at each discrete point into the optimal control model to obtain the wind power control strategy and the energy storage control strategy of the wind storage system. ​ In particular, given the system state variables x k , the explicit optimal control policy is solved offline by a receding horizon optimization strategy, following these steps:​ (1) in k acquiring system state x at k ); (2) Considering the discrete linear time-invariant system of discrete state space equation, the linear transfer property can be obtained: ， wherein, is the control quantity at the k+i-j-1 discrete time instant, is the disturbance quantity at the k+i-j-1 discrete time instant, with j = 0,..., i-1, denotes the matrix A discretized, denotes the matrix B discretized, denotes the matrix D discretized; The real-time state data and the state variable of each discrete point are input into the optimization control model, and the constraint condition of the linear transfer property of the offline system is added to obtain the expression: ， wherein, is the optimal target, x k is the initial state at the kth moment, k U k is the control sequence at the kth moment, represents U k is the matrix after discretization, represents the constraint condition;​​​​ (3) The first control quantity calculated is applied to the multi-objective optimization control of the frequency tracking of the wind storage system. (4) move the prediction horizon one step forward, and k repeat the above process at time +1, and thus obtain the optimal control strategy in the form of piecewise affine function: ， wherein, denotes the state feedback gain matrix, embodying the system states linear relationship of direct drive control variables; after completing a control region calculation, the system switches to the next region until the entire feasible solution space is covered.

Images