Control method and device of pumped storage power station, storage medium and program product

Abstract

The embodiment of the invention discloses a control method and device of a pumped storage power station, a storage medium and a program product. The method comprises the following steps: acquiring current operation data of the pumped storage power station in real time; the current operation data is preprocessed; an ESNN model is constructed; based on the preprocessed current operation data and the ESNN model, an opening degree gear is obtained; and controlling a speed regulator to adjust the guide vane opening of the pumped storage unit based on the opening gear. According to the method, the opening degree gear can be automatically adjusted according to the real-time condition of the pumped storage unit, then the speed regulator is controlled to adjust the guide vane opening degree of the pumped storage unit, the flexible control requirement of the pumped storage power station is met, and the operation reliability and control precision of the pumped storage power station are improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of pumped - storage power generation, and particularly to a control method, device, storage medium, and program product for a pumped - storage power station. Background Art

[0002] With the continuous expansion of the scale of the power system and the rapid development of new energy, the stability and economy of the power system are facing unprecedented challenges. As one of the important peak - shaving power sources in the power system, pumped - storage power stations undertake the important tasks of balancing the grid load and improving system stability. However, the existing pumped - storage AGC control technology usually adopts the traditional PID (Proportional - Integral - Derivative) control algorithm, whose parameters are difficult to optimize, and it is difficult to adapt to the rapidly changing grid conditions and the volatility of new - energy output, resulting in low control accuracy and poor adaptability under complex working conditions. Moreover, the access of new energy puts higher requirements on the speed and robustness of the AGC control system. The existing control strategies are difficult to effectively respond to the rapid changes in new - energy output, affecting the stable operation of the power grid. With the aging of equipment and the change of operating conditions such as water head, the characteristics of pumped - storage units will also change accordingly, which requires the AGC control system to be able to adaptively adjust the control strategy to maintain the efficient operation of pumped - storage units and the stability of the power grid. Summary of the Invention

[0003] In view of this, the embodiments of the present disclosure provide a control method, device, storage medium, and program product for a pumped - storage power station, which can automatically adjust the opening position according to the real - time situation of the pumped - storage unit, and then control the governor to adjust the guide - vane opening of the pumped - storage unit to meet the flexible control requirements of the pumped - storage power station.

[0004] In a first aspect, the embodiments of the present disclosure provide a control method for a pumped - storage power station, adopting the following technical solution:

[0005] Obtain the current operation data of the pumped - storage power station in real time;

[0006] Pre - process the current operation data;

[0007] Construct an ESNN model;

[0008] Based on the pre - processed current operation data and the ESNN model, obtain the opening position;

[0009] Based on the opening position, control the governor to adjust the guide - vane opening of the pumped - storage unit.

[0010] Optionally, the current operation data includes at least one operation parameter among the current power of the pumped-storage unit, the water head at the inlet, the water head at the outlet, the output command, the AGC power command, the control command at the previous moment, the flow rate of the penstock, and the efficiency of the pumped-storage unit.

[0011] Optionally, the preprocessing includes data cleaning and data normalization.

[0012] Optionally, the ESNN model includes an input layer, a hidden layer, and an output layer;

[0013] The input layer receives the operation parameters, converts the operation parameters into pulse signals of the input layer pulse neurons; transmits the pulse signals of the input layer pulse neurons to the hidden layer; wherein, each operation parameter corresponds to a group of input layer pulse neurons, and the firing frequency of the pulse signals of each group of input layer pulse neurons changes proportionally according to the magnitude of the operation parameter;

[0014] The hidden layer adopts a three-layer recursive structure, and each layer of the recursive structure contains a preset number of hidden layer pulse neurons; based on the pulse signals of the input layer pulse neurons and the hidden layer state at the previous moment, generates the pulse signals of the hidden layer pulse neurons; transmits the pulse signals of the hidden layer pulse neurons to the output layer;

[0015] The output layer includes multiple output layer pulse neurons, and each output layer pulse neuron corresponds to a fixed gear; based on the pulse signals of the hidden layer pulse neurons, generates the pulse signals of the output layer pulse neurons; counts the number of times the pulse signals of each output layer pulse neuron are sent within the sampling period; based on multiple fixed gears and multiple sending times, obtains the opening gear.

[0016] Optionally, the control method of the pumped-storage power station further includes:

[0017] Randomly generate a population, set the maximum number of evolutionary generations, the population includes M original individuals, each original individual contains a preset number of genes, each gene represents a parameter, and the parameter is represented by a floating point number;

[0018] In each iteration, configure the M original individuals into the ESNN model for simulation testing, and calculate the fitness of the M original individuals;

[0019] Randomly select k parent individuals from the M original individuals, among the k parent individuals, screen out the parent individual with the maximum fitness, and retain the remaining k - 1 parent individuals;

[0020] Judge whether the number of parent individuals is equal to M;

[0021] If it is equal to M, then according to the crossover probability g, randomly divide the M parent individuals into two groups. The first group of individuals includes g×M parent individuals, and the second group of individuals includes (1 - g)×M parent individuals;

[0022] If it is not equal to M, then continue to randomly select k parent individuals from the M original individuals;

[0023] Use the simulated binary crossover operator to randomly cross the parent individuals in the first group to generate 2×(g×M) offspring individuals;

[0024] According to the mutation probability, perturb the genes of the (1 - g)×M parent individuals and the 2×(g×M) offspring individuals to obtain 2×M mutant individuals;

[0025] Configure the 2×M mutant individuals into the ESNN model for simulation testing, and calculate the fitness of the 2×M mutant individuals;

[0026] Sort the 2×M mutant individuals in ascending order of fitness, and select the first M mutant individuals as the new M original individuals;

[0027] When the maximum number of generations of evolution is reached or within consecutive L iteration cycles, if the overall fitness improvement of the population is not greater than the preset amplitude threshold, configure the original individual with the minimum fitness among the new M original individuals into the ESNN model.

[0028] Optionally, the obtaining of the opening position includes:

[0029] Based on the multiple transmission times and the weights of the output layer pulse neurons preset, obtain the weighted average number of times;

[0030] Judge whether the weighted average number of times is equal to the average value of any two transmission times;

[0031] If so, obtain the two fixed positions corresponding to the two transmission times, and the opening position is equal to the average value of the two fixed positions;

[0032] If not, obtain the transmission time closest to the weighted average number of times, denoted as the closest transmission time;

[0033] The opening position is equal to the fixed position corresponding to the closest transmission time.

[0034] Optionally, the calculation formula for the weighted average number of times is:

[0035]

[0036]

[0037] Among them, QP is the number of weighted average times; QH is the number of weighted pulses; QZ is the total weight of the multiple output layer pulse neurons; i is the serial number of the output layer pulse neuron; I is the total number of output layer pulse neurons; Q i is the weight of the i-th output layer pulse neuron; M i is the number of transmissions of the pulse signal of the i-th output layer pulse neuron within the sampling period.

[0038] Optionally, the control method of the pumped storage power station further includes:

[0039] Setting the key components of the pumped storage power station as a plurality of visualization components, and storing the plurality of visualization components in a preset component library;

[0040] Connecting the plurality of visualization components by using logical connections to form a visualization configuration editor;

[0041] The visualization configuration editor is used to control the operation of the pumped storage power station;

[0042] Among them, the key components include at least one of a pumped storage unit, a governor, and an ESNN controller, and the ESNN controller is encapsulated by the ESNN model.

[0043] In a second aspect, an embodiment of the present disclosure further provides a control system for a pumped storage power station, adopting the following technical solution:

[0044] A data acquisition module, configured to acquire the current operation data of the pumped storage power station in real time;

[0045] A preprocessing module, configured to preprocess the current operation data;

[0046] A model construction module, configured to construct an ESNN model;

[0047] A gear position acquisition module, configured to acquire an opening gear position based on the preprocessed current operation data and the ESNN model;

[0048] A unit adjustment module, configured to control the governor to adjust the guide vane opening of the pumped storage unit based on the opening gear position.

[0049] In a third aspect, an embodiment of the present disclosure further provides a computer device, adopting the following technical solution:

[0050] The computer device includes:

[0051] At least one processor; and,

[0052] A memory communicatively connected to the at least one processor; wherein,

[0053] The memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the control method of the pumped-storage power station described in any one of the above.

[0054] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium storing computer instructions for causing a computer to execute the control method of the pumped-storage power station described in any one of the above.

[0055] In a fifth aspect, an embodiment of the present disclosure further provides a computer program product including computer programs / instructions, and when the computer programs / instructions are executed by a processor, the steps of the method described in any one of the above are implemented.

[0056] The control method of the pumped-storage power station provided by the embodiment of the present disclosure preprocesses by obtaining the current operation data of the pumped-storage power station in real time, and provides input for the constructed ESNN model in a timely manner, so that the ESNN model can quickly process the input data and generate opening positions, and the governor can be controlled according to the opening positions to adjust the guide vane opening of the pumped-storage unit, realizing the adaptive adjustment control of the pumped-storage power station. Through the intelligent processing ability of the ESNN model, this method can more accurately predict and respond to changes in grid demand, improve the output regulation accuracy of the pumped-storage power station, and the real-time processing method can shorten the delay from data acquisition to execution of control actions, improving the overall response speed of intelligent control. Adjusting the guide vane opening of the pumped-storage unit according to the current operation data, thereby adjusting the output of the pumped-storage unit, can optimize the use of water resources and electric energy, improve energy utilization efficiency, help maintain the stability of the grid frequency, and enhance the adaptability of the grid to load fluctuations. This automated control method can reduce the dependence on operators, reduce the possibility of human operation errors, improve the reliability and control accuracy of the operation of the pumped-storage power station, have good adaptability to various complex working conditions, and can avoid frequent start-stop and large-scale adjustments, reducing the mechanical stress on the pumped-storage unit and the governor, and helping to extend the service life of the equipment. This intelligent control method helps to reduce ineffective energy consumption, improve the operation efficiency of the power station, achieve energy conservation and emission reduction, provide better peak shaving and frequency support for the access of new energy (such as wind energy, solar energy), and help to build a diversified power supply system.

[0057] The above description is only an overview of the technical solutions of the present disclosure. In order to understand the technical means of the present disclosure more clearly, it can be implemented according to the content of the description. In order to make the above and other purposes, features and advantages of the present disclosure more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the drawings, is described in detail as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present disclosure. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0059] Figure 1 It is a schematic flowchart of the control of the pumped storage power station provided by the embodiment of the present disclosure;

[0060] Figure 2 It is a schematic flowchart of the ESNN model training method provided by the embodiment of the present disclosure;

[0061] Figure 3 It is a schematic flowchart of the opening position obtaining method provided by the embodiment of the present disclosure;

[0062] Figure 4 It is a schematic block diagram of the control system of the pumped storage power station provided by the embodiment of the present disclosure;

[0063] Figure 5 It is a schematic structural diagram of a computer device provided by the embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] The embodiments of the present disclosure will be described in detail below with reference to the drawings.

[0065] It should be clear that the embodiments of the present disclosure are illustrated by specific specific examples below. Those skilled in the art can easily understand other advantages and effects of the present disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present disclosure without creative efforts belong to the scope of protection of the present disclosure.

[0066] Note that the following description pertains to various aspects of embodiments within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of the aspects set forth herein can be used to implement a device and / or practice a method. Additionally, this device can be implemented and this method can be practiced using other structures and / or functionality in addition to one or more of the aspects set forth herein.

[0067] It should also be noted that the diagrams provided in the following embodiments merely illustrate the basic concept of the present disclosure schematically. Only the components related to the present disclosure are shown in the diagrams, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0068] In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the aspects described can be practiced without these specific details.

[0069] Referring to Figure 1 , the present disclosure provides a control method for a pumped-storage power station, including the following steps:

[0070] S1: Obtain the current operating data of the pumped-storage power station in real time;

[0071] S2: Preprocess the current operating data;

[0072] S3: Construct an ESNN model;

[0073] S4: Obtain the opening position based on the preprocessed current operating data and the ESNN model;

[0074] S5: Control the governor based on the opening position to adjust the guide vane opening of the pumped-storage unit.

[0075] The control method of the pumped-storage power station disclosed in the present disclosure preprocesses the currently operating data of the pumped-storage power station obtained in real time to provide input to the constructed ESNN model in a timely manner, enabling the ESNN model to quickly process the input data and generate the opening position. According to the opening position, the governor can be controlled to adjust the guide vane opening of the pumped-storage unit, realizing the adaptive adjustment control of the pumped-storage power station. Through the intelligent processing ability of the ESNN model, this method can more accurately predict and respond to changes in grid demand, improve the accuracy of the output regulation of the pumped-storage power station, and the real-time processing method can shorten the delay from data acquisition to the execution of control actions, enhancing the overall response speed of intelligent control. Adjusting the guide vane opening of the pumped-storage unit according to the currently operating data, thereby adjusting the output of the pumped-storage unit, can optimize the use of water resources and electric energy, improve energy utilization efficiency, contribute to maintaining the stability of the grid frequency, and enhance the adaptability of the grid to load fluctuations.

[0076] In summary, this automated control method can reduce the dependence on operators, reduce the possibility of human operation errors, improve the reliability and control accuracy of the operation of the pumped-storage power station, has good adaptability to various complex working conditions, and can avoid frequent start-stop and large-scale adjustments, reducing the mechanical stress on the pumped-storage unit and the governor, which helps to extend the service life of the equipment. This intelligent control method helps to reduce ineffective energy consumption, improve the operation efficiency of the power station, achieve energy conservation and emission reduction, provides better peak shaving and frequency support for the access of new energy (such as wind energy and solar energy), and contributes to the construction of a diversified power supply system.

[0077] In S1, the currently operating data includes at least one operating parameter among the current power of the pumped-storage unit, the water head at the inlet, the water head at the outlet, the output command, the AGC power command, the control command at the previous moment, the pressure pipeline flow rate, and the efficiency of the pumped-storage unit.

[0078] Among them, the current power of the pumped-storage unit represents the current power generation or pumping power of the pumped-storage unit; the inlet head refers to the head height when water enters the pumped-storage unit or the water turbine from the downstream reservoir, that is, the height difference at which the potential energy of the water flow is converted into mechanical energy and electrical energy; the outlet head refers to the head height when the pumped-storage unit or the water turbine releases water into the upstream reservoir or the generator, which is also the height difference at which the potential energy of the water flow is converted; the output instruction refers to the output power target set by the pumped-storage unit according to the grid demand or the dispatching instruction; the AGC power instruction is the power regulation instruction issued by the automatic generation control (AGC) system, which is used to adjust the output of the pumped-storage unit to maintain the grid frequency and stability; the previous control instruction is the control instruction of the pumped-storage unit given by the AGC system in the previous time period, which is used to analyze the effect of the control strategy and make adjustments; the pressure pipeline flow refers to the water flow through the pressure pipeline (connecting the upper reservoir and the lower reservoir), which affects the operation efficiency of the pumped-storage unit; the efficiency of the pumped-storage unit refers to the efficiency of the pumped-storage unit in converting energy, and the energy loss ratio in the whole cycle process of the pumped-storage unit from water absorption, pumping, energy storage, power generation to the water pump reflects the efficiency of converting energy.

[0079] In S2, the current operation data is preprocessed, and the preprocessing includes data cleaning and data normalization. Among them, data cleaning refers to identifying and removing outliers in the current operation data and filling in missing values in the current operation data to ensure data quality; data normalization refers to standardizing the operation parameters to the range of [0, 1] for the convenience of processing by the ESNN model.

[0080] In S3, the ESNN model includes an input layer, a hidden layer, and an output layer. Among them, the input layer includes multiple groups of spiking neurons, and each group contains 20 spiking neurons. For the convenience of distinction, the spiking neurons in the input layer are denoted as input layer spiking neurons. The input layer receives the operation parameters and converts the operation parameters into the spike signals of the input layer spiking neurons. The spike signals of the input layer spiking neurons are transmitted to the hidden layer. Each operation parameter corresponds to a group of input layer spiking neurons, and the firing frequency of the spike signals of each group of input layer spiking neurons changes proportionally according to the magnitude of the operation parameter. That is to say, the input layer spiking neurons encode information by emitting spike signals, and the Gaussian frequency modulation coding method is used to convert the operation parameters into the spike signals of the spiking neurons. The larger the operation parameter, the higher the firing frequency of the corresponding spike signals of the input layer spiking neurons. The value range of the firing frequency is [0, 200] Hz, which means that when the input value changes within the range of [0, 1], the firing frequency of the corresponding input layer spiking neurons will change from 0 Hz to 200 Hz.

[0081] Among them, the hidden layer adopts a three-layer recursive structure. Each layer of the recursive structure contains a preset number of spiking neurons. For the convenience of distinction, the spiking neurons in the hidden layer are denoted as hidden layer spiking neurons, and the preset number is 80. The hidden layer spiking neurons of each layer of the recursive structure are recursively connected to introduce the ability of temporal memory. The state of the hidden layer at the previous moment is used as the input at the current moment to achieve the continuity of the state. A full connection is adopted between adjacent layers of the recursive structure, and the connection weights between these hidden layer spiking neurons are adjustable to adapt to different information transmission requirements. Each hidden layer spiking neuron in the hidden layer is an integrate-and-fire unit, which means that the hidden layer spiking neuron emits a pulse signal when the membrane potential exceeds the preset threshold. For different hidden layer spiking neurons, the preset threshold can be the same or different, and the preset threshold is adjustable.

[0082] Among them, the output layer includes a plurality of spiking neurons. For the convenience of distinction, the spiking neurons in the output layer are denoted as output layer spiking neurons, and each output layer spiking neuron corresponds to a fixed gear. In this disclosure, 11 output layer spiking neurons and 11 fixed gears are specifically set. These 11 fixed gears are designed in the range from 0% to 100% with a step of 10%, covering the entire range of the governor opening. The first fixed gear is 0%, the second fixed gear is 10%, the third fixed gear is 20%, the fourth fixed gear is 30%, the fifth fixed gear is 40%, the sixth fixed gear is 50%, the seventh fixed gear is 60%, the eighth fixed gear is 70%, the ninth fixed gear is 80%, the tenth fixed gear is 90%, and the eleventh fixed gear is 100%. Based on the pulse signals of the hidden layer spiking neurons, the pulse signals of the output layer spiking neurons are generated. That is to say, after the pulse signals of the hidden layer spiking neurons are transmitted to the output layer spiking neurons, it may cause a change in the membrane potential of the output layer spiking neurons, and the change in the membrane potential triggers the output layer spiking neurons to emit pulse signals. The number of times of sending pulse signals of each output layer spiking neuron within the sampling period is counted, and based on the multiple fixed gears and the multiple sending times, the opening gear is obtained. That is to say, after the preprocessed current operating data is input into the ESNN model, the ESNN model outputs the opening gear.

[0083] Furthermore, the genetic algorithm is used to train the ESNN model. During the training process, parameters such as the connection weights and preset thresholds of the hidden layer are optimized. Refer to Figure 2 the schematic flowchart of the ESNN model training method shown in

[0084] S31: Randomly generate a population, set the maximum number of generations of evolution. The population includes M original individuals, each original individual contains a preset number of genes, each gene represents a parameter, and the parameter is represented by a floating point number;

[0085] S32: In each iteration, configure M original individuals into the ESNN model for simulation testing, and calculate the fitness of the M original individuals;

[0086] S33: Randomly select k parent individuals from the M original individuals. Among the k parent individuals, eliminate the parent individual with the maximum fitness, and retain the remaining k - 1 parent individuals;

[0087] S34: Determine whether the number of parent individuals is equal to M; if so, execute S35; if not, return to S33;

[0088] S35: Randomly divide the M parent individuals into two groups of individuals according to the crossover probability g. The first group of individuals includes g×M parent individuals, and the second group of individuals includes (1 - g)×M parent individuals;

[0089] S36: Use the simulated binary crossover operator to randomly cross the parent individuals in the first group of individuals to generate 2×(g×M) offspring individuals;

[0090] S37: According to the mutation probability, perturb the genes in (1 - g)×M parent individuals and 2×(g×M) offspring individuals to obtain 2×M mutant individuals;

[0091] S38: Configure the 2×M mutant individuals into the ESNN model for simulation testing, and calculate the fitness of the 2×M mutant individuals;

[0092] S39: Sort the 2×M mutant individuals in ascending order of fitness, and select the first M mutant individuals as the new M original individuals;

[0093] S310: Determine whether the maximum number of generations of evolution is reached or whether the overall fitness improvement of the population is not greater than the preset amplitude threshold within L consecutive iteration cycles; if either condition is met, execute S311; if neither condition is met, return to S32;

[0094] S311: Configure the original individual with the minimum fitness among the new M original individuals into the ESNN model.

[0095] In S31, the population is initially P(0), and the initial population is randomly generated. M is 50. Each original individual in the population represents a parameter set of the ESNN model. The parameter set includes multiple connection weights and multiple preset thresholds. One parameter is a gene. In this disclosure, each original individual includes 400 genes. The encoding method of the individual is real - number encoding. Therefore, each parameter (gene) is represented by a floating - point number. The maximum number of generations of evolution is T. In this disclosure, T is 100.

[0096] In S32, M original individuals are configured into the ESNN model, and historical operation data is input to conduct a simulation test on the ESNN model. The time for the simulation test is 1 hour. After the simulation test ends, the control performance of each individual is evaluated, and the fitness of each individual is calculated. The specific formula is as follows:

[0097] Fitness(m) = w1 * RMSE + w2 * ITAE + w3 * Ms;

[0098] The above formula is the fitness function. Fitness(m) is the fitness of the m-th original individual; m is the individual number, 1 ≤ m ≤ M; RMSE is the root mean square adjusted error, which measures the deviation between the model prediction value and the actual value; ITAE is the integral of time multiplied by the absolute error, which considers the cumulative effect of the error over time; Ms is the overshoot, which measures the overshoot degree of the system response; w1, w2, and w3 are all weighting coefficients used to balance the importance of different performance indicators. The fitness function comprehensively considers the regulation accuracy, response speed, and robustness of the ESNN model. This fitness function can comprehensively evaluate the control effect, and the smaller the fitness, the better the individual.

[0099] In S33, a smaller k value will increase the selection pressure but may reduce the diversity of the population, while a larger k value will increase the diversity of the population but may reduce the selection pressure. In this disclosure, k is 3, which can achieve a balance between the selection pressure (i.e., the tendency to select the individual with the highest fitness) and the diversity (i.e., maintaining the diversity of different individuals in the population). The larger the fitness, the worse the individual. Therefore, among the randomly selected k parent individuals, the parent individual with the largest fitness is screened out.

[0100] In S35, g is 0.8, g × M = 40, and (1 - g) × M = 20.

[0101] In S36, the simulated binary crossover operator (SBX operator, SBX stands for Simulated Binary Crossover) is adopted to randomly cross the parent individuals in the first group of individuals to generate offspring individuals. For every two parent individuals crossed, two offspring individuals are generated until 2 × (g × M) offspring individuals are generated, and 2 × (g × M) = 80.

[0102] Through the crossover probability, only some parent individuals participate in the crossover in each iteration to generate new offspring individuals, which helps to balance local search and global exploration.

[0103] In S37, the polynomial mutation operator is used to perturb the genes in g×M parent individuals and 2×(g×M) offspring individuals. The mutation probability is set to 0.1, which means that each gene has a 10% probability of being mutated. The mutation amplitude follows a polynomial probability distribution, which usually means that the mutation amplitude decreases as the gene value increases. Mutation helps the algorithm jump out of local optima and increases the global search ability. After perturbing the genes, new individuals are obtained, denoted as mutated individuals, with a quantity of 2×M.

[0104] In S38, the calculation principle of the fitness of 2×M mutated individuals is the same as that of the fitness of M individuals in S32, and will not be elaborated here.

[0105] In S39, the top M mutated individuals selected are regarded as elite individuals and are used as the new M original individuals for the next iteration. This elite retention strategy can prevent the loss of excellent individuals and accelerate the convergence speed.

[0106] In S310, L is 20, and the calculation formula for the overall fitness improvement amplitude of the population is as follows:

[0107]

[0108] where t is the current iteration number; P(t) is the overall fitness improvement amplitude of the population in the t-th iteration; F(t) is the average fitness of all original individuals in the population in the t-th iteration; F(t - 1) is the average fitness of all original individuals in the population in the (t - 1)-th iteration.

[0109] The first condition is that the current iteration number reaches the maximum number of evolutionary generations. The second condition is that the overall fitness improvement amplitude of the population is not greater than the preset amplitude threshold within L consecutive iteration cycles. As long as either of the two conditions is met, the iteration ends and S311 is executed. If neither of the two conditions is met, return to S32 to continue the iteration. Among them, if the second condition is met, it is considered that the population tends to converge and the iteration ends prematurely.

[0110] In S311, select the original individual with the minimum fitness among the new M original individuals, and the genes contained in this original individual are used as the final parameter configuration of the ESNN model into the ESNN model.

[0111] Refer to Figure 3 Referring to the flowchart of the opening position gear acquisition method shown, in S4, based on multiple fixed gears and multiple transmission times, obtain the opening position gear, including the following steps:

[0112] S41: Based on multiple transmission times and the weights of the output layer pulse neurons preset, obtain the weighted average number;

[0113] S42: Determine whether the weighted average number of times is equal to the average of any two transmission times; if it is S43, then execute; if not, then execute S44;

[0114] S43: Obtain the two fixed gears corresponding to the two transmission times, and the opening gear is equal to the average of the two fixed gears;

[0115] S44: Obtain the transmission time closest to the weighted average number of times, denoted as the closest transmission time;

[0116] S45: The opening gear is equal to the fixed gear corresponding to the closest transmission time.

[0117] In S41, the calculation formula for the weighted average number of times is as follows:

[0118]

[0119] where QP is the weighted average number of times; QH is the weighted pulse number; QZ is the total weight of multiple output layer pulse neurons; i is the serial number of the output layer pulse neuron; I is the total number of output layer pulse neurons; Q i is the weight of the i-th output layer pulse neuron; M i is the number of times the pulse signal of the i-th output layer pulse neuron is transmitted within the sampling period.

[0120] An example of S42 - S45 is as follows: If there is 1 output layer pulse neuron, the number of times the pulse signal is transmitted within the sampling period is 10, and its corresponding fixed gear is 10%, and there is another 1 output layer pulse neuron, the number of times the pulse signal is transmitted within the sampling period is 20, and its corresponding fixed gear is 20%, and the weighted average number of times is 15. 15 is between 10 and 20, that is, the weighted average number of times is equal to the average of the two transmission times, then the opening gear is equal to the average of 10% and 20%, that is, 15%. If the weighted average number of times is not equal to the average of any two transmission times, then among the number of times the pulse signals of 11 output layer pulse neurons are transmitted within the sampling period, select the transmission time closest to the weighted average number of times, and the opening gear is equal to the fixed gear corresponding to this transmission time.

[0121] Although the present disclosure is designed with multiple discrete fixed gears, by calculating the weighted average number of times, the outputs of the pulse neurons in each output layer are integrated and mapped to a final unified opening gear. During the weighting process, the error caused by discrete control steps can be reduced, the control accuracy can be improved, and smoother and more precise control can be provided. The firing frequency of the pulse signals of the pulse neurons in the output layer can reflect the time dynamic changes. Through weighted averaging, these changes can be captured and converted into control commands for the governor. By allocating different weights, the expression ability of the ESNN model for complex control tasks can be enhanced, especially when multiple inputs need to be considered simultaneously. Moreover, the ESNN model can learn the optimal control strategies under different working conditions, enabling the system to adapt to various complex grid operating conditions, providing a basis for the intelligent upgrade and future technical expansion of pumped-storage power stations.

[0122] In S5, an opening command is generated based on the obtained opening gear and transmitted to the governor, and the governor adjusts the guide vane opening of the pumped-storage unit according to the opening command.

[0123] Optionally, a visualization component for key components in the pumped-storage power station is designed using the WPF framework to ensure that the graphical representation of each visualization component is intuitive and easy for users to identify and operate. Among them, the key components include at least one of the pumped-storage unit, the governor, and the ESNN controller. Of course, it can also include a water turbine, a generator, etc. Among them, the ESNN controller is encapsulated by the ESNN model. A component library is created, and the component library includes visualization components of key components in the pumped-storage power station. Users are allowed to adjust the characteristics of the visualization components, such as input-output characteristics and control logic, by adjusting component parameters according to actual needs, and an intuitive component parameter setting interface is provided to ensure that users can easily customize the visualization components. Support users to connect different visualization components through logical connections to form a complete control process. The logical connections define the interaction and control logic between the visualization components, realizing the visual representation of the automatic control strategy. These visualization components and logical connections constitute a visual configuration editor, which is used to guide the actual operation of the water storage power station. When the visual configuration editor runs, it collects the current operation data in real time, calculates the opening gear, and realizes automatic generation control.

[0124] Develop a real-time curve control to plot key process variables such as the power of the pumped-storage unit, the inlet head, the outlet head, and the opening gear, involving statistical analysis functions. This function calculates the performance indicators of automatic generation control (AGC), such as regulation deviation and qualification rate, and then forms a detailed report for in-depth statistical analysis and performance evaluation. The visual configuration editor, the real-time curve control, and the statistical analysis function constitute the pumped-storage AGC configuration development platform.

[0125] Deploy the pumped-storage AGC configuration development platform for engineering, including deploying the platform to the control center server of the pumped-storage power station, connecting to the lower computer using the OPC protocol to collect operation data in real time, and ensuring the integration of the interface of the pumped-storage AGC configuration development platform with the historical database, so as to realize the functions of data archiving and historical curve playback, and provide comprehensive monitoring and analysis capabilities for the operation of the pumped-storage power station.

[0126] In the visual configuration editor, construct the control logic diagram of automatic generation control (AGC) according to the actual topological structure of the pumped-storage power station, set the parameters of each device including the rated values of the water turbine and generator and the PID parameters of the governor, configure the parameters of the ESNN intelligent controller such as the input and output scales, connection weights and thresholds, etc., and perform point mapping to ensure that the variables in the visual configuration editor correspond one-to-one with the actual data sources, so as to ensure the smoothness of the data link.

[0127] Conduct off-line simulation tests, including selecting typical historical operation conditions such as large-scale load increase and decrease of AGC and start-stop of pumped-storage units, using the simulation box model to construct a complete AGC system covering components such as water turbines, governors, and ESNN controllers, running off-line simulation to evaluate whether the control performance meets standards such as regulation accuracy and response time, and analyzing the simulation curves to optimize the logic and parameters of the pumped-storage AGC configuration development platform, so as to improve the overall control effect.

[0128] Conduct on-line trial operation, which involves selecting a period when the operation of the pumped-storage unit is relatively stable to switch to the ESNN intelligent AGC control mode. Initially, dual-mode redundancy operation can be carried out with the traditional PID control to compare the performance differences in real time. Continuously run the trial operation for about 1 week to evaluate the actual effect of the AGC control, especially pay attention to the fluctuations of the grid frequency and the power of the pumped-storage unit. If the trial operation results meet the expectations, the intelligent AGC mode can be fully switched, and the control parameters can be adjusted in a timely manner according to seasonal changes and load conditions.

[0129] Strictly assess the control quality of automatic generation control (AGC) every month and compile an evaluation report containing indicators such as the CPS qualification rate and adjustment absolute error. Diagnose problems during the non-compliant periods to analyze the reasons and formulate optimization measures. Long-term track the AGC control effect and continuously improve the pumped-storage AGC configuration development platform. At the same time, summarize the operation characteristics of the power station by analyzing the AGC operation history through big data, form an expert control knowledge base, provide guidance for subsequent intelligent upgrades, and realize effect evaluation and optimization.

[0130] An example of the above is as follows: A large pumped-storage power station has been transformed for automatic generation control (AGC) by adopting the above-mentioned pumped-storage AGC configuration development platform. During this transformation process, AGC configuration development has been carried out for 4 reversible units with a single-unit capacity of 300 MW in the power station. Using the visual configuration editor, a control logic covering "primary frequency modulation - AGC - load distribution" has been constructed to ensure the systematicness and coordination of the control process.

[0131] In terms of the intelligent control core, the ESNN controller is designed with a structure of 8 input nodes and 1 output node. The input nodes cover key parameters such as the head at the inlet, the head at the outlet, and the output command, while the output node is responsible for generating the opening command of the governor. The hidden layer of the ESNN controller consists of 3 layers, with each layer containing 80 spiking neurons, forming a deep learning model that can handle complex control tasks. By using the measured operation data of the past year for offline training, the connection weights and thresholds of the ESNN controller have been optimized, and finally, the parameter configuration with the optimal performance has been obtained.

[0132] The simulation test results show that the ESNN intelligent AGC control scheme performs excellently under multiple typical working conditions, and the CPS1 index exceeds 90% in all cases, significantly better than the standard requirements. In practical applications, the start-stop frequency of the pumped-storage units has been reduced by 20%, and the peak regulation capacity has been increased by 15%. After one month of on-site trial operation, compared with the traditional PID control, the grid frequency qualification rate has been increased by 5 percentage points, and the active power response time has been reduced by 2 seconds. After six months of continuous monitoring, the AGC control operates stably, significantly improving the peak regulation and grid connection support capabilities of the pumped-storage power station, creating favorable conditions for the consumption of new energy.

[0133] From deployment and implementation to evaluation and optimization, each link needs to comprehensively consider the engineering reality, actively apply new technical means, and continuously improve the control scheme. Practice has proved that the pumped-storage AGC configuration development platform with ESNN as the core can significantly improve the automatic control level of the pumped-storage power station, achieving the goals of energy conservation, efficiency improvement, safety, and reliability. Moreover, this platform is expected to integrate and strengthen frontier technologies such as reinforcement learning and knowledge graphs, and continuously optimize using operation big data, evolving into a more intelligent and adaptive AGC system. This is of great significance and broad prospects for building a clean and low-carbon power system with new energy as the main body. Through continuous technological innovation and system optimization, pumped-storage power stations will be able to better adapt to the development needs of future power systems and make greater contributions to the realization of sustainable energy development.

[0134] Refer to Figure 4 , the present disclosure provides a control system for a pumped-storage power station, including:

[0135] A data acquisition module 101 for acquiring the current operation data of the pumped-storage power station in real time;

[0136] A preprocessing module 102 for preprocessing the current operation data;

[0137] A model construction module 103 for constructing an ESNN model;

[0138] A gear position acquisition module 104 for acquiring the opening gear position based on the preprocessed current operation data and the ESNN model;

[0139] A unit adjustment module 105 for controlling the governor to adjust the guide vane opening of the pumped-storage unit based on the opening gear position.

[0140] The various variation modes and specific examples in the above-provided control method of the pumped-storage power station are equally applicable to the control system of the pumped-storage power station provided by the present disclosure. Through the foregoing detailed description of the control method of the pumped-storage power station, those skilled in the art can clearly know the implementation method of the control system of the pumped-storage power station. For the sake of simplicity of the specification, it will not be elaborated herein.

[0141] The computer device according to an embodiment of the present disclosure includes a memory and a processor. The memory is used to store non-temporary computer-readable instructions. Specifically, the memory may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0142] The processor may be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and may control other components in the computer device to perform desired functions. In an embodiment of the present disclosure, the processor is used to run the computer-readable instructions stored in the memory, so that the computer device executes all or part of the steps of the control method of the pumped-storage power station in the foregoing embodiments of the present disclosure.

[0143] Those skilled in the art should understand that, in order to solve the technical problem of how to obtain a good user experience effect, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included in the protection scope of the present disclosure.

[0144] Such as Figure 5 is a schematic structural diagram of a computer device provided by an embodiment of the present disclosure. It shows a schematic structural diagram suitable for implementing the computer device in the embodiments of the present disclosure.Figure 5 The computer device shown is merely an example and should not impose any limitations on the functions and scope of use of the embodiments of the present disclosure.

[0145] As Figure 5 shown, the computer device may include a processor (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage device into a random access memory (RAM). In the RAM, various programs and data required for the operation of the computer device are also stored. The processor, the ROM, and the RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0146] Generally, the following devices may be connected to the I / O interface: an input device including, for example, a sensor or a visual information acquisition device, etc.; an output device including, for example, a display screen, etc.; a storage device including, for example, a magnetic tape, a hard disk, etc.; and a communication device. The communication device may allow the computer device to communicate with other devices (such as edge computing devices) wirelessly or wiredly to exchange data. Although Figure 5 a computer device with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had.

[0147] In particular, according to the embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, the embodiments of the present disclosure include a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program may be downloaded and installed from a network through the communication device, or installed from a storage device, or installed from the ROM. When the computer program is executed by the processor, all or part of the steps of the control method of the pumped-storage power station according to the embodiments of the present disclosure are executed.

[0148] For the detailed description of this embodiment, reference may be made to the corresponding descriptions in the foregoing embodiments, and details will not be repeated here.

[0149] A computer-readable storage medium according to an embodiment of the present disclosure stores non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are run by a processor, all or part of the steps of the control methods of the pumped-storage power station according to the foregoing embodiments of the present disclosure are executed.

[0150] The above computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROMs and DVDs), magneto-optical storage media (e.g., MOs), magnetic storage media (e.g., magnetic tapes or external hard drives), media with built-in rewritable non-volatile memories (e.g., memory cards), and media with built-in ROMs (e.g., ROM cartridges).

[0151] For a detailed description of this embodiment, reference may be made to the corresponding descriptions in the foregoing embodiments, which will not be repeated herein.

[0152] The basic principles of the present disclosure have been described above in connection with specific embodiments. However, it should be noted that the advantages, benefits, effects, etc. mentioned in the present disclosure are merely examples and not limitations, and it cannot be considered that these advantages, benefits, effects, etc. are essential for each embodiment of the present disclosure. Additionally, the specific details disclosed above are for illustrative and facilitating understanding purposes only and are not limitations. The above details do not limit the present disclosure to necessarily adopting the above specific details for implementation.

[0153] In the present disclosure, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The block diagrams of the devices, apparatuses, equipment, and systems involved in the present disclosure are only illustrative examples and do not intend to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open-ended terms meaning "including but not limited to" and can be used interchangeably with each other. The words "or" and "and" used herein refer to the word "and / or" and can be used interchangeably with each other, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to" and can be used interchangeably with each other.

[0154] In addition, as used herein, the "or" used in the listing of items starting with "at least one" indicates a disjunctive listing, so that for example, the listing of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the term "exemplary" does not mean that the described examples are preferred or better than other examples.

[0155] It should also be noted that in the systems and methods of the present disclosure, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present disclosure.

[0156] Various changes, substitutions, and alterations to the technology described herein may be made without departing from the teachings defined by the appended claims. Additionally, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, compositions of events, means, methods, and acts described above. Processes, machines, manufactures, compositions of events, means, methods, or acts that are currently available or later to be developed that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein may be utilized. Accordingly, the appended claims include such processes, machines, manufactures, compositions of events, means, methods, or acts within their scope.

[0157] The foregoing description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the disclosure. Thus, the present disclosure is not intended to be limited to the aspects shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0158] The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of the disclosure to the form disclosed herein. Although several example aspects and embodiments have been discussed above, those skilled in the art will recognize some of their variations, modifications, alterations, additions, and subcombinations.

Claims

1. A control method for a pumped storage power station, characterized in that: include: Obtain current operating data of pumped storage power stations in real time; Preprocessing the current operation data; Build ESNN model; Based on the preprocessed current operating data and the ESNN model, obtaining the opening gear position; The speed regulator is controlled to adjust the guide vane opening of the pumped storage unit based on the opening gear position.

2. The control method of a pumped storage power station according to claim 1, characterized in that: The current operating data includes at least one operating parameter of the pumped-storage unit's current power, water inlet head, water outlet head, output instruction, AGC power instruction, control instruction at the last moment, pressure pipeline flow, and efficiency of the pumped-storage unit.

3. The control method of a pumped storage power station according to claim 2, characterized in that: The ESNN model includes an input layer, a hidden layer and an output layer; The input layer receives the operating parameters and converts the operating parameters into pulse signals of the pulse neurons of the input layer; Transmitting the pulse signal of the input layer pulse neuron to the hidden layer; wherein each of the operating parameters corresponds to a group of input layer pulse neurons, and the emission frequency of the pulse signal of each group of input layer pulse neurons changes in proportion to the size of the operating parameter; The hidden layer adopts a three-layer recursive structure, each layer of the recursive structure includes a preset number of hidden layer pulse neurons; based on the pulse signal of the input layer pulse neuron and the hidden layer state at the previous moment, the pulse signal of the hidden layer pulse neuron is generated; the pulse signal of the hidden layer pulse neuron is transmitted to the output layer; The output layer includes multiple output layer pulse neurons, each output layer pulse neuron corresponds to a fixed gear; based on the pulse signal of the hidden layer pulse neuron, the pulse signal of the output layer pulse neuron is generated; the number of times the pulse signal of each output layer pulse neuron is sent within a sampling period is counted; based on multiple fixed gears and multiple sending times, the opening gear is obtained.

4. The control method of a pumped storage power station according to claim 1, characterized in that: Also includes: A population is randomly generated, and a maximum evolutionary generation is set, wherein the population includes M original individuals, each original individual contains a preset number of genes, each gene represents a parameter, and the parameter is represented by a floating point number; In each iteration, the M original individuals are configured into the ESNN model for simulation testing, and the fitness of the M original individuals is calculated; Randomly select k parent individuals from the M original individuals, and among the k parent individuals, screen out the parent individual with the largest fitness, and retain the remaining k-1 parent individuals; Determine whether the number of parent individuals is equal to M; If it is equal to M, then according to the crossover probability g, the M parent individuals are randomly divided into two groups of individuals. The first group of individuals includes g×M parent individuals, and the second group of individuals includes (1-g)×M parent individuals. If it is not equal to M, then continue to randomly select k parent individuals from the M original individuals; The simulated binary crossover operator is used to perform random crossover on the parent individuals in the first group of individuals to generate 2×(g×M) offspring individuals; According to the mutation probability, the genes in the (1-g)×M parent individuals and the 2×(g×M) offspring individuals are disturbed to obtain 2×M mutant individuals; The 2×M mutant individuals are configured into the ESNN model for simulation testing, and the fitness of the 2×M mutant individuals is calculated; The 2×M mutant individuals are sorted in order of fitness from small to large, and the first M mutant individuals are selected as the new M original individuals; When the maximum evolutionary generation is reached or the overall fitness of the population is improved by no more than a preset threshold within L consecutive iteration cycles, the original individual with the smallest fitness among the new M original individuals is configured into the ESNN model.

5. The control method of a pumped storage power station according to claim 3, characterized in that: The step of obtaining the opening gear position based on a plurality of fixed gear positions and a plurality of sending times includes: Obtaining a weighted average number of times based on the multiple transmission times and the preset weights of the output layer pulse neurons; Determine whether the weighted average number of times is equal to the average of any two sending times; If yes, then obtain two fixed gears corresponding to the two sending times, and the opening gear is equal to the average value of the two fixed gears; If not, then obtain the number of transmissions closest to the weighted average number of transmissions, and record it as the closest number of transmissions; The opening gear is equal to the fixed gear corresponding to the closest number of transmission times.

6. The control method of a pumped storage power station according to claim 5, characterized in that: The calculation formula of the weighted average number of times is: Wherein, QP is the weighted average number; QH is the weighted pulse number; QZ is the sum of the weights of the plurality of output layer pulse neurons; i is the sequence number of the output layer pulse neurons; I is the total number of output layer pulse neurons; Q i is the weight of the spiking neuron in the i-th output layer; M i is the number of times the pulse signal of the ith output layer pulse neuron is sent within the sampling period.

7. The control method of a pumped storage power station according to claim 1, characterized in that: Also includes: The key components of the pumped storage power station are set as a plurality of visualization components, wherein the plurality of visualization components are stored in a preset component library; Connecting the plurality of visualization components by using logic lines to form a visualization configuration editor; The visual configuration editor is used to control the operation of the pumped storage power station; Among them, the key components include at least one of a pumped storage unit, a speed regulator, and an ESNN controller, and the ESNN controller is encapsulated by the ESNN model.

8. A computer device, characterized in that: The computer device comprises: 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, and the instructions are executed by the at least one processor so that the at least one processor can execute the control method of the pumped-storage power station described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the control method of a pumped-storage power station as described in any one of claims 1-7.

10. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images