Pumped storage wind-solar-hydro microgrid reactive power coordinated control method and system

Abstract

The application discloses a kind of wind light water microgrid reactive power coordinated control method and system containing pumped storage, method constructs "model predictive control layer-improved chaos evolution optimization layer-equipment execution layer" three-layer collaborative control architecture: model predictive control layer is based on ultra-short-term prediction and voltage sensitivity model, the power feasible adjustment interval of each reactive power compensation device in future period is solved by rolling;Improved chaos evolution optimization layer is in each control cycle, with the feasible interval as hard constraint, using hot start, two-dimensional memristor hyperchaotic mapping and random crossover variation mechanism, quickly solve the optimal discrete switching of capacitor reactor and the optimal continuous reactive power output of pumped storage unit;Equipment execution layer issues instructions and collects operation data, realizes closed-loop correction.The application synchronously improves the voltage stability, reactive power economy and control response speed of microgrid under the premise of guaranteeing online real-time by hierarchical collaboration and series acceleration strategy.

Description

Technical Field

[0001] This invention belongs to the field of microgrid operation control and voltage reactive power optimization technology, and particularly relates to a method and system for reactive power coordinated control of wind, solar and hydropower microgrids including pumped storage. Background Technology

[0002] With the increasing penetration of intermittent renewable energy sources such as wind and solar power in microgrids, their output fluctuations pose a severe challenge to system voltage stability. To maintain voltage levels, microgrids typically employ various reactive power compensation devices: capacitor banks have the advantages of low cost and large capacity, but their switching is a discrete action, and their response speed is limited by mechanical switches; pumped storage units combine power generation and pumping functions, and their excitation systems can provide continuous and smooth reactive power regulation capabilities, but their response exhibits mechanical inertia (on the order of hundreds of milliseconds). These two types of devices differ significantly in response characteristics and control variable types, constituting a complex problem of discrete-continuous hybrid, multi-timescale coordinated control.

[0003] Existing control strategies are ineffective in addressing this problem: traditional droop control or proportional allocation strategies cannot achieve global optimization and are prone to causing conflicts between devices; centralized optimization using a single intelligent optimization algorithm (such as particle swarm optimization or genetic algorithm) can handle mixed variables, but the computation time is long and it is difficult to meet the real-time control requirements of microgrids at the second or even sub-second level; model predictive control (MPC) is good at handling dynamic systems and constraints at multiple time scales, but directly using MPC to solve optimization problems with a large number of discrete variables will result in mixed integer programming, which has high computational complexity.

[0004] Therefore, there is an urgent need for a new collaborative control method that can integrate the dynamic coordination advantages of MPC with the global optimization capability of intelligent optimization algorithms, and effectively ensure the real-time performance of online calculations, so as to achieve efficient, stable and economical operation of microgrids with heterogeneous reactive power compensation resources. Summary of the Invention

[0005] This invention aims to overcome the shortcomings of existing technologies and provide a reactive power coordination control method and system for wind-solar-hydro microgrids including pumped storage. The core concept of this invention is as follows: A hierarchical collaborative architecture is adopted. At the upper layer, Multi-Time-Scale Dynamic Coordination (MPC) is used to handle system-level multi-time-scale dynamic coordination problems and generate refined and compact feasible power ranges for equipment. At the lower layer, an improved chaotic evolutionary optimization algorithm is used to quickly solve the hybrid optimization problem of discrete switching of capacitors and reactors and continuous power output of pumped storage within the compressed search space provided by MPC. Through a series of engineering acceleration strategies, such as hot-start initialization, strong search space constraints, and linearization of fitness evaluation, it is ensured that the lower-level optimizer can complete the calculation within the control cycle of MPC, thereby balancing global optimization quality and online real-time performance.

[0006] In a first aspect, the present invention provides a reactive power coordinated control method for a wind-solar-hydro microgrid including pumped storage, comprising:

[0007] The ultra-short-term power output prediction data of wind and solar distributed power sources are obtained according to the preset wind and solar power output prediction model, and the real-time network status data is obtained according to the preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the goal of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve.

[0008] The discretized prediction model is solved in a rolling manner to output the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups.

[0009] Within each control cycle, the feasible adjustment range of the current output power is taken as a hard constraint. An improved chaotic evolutionary optimization algorithm is used to solve the optimal switching state of the capacitor reactor group and the optimal reactive power output of the pumped storage unit, forming an equipment-level collaborative control strategy.

[0010] According to the equipment-level collaborative control strategy, switching commands are issued to the switching devices of the capacitor and reactor group, and reactive power setpoint commands are issued to the excitation system of the pumped storage unit. Actual operating data of each device are collected at a frequency of not less than 1kHz.

[0011] Based on the actual operating data, the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm are updated and adaptively corrected online. The next round of reactive power coordination control is then carried out based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

[0012] Secondly, the present invention provides a reactive power coordinated control system for a wind-solar-hydro microgrid including pumped storage, comprising:

[0013] The acquisition module is configured to acquire ultra-short-term power output prediction data of wind and solar distributed power sources according to a preset wind and solar power output prediction model, and acquire real-time network status data according to a preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the goal of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve.

[0014] The output module is configured to solve the discretized prediction model in a rolling manner and output the reactive power reference trajectory and the corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups.

[0015] The solution module is configured to use the feasible adjustment range of the current output power as a hard constraint in each control cycle, and use an improved chaotic evolutionary optimization algorithm to solve the optimal switching state of the capacitor reactor group and the optimal reactive power output of the pumped storage unit, thus forming an equipment-level collaborative control strategy.

[0016] The data acquisition module is configured to issue switching commands to the switching device of the capacitor and reactor group and reactive power setpoint commands to the excitation system of the pumped storage unit according to the equipment-level collaborative control strategy, and to acquire the actual operating data of each device at a frequency of not less than 1kHz.

[0017] The update module is configured to update and adaptively correct the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm online based on the actual operating data, and to perform the next round of reactive power coordination control based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

[0018] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the reactive power coordination control method for a wind-solar-hydro microgrid containing pumped storage according to any embodiment of the present invention.

[0019] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the reactive power coordination control method for a wind-solar-hydro microgrid containing pumped storage according to any embodiment of the present invention.

[0020] This application discloses a reactive power coordinated control method and system for a wind-solar-hydro microgrid with pumped storage, applicable to microgrids containing pumped storage units, wind and solar power generation, and capacitor-reactor compensation devices. The method constructs a three-layer collaborative control architecture: a model predictive control layer, an improved chaotic evolutionary optimization layer, and an equipment execution layer. The model predictive control layer, based on ultra-short-term prediction and voltage sensitivity models, continuously solves the feasible power adjustment range of each reactive power compensation device in future time periods. The improved chaotic evolutionary optimization layer, within each control cycle, uses the feasible range as a hard constraint and employs hot start, two-dimensional memristor hyperchaotic mapping, and random crossover mutation mechanisms to quickly solve for the optimal discrete switching of capacitors and reactors and the optimal continuous reactive power output of pumped storage units. The equipment execution layer issues commands and collects operating data to achieve closed-loop correction. Through layered collaboration and a series of acceleration strategies, this invention simultaneously improves the voltage stability, reactive power economy, and control response speed of the microgrid while ensuring online real-time performance. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart of a reactive power coordinated control method for a wind-solar-hydro microgrid including pumped storage, provided as an embodiment of the present invention;

[0023] Figure 2 A structural block diagram of a reactive power coordinated control system for a wind-solar-hydro microgrid including pumped storage is provided in an embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Please see Figure 1 The diagram shows a flowchart of a reactive power coordination control method for a wind-solar-hydro microgrid containing pumped storage, as described in this application.

[0027] like Figure 1 As shown, the reactive power coordinated control method for wind-solar-hydro microgrids including pumped storage specifically includes the following steps:

[0028] Step S101: Obtain ultra-short-term power output prediction data of wind and solar distributed power sources according to the preset wind and solar power output prediction model, and obtain real-time network status data according to the preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, construct a discretized prediction model with the objectives of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve.

[0029] In this step, a voltage sensitivity matrix based on analytical methods is established to quantify the impact of reactive power injection on node voltage, namely:

[0030] ,

[0031] In the formula, for conjugate, Let be the voltage vector at node i. Let be the magnitude of the voltage vector at node i. The reactive power injected into node l Let be the voltage at node i;

[0032] The power control loop of the pumped storage unit is simplified to a first-order lag function, and the first reactive power output prediction model of the pumped storage unit is constructed, with the expression as follows:

[0033] ,

[0034] In the formula, For the predicted reactive power output of pumped storage units, The reactive power control time constant for pumped storage units. For the reactive power increment of pumped storage units, This is a reference value for the reactive power output increment of pumped storage units;

[0035] The power control loop of the capacitor-reactor bank is simplified to a first-order lag function, and a second reactive power output prediction model for the capacitor-reactor bank is constructed, expressed as:

[0036] ,

[0037] In the formula, The reactive power output of the capacitor reactor is predicted by the model. The reactive power control time constant of the capacitor-reactor bank. This represents the reactive power increment of the capacitor-reactor group. This is a reference value for the reactive power output increment of the capacitor-reactor group;

[0038] Based on the first reactive power output prediction model and the second reactive power output prediction model, a system state-space prediction model incorporating the dynamic characteristics of the pumped storage unit and the capacitor-reactor group is established, with the following expression:

[0039] ,

[0040] ,

[0041] ,

[0042] ,

[0043] ,

[0044] ,

[0045] In the formula, for Regarding the coefficient of time, This is the current state variable. To control variables, , , All are intermediate variables. This represents the reactive power output increment of the first group of capacitor reactors. This is the reference value for the reactive power output increment of the first group of capacitor reactors. This is the reference value for the reactive power output increment of the i-th group of capacitor reactors. This is a reference value for the reactive power output increment of pumped storage units. The reactive power control time constant of the capacitor-reactor bank. The reactive power control time constant for pumped storage units. It is the identity matrix;

[0046] Set the sampling period to The system state-space prediction model is discretized using the forward Euler method to obtain the discretized prediction model, which is expressed as follows:

[0047] ,

[0048] ,

[0049] ,

[0050] ,

[0051] In the formula, Let k+1 be the state variable. , , Let k be the state variable, control variable, and output variable at time k, respectively. The system matrix is ​​the discretized matrix. The input matrix is ​​discretized. The output matrix after discretization. The sampling period is It is the integral variable.

[0052] It should be noted that the objective function of the discretized prediction model, which aims to minimize the voltage deviation at the grid connection point and maximize the reactive power reserve, is:

[0053] ,

[0054] ,

[0055] ,

[0056] ,

[0057] In the formula, Let be the overall objective function. The objective function is to minimize the voltage deviation. The objective function is to maximize the reactive power reserve of the capacitor reactor. To predict the time domain, This is the reactive power output weighting coefficient for the capacitor reactor. The output increment of the capacitor-reactor at time k. For discrete time steps, The weighting factor for the grid connection point voltage. This represents the difference between the grid connection point voltage at time k and the reference value. The weighting coefficient for the target terminal voltage of pumped storage units. The difference between the pumped storage unit terminal voltage at time k and the reference value. The initial measured value of the grid connection point voltage. Let k be the increment of the reactive power of the pumped storage unit. This is the reference voltage value at the grid connection point. This represents the measured value of the pumped storage unit's terminal voltage at the initial moment. This is the reference value for the terminal voltage of the pumped storage unit. The voltage at the grid connection point. This refers to the terminal voltage of the pumped storage unit.

[0058] Specifically, the capacity constraint is expressed as:

[0059] ,

[0060] In the formula, for The reactive power output of the capacitor reactor at any given time; Provide the AVC with the output active power of the pumped storage power station units; for Pumped storage units constantly output reactive power; , These represent the minimum and maximum reactive power output of the capacitor reactor, respectively. , These are the minimum and maximum reactive power outputs of the pumped storage unit, respectively. , These are the minimum and maximum active power outputs of the unit, respectively.

[0061] The line power constraint is expressed as:

[0062] ,

[0063] In the formula: Let be the active power flowing through line w at time k; Let be the reactive power flowing through line w at time k; This represents the maximum apparent power that the line can transmit.

[0064] ,

[0065] In the formula, , These represent the minimum and maximum rates of change in the power generation of pumped storage units, respectively. , These represent the minimum and maximum rates of change of pumping power for pumped storage units, respectively. , These represent the power generation and pumping power of the pumped storage unit at time t, respectively.

[0066] Step S102: The discretized prediction model is solved in a rolling manner to output the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups.

[0067] In this step, the first control variable in the optimal control sequence obtained by solving the current control cycle, namely the reactive power setpoint command of the pumped storage unit and the switching command of the capacitor reactor group, are sent to the corresponding equipment respectively.

[0068] After waiting for a sampling period, new system state quantities are obtained through the data acquisition system. These system state quantities include the voltage of each node, the actual output of the pumped storage unit, and the actual state of the capacitor reactor.

[0069] Using the new system state variables as the initial state for the next control cycle optimization, the construction and solution of the discretized prediction model are repeated to form a rolling optimization process.

[0070] In step S103, within each control cycle, the feasible adjustment range of the current output power is used as a hard constraint, and an improved chaotic evolutionary optimization algorithm is used to solve for the optimal switching state of the capacitor reactor group and the optimal reactive power output value of the pumped storage unit, thus forming an equipment-level collaborative control strategy.

[0071] In this step, within each control cycle, the feasible adjustment range of the current output power is used as a hard constraint. An improved chaotic evolutionary optimization algorithm is used to solve for the optimal switching state of the capacitor reactor group and the optimal reactive power output of the pumped storage unit, thus forming an equipment-level collaborative control strategy.

[0072] The specific steps are as follows:

[0073] Step A: Perform population initialization and warm start. The optimal solution obtained in the previous control cycle is used as the core initial individual of the population in the current cycle, and the remaining individuals are randomly generated within the feasible region. This mechanism fully utilizes the strong correlation between optimization problems in adjacent control cycles, retains historical search information, significantly reduces the number of iterations required for convergence, and adapts to the real-time requirements of MPC fast rolling optimization.

[0074] Step B: Perform interval mapping and generate chaotic samples. In each iteration, two individuals are randomly selected and normalized to the domain of the chaotic system using a linear mapping, ensuring that their value range meets the requirements of the chaotic iteration. Subsequently, multiple sets of chaotic sequences are generated using a two-dimensional exponential discrete memristor hyperchaotic mapping. This mapping has two positive Lyapunov exponents, exhibiting strong randomness, unpredictability, and ergodicity, providing rich exploration directions for optimization searches.

[0075] Step C: Calculate the evolutionary direction and perform mutation operations. Based on the differences between the generated chaotic samples and the current individual, multiple evolutionary directions are calculated. The algorithm randomly selects a mutation strategy based on the current individual or the global optimal solution with a preset probability, and combines it with a randomly generated step size factor to generate mutated individuals.

[0076] Step D: Perform crossover and selection operations. A binomial crossover mechanism is used to merge mutated individuals with original individuals to generate experimental individuals. During the crossover process, the crossover rate is randomly generated in each iteration, further enhancing the diversity of the search and avoiding premature convergence of the algorithm.

[0077] Step E: Repeat steps B through D until the maximum number of evaluations is reached or the convergence condition is met. Output the optimal solution for the current cycle, i.e., the optimal discrete switching state of the capacitor bank and the optimal continuous reactive power output of the pumped storage unit.

[0078] Specifically, the population initialization and warm-start settings are as follows: Population size Np = 50, number of chaotic samples N = 5, maximum function evaluation times MaxFES = 10000, and the chaos parameter is fixed at k = 2.66. =1.0, random crossover rate and step size factor In each iteration, individuals are uniformly and randomly generated from the interval [0,1]. A warm-start mechanism is adopted, using the optimal solution obtained in the previous MPC control cycle as the core initial individual of the current population, and the remaining individuals are randomly generated within the feasible region to utilize historical optimization information and accelerate algorithm convergence.

[0079] Interval mapping and chaotic sequence generation. To adapt the actual decision variables to the domain of the chaotic system, interval mapping is required. In each iteration, two sequences are randomly selected from the population. and Map it to the domain of a chaotic system:

[0080] ,

[0081] In the formula, lb and ub are the lower and upper bounds of the decision variable, respectively.

[0082] Generating chaotic sequences using a two-dimensional exponential discrete memristor hyperchaotic map:

[0083] ,

[0084] In the formula, and For the state variable at the next moment, For chaotic mapping parameters, and Let be the chaotic state variable, and e be the natural constant.

[0085] N chaotic samples are generated through this mapping, and then inversely mapped back to the decision space to obtain... and , and These correspond to the values ​​of the two chaotic samples in the decision space.

[0086] Evolutionary direction calculation and mutation operation. Calculating the evolutionary direction based on the difference between chaotic samples and the current individual:

[0087] ,

[0088] In the formula, n = 1, 2, ..., N, and Individuals and individuals The nth evolutionary direction.

[0089] The algorithm selects one of two mutation strategies with a probability p=0.5. The first strategy is based on the mutation of the current individual:

[0090] ,

[0091] The second type is mutation based on the global optimal solution:

[0092] ,

[0093] In the formula, Step size factor This is the current globally optimal solution. and These are the two new individuals obtained after the mutation operation. for The first chaotic sample generated, for The first chaotic sample generated;

[0094] Crossover and selection operations. The crossover operation uses a binomial crossover method to merge the mutated individuals with the original individuals. For each dimension j, the generation rule for the experimental individuals is:

[0095] ,

[0096] In the formula, for The components of the experimental individual in the j-th dimension, for The components of the variant individuals in the j-th dimension, for The component of the original individual in the j-th dimension, The numbers are uniformly random numbers in the range [0,1]. For random crossover rate, The dimension index is randomly selected.

[0097] Selection and Population Update. The fitness of the trial individuals is compared with that of the original individuals using a greedy criterion to update the population.

[0098] ,

[0099] In the formula, This is the fitness function, used to evaluate the quality of the solution.

[0100] Iteration Termination and Optimal Output. Repeat the iteration until the maximum number of evaluations is reached or the convergence condition is met. Output the optimal solution for the current period, i.e., the optimal discrete switching state of the capacitor bank and the optimal continuous reactive power output of the pumped storage unit.

[0101] Step S104: According to the equipment-level collaborative control strategy, a switching command is issued to the switching device of the capacitor reactor group, a reactive power setpoint command is issued to the excitation system of the pumped storage unit, and the actual operating data of each device is collected at a frequency of not less than 1kHz.

[0102] In this step, according to the equipment-level collaborative control strategy, switching commands are issued to the switching device of the capacitor reactor group, and reactive power setpoint commands are issued to the excitation system of the pumped storage unit. The actual operating data of each device is collected at a frequency of not less than 1kHz.

[0103] Specifically, the equipment execution layer performs anti-jitter filtering on the capacitor and reactor switching commands. When the switching commands of adjacent control cycles reverse, a one-cycle delay is automatically inserted before execution to avoid frequent equipment operation caused by minor fluctuations in commands or measurement noise, thereby extending the mechanical life of the equipment.

[0104] Step S105: Based on the actual operating data, the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm are updated and adaptively corrected online, and the next round of reactive power coordination control is carried out based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

[0105] Specifically, the system includes periodic model updates: an online incremental learning process is performed every 15 minutes using the latest operational data to predict the ultra-short-term wind and solar power output; sensitivity matrix updates: the voltage sensitivity matrix is ​​recalculated every 5 minutes or when the network topology changes significantly; and adaptive algorithm parameters: based on the historical records of recent optimization results, the random crossover rate, step size factor, and number of chaotic samples in the improved chaotic evolutionary optimization algorithm are dynamically adjusted to maintain a balance between exploration and development and adapt to changes in the system's operating state.

[0106] In summary, the method of this application, firstly, at the system-level dynamic coordination level, establishes a discretized prediction model based on wind and solar ultra-short-term output prediction data and voltage sensitivity matrix, aiming to minimize grid connection point voltage deviation and maximize reactive power reserve. Through rolling optimization, it generates refined power feasible adjustment ranges for each reactive power compensation device over multiple future control cycles. This mechanism effectively quantifies the impact of reactive power injection on node voltage and fully considers the dynamic characteristic differences between the millisecond-level continuous response of pumped storage units and the discrete switching of capacitors and reactors. It defines the optimized dynamic feasible domain at the system level, providing a compact and reasonable search space for lower-level optimization, avoiding conflicts between device actions, and significantly improving the collaborative control accuracy across multiple time scales.

[0107] Secondly, at the mixed-variable optimization solution level, the improved chaotic evolutionary optimization layer, within the compressed search space provided by the model predictive control layer, uses the power feasible interval as a hard constraint and integrates hot-start initialization, two-dimensional exponential discrete memristor hyperchaotic mapping, and random crossover and mutation mechanisms to quickly solve for the optimal discrete switching state of the capacitor-reactor group and the optimal continuous reactive power output of the pumped storage unit. Specifically, the hot-start mechanism fully utilizes the strong correlation between optimization problems in adjacent control cycles, retains historical search information, and significantly reduces the number of iterations required for convergence; the two-dimensional memristor hyperchaotic mapping has two positive Lyapunov exponents, exhibiting strong randomness, unpredictability, and ergodicity, providing rich exploration directions for optimization search; the dynamic generation of random crossover rate and variable asynchronous long factor enhances population diversity and avoids premature convergence. The organic combination of these strategies fundamentally solves the bottleneck of long computation time and difficulty in online application of traditional centralized optimization algorithms, achieving millisecond-level fast optimization while ensuring solution accuracy, meeting the real-time control requirements of microgrids.

[0108] Furthermore, at the closed-loop correction and adaptive level, the equipment execution layer collects actual operating data of each device at a high frequency of no less than 1kHz, and introduces anti-jitter filtering processing for capacitor and reactor switching commands—when the switching commands of adjacent control cycles reverse, a delay is automatically inserted before execution, effectively suppressing frequent equipment actions caused by measurement noise or small fluctuations in commands, and significantly extending the mechanical life of the equipment. The closed-loop update and correction module performs periodic online incremental learning of the wind and solar power output prediction model based on the feedback of actual operating data, recalculates the voltage sensitivity matrix triggered by topological changes, and dynamically adjusts the random crossover rate, step size factor, and number of chaotic samples of the improved chaotic evolution optimization algorithm based on historical effects. This achieves real-time correction of model deviations and accurate tracking of operating status, ensuring the robustness and adaptability of the control strategy under different operating conditions.

[0109] In summary, this invention, through the organic combination of a hierarchical collaborative architecture and a series of acceleration strategies, simultaneously improves the voltage stability, reactive power regulation economy, and control response speed of microgrids while ensuring the online real-time performance of the algorithm: voltage deviation is effectively suppressed, reactive power reserve capacity is maximized, the number of equipment actions is significantly reduced, and control command response time is compressed to the millisecond level. Ultimately, it provides an efficient, reliable, and engineerable solution for the safe, stable, and economical operation of microgrids containing pumped storage and distributed wind and solar power.

[0110] Please see Figure 2 The diagram shows a structural block diagram of a reactive power coordinated control system for a wind-solar-hydro microgrid containing pumped storage, as described in this application.

[0111] like Figure 2 As shown, the reactive power coordinated control system 200 for wind, solar and hydro microgrids includes an acquisition module 210, an output module 220, a solution module 230, a data acquisition module 240, and an update module 250.

[0112] The acquisition module 210 is configured to acquire ultra-short-term power output prediction data of wind and solar distributed power sources according to a preset wind and solar power output prediction model, and acquire real-time network status data according to a preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the objectives of minimizing the voltage deviation at the grid connection point and maximizing reactive power reserve. The output module 220 is configured to solve the discretized prediction model in a rolling manner, and output the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped switching capacitor reactor groups. The solution module 230 is configured to use the currently output feasible power adjustment range as a hard constraint in each control cycle. The system employs an improved chaotic evolutionary optimization algorithm to solve for the optimal switching state of the capacitor bank and the optimal reactive power output of the pumped storage unit, forming an equipment-level collaborative control strategy. The acquisition module 240 is configured to issue switching commands to the capacitor bank's switching device and reactive power setpoint commands to the pumped storage unit's excitation system according to the equipment-level collaborative control strategy, and to acquire actual operating data of each device at a frequency of not less than 1 kHz. The update module 250 is configured to update and adaptively correct the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolutionary optimization algorithm online based on the actual operating data, and to perform the next round of reactive power coordination control based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolutionary optimization algorithm.

[0113] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0114] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the reactive power coordination control method for a wind-solar-hydro microgrid containing pumped storage in any of the above method embodiments.

[0115] In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows:

[0116] The ultra-short-term power output prediction data of wind and solar distributed power sources are obtained according to the preset wind and solar power output prediction model, and the real-time network status data is obtained according to the preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the goal of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve.

[0117] The discretized prediction model is solved in a rolling manner to output the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups.

[0118] Within each control cycle, the feasible adjustment range of the current output power is taken as a hard constraint. An improved chaotic evolutionary optimization algorithm is used to solve the optimal switching state of the capacitor reactor group and the optimal reactive power output of the pumped storage unit, forming an equipment-level collaborative control strategy.

[0119] According to the equipment-level collaborative control strategy, switching commands are issued to the switching devices of the capacitor and reactor group, and reactive power setpoint commands are issued to the excitation system of the pumped storage unit. Actual operating data of each device are collected at a frequency of not less than 1kHz.

[0120] Based on the actual operating data, the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm are updated and adaptively corrected online. The next round of reactive power coordination control is then carried out based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

[0121] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of a reactive power coordinated control system for a wind-solar-hydro microgrid containing pumped storage. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely disposed relative to a processor, which can be connected to the reactive power coordinated control system for a wind-solar-hydro microgrid containing pumped storage via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0122] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3 Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the reactive power coordinated control method for a wind-solar-hydro microgrid containing pumped storage, as described in the above method embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the reactive power coordinated control system for a wind-solar-hydro microgrid containing pumped storage. The output device 340 may include a display screen or other display device.

[0123] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0124] In one implementation, the aforementioned electronic device is applied in a reactive power coordinated control system for a wind-solar-hydro microgrid containing pumped storage, serving as a client, and includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0125] The ultra-short-term power output prediction data of wind and solar distributed power sources are obtained according to the preset wind and solar power output prediction model, and the real-time network status data is obtained according to the preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the goal of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve.

[0126] The discretized prediction model is solved in a rolling manner to output the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups.

[0127] Within each control cycle, the feasible adjustment range of the current output power is taken as a hard constraint. An improved chaotic evolutionary optimization algorithm is used to solve the optimal switching state of the capacitor reactor group and the optimal reactive power output of the pumped storage unit, forming an equipment-level collaborative control strategy.

[0128] According to the equipment-level collaborative control strategy, switching commands are issued to the switching devices of the capacitor and reactor group, and reactive power setpoint commands are issued to the excitation system of the pumped storage unit. Actual operating data of each device are collected at a frequency of not less than 1kHz.

[0129] Based on the actual operating data, the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm are updated and adaptively corrected online. The next round of reactive power coordination control is then carried out based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

[0130] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A reactive power coordinated control method for a wind-solar-hydro microgrid including pumped storage, characterized in that, include: The ultra-short-term power output prediction data of wind and solar distributed power sources are obtained according to the preset wind and solar power output prediction model, and real-time network status data are obtained according to the preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the goal of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve. Specifically, the power control loops of the pumped storage unit and the grouped switching capacitor and reactor group are simplified to first-order lag functions. A system state-space prediction model containing the dynamic characteristics of the pumped storage unit and the dynamic characteristics of the capacitor and reactor group is constructed. The system state-space prediction model is then discretized using the forward Euler method to obtain the discretized prediction model. The discretized prediction model is solved in a rolling manner to output the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups. Within each control cycle, the feasible adjustment range of the current output power is used as a hard constraint. An improved chaotic evolutionary optimization algorithm is employed to solve for the optimal switching state of the capacitor bank and the optimal reactive power output of the pumped storage unit, forming an equipment-level collaborative control strategy, specifically including: Step A: Perform population initialization and hot start. The optimal solution obtained in the previous control cycle is used as the core initial individual of the population in the current cycle, and the remaining individuals are randomly generated within the feasible region. Step B: Perform interval mapping and chaotic sample generation. Use two-dimensional exponential discrete memristor hyperchaotic mapping to generate multiple sets of chaotic sequences, providing exploration directions for optimization search; Step C: Perform evolution direction calculation and mutation operation. Calculate the evolution direction based on the difference between the chaotic sample and the current individual, and randomly select a mutation strategy based on the current individual or the global optimal solution with a preset probability to generate mutated individuals. Step D: Perform crossover and selection operations, using a binomial crossover mechanism to fuse the mutated individuals with the original individuals to generate experimental individuals; Step E: Repeat steps B to D until the maximum number of evaluations is reached or the convergence condition is met, and output the optimal solution for the current period. According to the equipment-level collaborative control strategy, switching commands are issued to the switching devices of the capacitor and reactor group, and reactive power setpoint commands are issued to the excitation system of the pumped storage unit. Actual operating data of each device are collected at a frequency of not less than 1kHz. Based on the actual operating data, the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm are updated and adaptively corrected online. The next round of reactive power coordination control is then carried out based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

2. The reactive power coordinated control method for a wind-solar-hydro microgrid including pumped storage as described in claim 1, characterized in that, The discretized prediction model constructed based on the ultra-short-term power output prediction data and the real-time network status data, with the objectives of minimizing grid connection point voltage deviation and maximizing reactive power reserve, includes: The power control loop of the pumped storage unit is simplified to a first-order lag function, and the first reactive power output prediction model of the pumped storage unit is constructed, with the expression as follows: ， In the formula, For the predicted reactive power output of pumped storage units, The reactive power control time constant for pumped storage units. For the reactive power increment of pumped storage units, This is a reference value for the reactive power output increment of pumped storage units; The power control loop of the capacitor-reactor bank is simplified to a first-order lag function, and a second reactive power output prediction model for the capacitor-reactor bank is constructed, expressed as: ， In the formula, The reactive power output of the capacitor reactor is predicted by the model. The reactive power control time constant of the capacitor-reactor bank. This represents the reactive power increment of the capacitor-reactor group. This is a reference value for the reactive power output increment of the capacitor-reactor group; Based on the first reactive power output prediction model and the second reactive power output prediction model, a system state-space prediction model incorporating the dynamic characteristics of the pumped storage unit and the capacitor-reactor group is established, with the following expression: ， ， ， ， ， ， In the formula, for Regarding the coefficient of time, This is the current state variable. To control variables, , , All are intermediate variables. This represents the reactive power output increment of the first group of capacitor reactors. This is the reference value for the reactive power output increment of the first group of capacitor reactors. This is the reference value for the reactive power output increment of the i-th group of capacitor reactors. This is a reference value for the reactive power output increment of pumped storage units. The reactive power control time constant of the capacitor-reactor bank. The reactive power control time constant for pumped storage units. It is the identity matrix; Set the sampling period to The system state-space prediction model is discretized using the forward Euler method to obtain the discretized prediction model, which is expressed as follows: ， ， ， ， In the formula, Let k+1 be the state variable. , , Let k be the state variable, control variable, and output variable at time k, respectively. The system matrix is ​​the discretized matrix. The input matrix is ​​discretized. The output matrix after discretization. The sampling period is It is the integral variable.

3. The reactive power coordinated control method for a wind-solar-hydro microgrid including pumped storage as described in claim 1, characterized in that, in, The objective function of the discretized prediction model, which aims to minimize the voltage deviation at the grid connection point and maximize the reactive power reserve, is: ， ， ， ， In the formula, Let be the overall objective function. The objective function is to minimize the voltage deviation. The objective function is to maximize the reactive power reserve of the capacitor reactor. To predict the time domain, This is the reactive power output weighting coefficient for the capacitor reactor. The output increment of the capacitor-reactor at time k. For discrete time steps, The weighting factor for the grid connection point voltage. This represents the difference between the grid connection point voltage at time k and the reference value. The weighting coefficient for the target terminal voltage of pumped storage units. The difference between the pumped storage unit terminal voltage at time k and the reference value. The initial measured value of the grid connection point voltage. Let k be the increment of the reactive power of the pumped storage unit. This is the reference voltage value at the grid connection point. This represents the measured value of the pumped storage unit's terminal voltage at the initial moment. This is the reference value for the terminal voltage of the pumped storage unit. The voltage at the grid connection point. This refers to the terminal voltage of the pumped storage unit.

4. The reactive power coordinated control method for a wind-solar-hydro microgrid including pumped storage as described in claim 1, characterized in that, The rolling solution of the discretized prediction model outputs the reactive power reference trajectory and corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles, including: The first control variable in the optimal control sequence obtained by solving the current control cycle, namely the reactive power setpoint command of the pumped storage unit and the switching command of the capacitor and reactor group, is sent to the corresponding equipment respectively. After waiting for a sampling period, new system state quantities are obtained through the data acquisition system. These system state quantities include the voltage of each node, the actual output of the pumped storage unit, and the actual state of the capacitor reactor. Using the new system state variables as the initial state for the next control cycle optimization, the construction and solution of the discretized prediction model are repeated to form a rolling optimization process.

5. A reactive power coordinated control system for a wind-solar-hydro microgrid including pumped storage, characterized in that, include: The acquisition module is configured to acquire ultra-short-term power output prediction data of wind and solar distributed power sources according to a preset wind and solar power output prediction model, and acquire real-time network status data according to a preset voltage sensitivity matrix. Based on the ultra-short-term power output prediction data and the real-time network status data, a discretized prediction model is constructed with the objectives of minimizing the voltage deviation at the grid connection point and maximizing the reactive power reserve. Specifically, the power control loops of the pumped storage unit and the grouped switching capacitor and reactor group are simplified to first-order lag functions. A system state-space prediction model containing the dynamic characteristics of the pumped storage unit and the dynamic characteristics of the capacitor and reactor group is constructed. The system state-space prediction model is discretized by the forward Euler method to obtain the discretized prediction model. The output module is configured to solve the discretized prediction model in a rolling manner and output the reactive power reference trajectory and the corresponding feasible power adjustment range of each reactive power compensation device in multiple future control cycles. The reactive power compensation device includes pumped storage units and grouped capacitor reactor groups. The solution module is configured to use the feasible adjustment range of the current output power as a hard constraint in each control cycle, and employ an improved chaotic evolutionary optimization algorithm to solve for the optimal switching state of the capacitor bank and the optimal reactive power output of the pumped storage unit, forming an equipment-level collaborative control strategy, specifically including: Step A: Perform population initialization and hot start. The optimal solution obtained in the previous control cycle is used as the core initial individual of the population in the current cycle, and the remaining individuals are randomly generated within the feasible region. Step B: Perform interval mapping and chaotic sample generation. Use two-dimensional exponential discrete memristor hyperchaotic mapping to generate multiple sets of chaotic sequences, providing exploration directions for optimization search; Step C: Perform evolution direction calculation and mutation operation. Calculate the evolution direction based on the difference between the chaotic sample and the current individual, and randomly select a mutation strategy based on the current individual or the global optimal solution with a preset probability to generate mutated individuals. Step D: Perform crossover and selection operations, using a binomial crossover mechanism to fuse the mutated individuals with the original individuals to generate experimental individuals; Step E: Repeat steps B to D until the maximum number of evaluations is reached or the convergence condition is met, and output the optimal solution for the current period. The data acquisition module is configured to issue switching commands to the switching device of the capacitor and reactor group and reactive power setpoint commands to the excitation system of the pumped storage unit according to the equipment-level collaborative control strategy, and to acquire the actual operating data of each device at a frequency of not less than 1kHz. The update module is configured to update and adaptively correct the parameters of the wind and solar power output prediction model, the voltage sensitivity matrix, and the improved chaotic evolution optimization algorithm online based on the actual operating data, and to perform the next round of reactive power coordination control based on the updated parameters of the wind and solar power output prediction model and the updated improved chaotic evolution optimization algorithm.

6. An electronic device, characterized in that, include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method described in any one of claims 1 to 4.

Images