Flywheel energy storage-thermal power combined secondary frequency modulation multi-target control method and device for power grid frequency modulation specification

Abstract

The invention discloses a flywheel energy storage-thermal power combined secondary frequency modulation multi-target control method and device for power grid frequency modulation specifications, and relates to the field of power frequency modulation control, and the method comprises the steps: dividing the process of thermal power responding to an AGC instruction into a plurality of target time periods, and setting a combined system frequency modulation power curve according to the performance index of each time period; then, LSTM is used for predicting the frequency modulation capability of the thermal power generating unit, and logistic regression fitting is used for predicting the maximum charging and discharging power of a flywheel monomer; and constructing a target layer-comprehensive satisfaction model taking the power grid frequency modulation performance, the system frequency modulation cost and the flywheel loss as target layers. And in a preset output boundary, taking a comprehensive target layer-comprehensive satisfaction model as an optimization criterion, selecting an optimal load distribution scheme of the thermal power and the flywheel from the solution set by improving optimization of a multi-target genetic algorithm, and issuing the optimal load distribution scheme to a thermal power and flywheel system to execute power response. The high efficiency and economical efficiency of flywheel energy storage-thermal power combined secondary frequency modulation are improved at the same time.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power frequency modulation control, in particular to a flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method and device oriented to power grid frequency modulation specifications. BACKGROUND

[0002] Energy generation has randomness, intermittency and volatility, which leads to the decline of power grid inertia and the weakening of damping, and the power grid frequency changes greatly under load disturbance, which brings great potential problems to the long-term safe operation of the power system. The traditional thermal power unit and the hydroelectric generating unit are difficult to meet the requirements of continuous frequency modulation of the power grid due to slow start, low climbing rate and other problems, and frequent load modulation will cause unit pressure oscillation and increase wear and tear. Energy storage power can quickly absorb or output active power, and coupled energy storage system can assist traditional frequency modulation units to perform rapid frequency regulation, improve the quality and economy of power grid frequency modulation. Therefore, flexible control of the output distribution of energy storage and traditional thermal power units and the control strategy within the large-capacity energy storage array has become the key to improving the frequency modulation performance of traditional units, improving system economy and reducing losses.

[0003] In the load output allocation strategy of the combined energy storage and traditional frequency modulation unit, the existing traditional allocation methods mainly include the proportional allocation according to the dynamic adjustable capacity of the two systems and the allocation according to the frequency domain characteristics of the frequency modulation signal. Although both are relatively simple and easy to implement in engineering, they lack specific consideration of the grid frequency modulation performance evaluation and the frequency modulation cost and benefit, and do not take into account the loss of the traditional unit and the energy storage in different states. In view of the different goals of the secondary frequency modulation of the combined frequency modulation unit, many scholars have proposed various optimization schemes and control models to carry out extensive research on the load allocation output strategy. Based on PI control, the economic dispatch iteration of the secondary frequency modulation of the combined frequency modulation system is realized, and the iteration convergence and the update of the multiplier are promoted through local feedback control, effectively improving the economy of the secondary frequency modulation of the system. The energy storage system state of charge and output deviation are regionally divided by using fuzzy theory, and the economy of the energy storage system throughout the life cycle is evaluated by using the net present value method, and finally the particle swarm optimization algorithm is used to optimize the energy storage output of each instruction period. The above research has good optimization effect for the target, but fails to consider the control strategy from multiple aspects such as frequency modulation performance, frequency modulation cost, energy storage loss and SOE state. In order to consider multiple goals of AGC frequency modulation of the system, some scholars have proposed a double-layer model to realize the multi-objective optimization allocation of the load of the combined frequency modulation system. Based on the grid frequency modulation demand and the battery frequency modulation capacity, a double signal allocation mode is proposed for the double-layer fuzzy control strategy of the coordinated response of the frequency modulation unit to the automatic generation control instruction, which accurately allocates the power output depth of the frequency modulation unit, maintains the battery state of charge, and controls the frequency deviation fluctuation in a smaller range at a faster speed. Based on the energy storage life state economy representation and the frequency modulation limit division, a double-layer control strategy is proposed, which includes the upper layer dynamic balance bidding and the energy storage performance representation, the bottom layer flywheel energy storage system frequency modulation limit standard and the energy storage operation scheme in all scenarios, to improve the economy and frequency modulation performance of the combined frequency modulation system. The above control system has good optimization effect for the target, and also formulates the related energy storage energy management strategy to prevent the SOE from exceeding the limit. However, when the thermal power unit participates in AGC frequency modulation, the power system uses different frequency modulation performance indicators to evaluate and compensate the unit in the start-up period, climbing period and adjustment period, and the frequency modulation energy and power demand are different in different frequency modulation periods, so the load power allocation strategy of the combined frequency modulation system will cause differences in frequency modulation cost and loss.

[0004] In summary, in the aspect of flywheel energy storage-thermal power combined secondary frequency modulation, there is currently a lack of comprehensive consideration of multiple targets such as frequency modulation performance, cost and loss of the system in different periods under the grid frequency modulation specification (grid evaluation rules), and the real-time state of the two systems in different periods is not fully considered to affect the load allocation. SUMMARY

[0005] The application aims to provide a flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method and device facing grid frequency modulation specifications, which can improve the efficiency and economy of flywheel energy storage-thermal power combined secondary frequency modulation.

[0006] To achieve the above-mentioned purpose, the application provides the following solutions.

[0007] In a first aspect, the application provides a flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method facing grid frequency modulation specifications, which comprises: according to the grid frequency modulation specifications, dividing the frequency modulation process of the thermal power complete response primary AGC into multiple target periods based on the frequency modulation performance index; the target period comprises: a start-up period, a climbing period, an adjustment period and a non-assessment period; respectively assessing and compensating each period based on a preset performance index, and setting a combined frequency modulation system frequency modulation power curve according to the target value of the performance index; the preset performance index comprises: response time, regulation rate and regulation accuracy; predicting the thermal power unit frequency modulation capacity by using an LSTM model; linearly fitting the historical flywheel single charging and discharging power based on a logistic regression function, and predicting the maximum charging and discharging power of the flywheel single; constructing three target layers including the grid AGC frequency modulation performance index, the combined frequency modulation system secondary frequency modulation cost and the flywheel energy storage system loss, quantifying the achievement degree of each target layer by using a satisfaction fuzzy membership function, and dynamically adjusting the target weight by using an adaptive weight coefficient matrix to construct a target layer-comprehensive satisfaction model; taking the combined frequency modulation system frequency modulation power curve, the thermal power unit frequency modulation capacity and the maximum charging and discharging power of the flywheel single as the boundary conditions of the combined system output control amount, taking the target layer-comprehensive satisfaction model as the judgment criterion, adopting an improved multi-objective genetic algorithm to optimize the load distribution of the thermal power and the flywheel output distribution scheme, introducing a dynamic iteration number and an adaptive crossover probability to optimize the calculation efficiency while ensuring the optimization accuracy, and selecting the best thermal power and flywheel output distribution scheme from the Pareto solution set; issuing power instructions to the thermal power unit coordinated control system and the flywheel energy storage array energy management system according to the best thermal power and flywheel output distribution scheme, and executing power response.

[0008] In a second aspect, the application provides a computer device, which comprises a memory, a processor, a computer program stored on the memory and executable on the processor, and the processor executes the computer program to implement the flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method facing grid frequency modulation specifications.

[0009] According to the specific embodiments provided by the application, the following technical effects are disclosed.

[0010] The application predicts the frequency modulation capacity of the thermal power unit by adopting an LSTM model, and simultaneously predicts the maximum charge-discharge power of a flywheel unit by linear fitting based on a logistic regression function, so as to provide accurate boundary conditions for optimization distribution. The application further constructs a three-target optimization model including the frequency modulation performance, frequency modulation cost and flywheel loss, and constructs a comprehensive satisfaction model through a satisfaction fuzzy membership function and an adaptive weight matrix, so as to ensure the performance and take into account the economic effect. Moreover, the application takes the comprehensive satisfaction model as a criterion, adopts an improved multi-objective genetic algorithm for load distribution optimization, improves the efficiency while ensuring the accuracy through a dynamic iteration number and an adaptive crossover probability, and selects the best distribution scheme from a Pareto solution set to issue a power instruction for the thermal power unit and the flywheel energy storage system. The application takes into account the grid frequency modulation requirement and the safety and economy of the thermal power plant, significantly reduces the frequency modulation cost and equipment loss while ensuring the frequency modulation performance, and provides a feasible solution for the engineering application of large-scale flywheel energy storage on the power supply side. BRIEF DESCRIPTION OF DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0012] Figure 1 The unit frequency modulation demand analysis curve provided by the embodiment of the present application.

[0013] Figure 2 The flywheel unit maximum charge-discharge power curve provided by the embodiment of the present application Figure 3 The load distribution flowchart based on the improved multi-objective genetic algorithm provided by the embodiment of the present application.

[0014] Figure 4 The flywheel-thermal power combined AGC frequency modulation data 60min high frequency band frequency modulation simulation result schematic diagram in the strategy of the embodiment of the present application; wherein, Figure 4 (a) in the above is a grid frequency deviation curve; Figure 4 (b) in the above is a thermal power unit output power curve; Figure 4 (c) in the above is a flywheel energy storage system output power curve.

[0015] Figure 5 The flywheel-thermal power combined AGC frequency modulation data certain 4-hour typical period frequency modulation simulation result schematic diagram in the strategy of the embodiment of the present application; wherein, Figure 5 (a) in the above is a grid frequency deviation curve; Figure 5(b) in FIG. 1 is a flywheel energy storage system output power curve diagram. Figure 5 (b) in FIG. 1 is a flywheel energy storage system output power curve diagram.

[0016] Figure 6 FIG. 1 is a schematic diagram of flywheel-thermal combined AGC frequency modulation data of a certain 4-hour typical period in a comparison strategy 1 provided by an embodiment of the present application; wherein, Figure 6 (a) in FIG. 1 is a power grid frequency deviation curve diagram. Figure 6 (b) in FIG. 1 is a flywheel energy storage system output power curve diagram. Figure 6 (c) in FIG. 1 is a flywheel energy storage system output power curve diagram. Figure 6 (d) in FIG. 1 is a flywheel energy storage system output power curve diagram.

[0017] Figure 7 FIG. 2 is a schematic diagram of flywheel-thermal combined AGC frequency modulation data of a certain 4-hour typical period in a comparison strategy 2 provided by an embodiment of the present application; wherein, Figure 7 (a) in FIG. 2 is a 4h system frequency modulation cost box plot. Figure 7 (b) in FIG. 2 is a 4h system frequency modulation cost box plot. Figure 7 (c) in FIG. 2 is a 4h flywheel energy storage loss box plot Figure 8 FIG. 3 is a flowchart of a flywheel energy storage-thermal combined secondary frequency modulation multi-objective control method facing power grid frequency modulation specifications provided by an embodiment of the present application.

[0018] Figure 9 FIG. 4 is a structural schematic diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0020] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0021] Embodiment 1, as shown in FIG. 1, provides a flywheel energy storage-thermal combined secondary frequency modulation multi-objective control method facing power grid frequency modulation specifications, which comprises the following steps. Figure 8

[0022] ​S1. According to the power grid frequency regulation specifications, the frequency regulation process of a complete AGC response of a thermal power plant is divided into multiple target time periods based on the frequency regulation performance indicators; the target time periods include: the start-up period, the ramp-up period, the adjustment period, and the non-assessment period.

[0023] Optionally, the power grid frequency regulation standard is a standard set based on two detailed rules: the "Implementation Rules for the Management of Power Plant Grid-Connected Operation" and the "Implementation Rules for the Management of Ancillary Services of Grid-Connected Power Plants".

[0024] In practical applications, such as Figure 1 As shown. Before that time, the unit was operating stably at its output value. nearby, At that moment, the AGC control program issued a power rating of [value] to the unit. Upon receiving the start-up command, the unit issues instructions to the boiler and turbine, utilizing the thermal storage response output while simultaneously adjusting the coal feed rate, after a certain delay... Always Reliable Steps The adjustment dead zone, to The moment the power command range is first reached, then... It fluctuated slightly nearby and remained stable at [a certain level]. Nearby, until Once the system enters the stable dead zone, a complete power-up secondary frequency regulation operation is completed. According to the grid frequency regulation performance assessment and compensation details, response times are measured during the initial response, ramp-up, and stabilization phases of the thermal power plant's secondary frequency regulation. Adjustment rate Adjustment accuracy The assessment and compensation can be based on the target value of the performance index to set the ideal output curve of the system, that is, the frequency modulation power curve of the joint frequency modulation system.

[0025] S2. Each time period is assessed and compensated based on preset performance indicators, and the frequency modulation power curve of the joint frequency modulation system is set according to the target value of the performance indicators; the preset performance indicators include: response time, adjustment rate and adjustment accuracy.

[0026] S3. Predict the frequency regulation capability of thermal power units using LSTM models.

[0027] Furthermore, the LSTM model includes a first-level sub-model and a second-level sub-model. The first-level model predicts the main steam pressure and temperature based on historical operating condition parameter data. The second-level model predicts the actual power generation output of the thermal power unit based on the main steam pressure and temperature. The historical operating condition parameter data includes: coal quantity, air supply volume, feedwater flow rate, main steam pressure, and main steam temperature.

[0028] In practical applications, firstly, in the first-level sub-model, the historical sequences of coal quantity, air supply volume, feedwater flow rate, main steam pressure, and main steam temperature are used as model inputs. The main steam pressure and temperature of the next sampling period are then used as model outputs to characterize the boiler's response capability. The prediction process is as follows.

[0029] .

[0030] In the formula, , These represent the predicted values ​​of the main steam pressure and temperature at the next timestamp, respectively. , , These represent the values ​​of coal feed rate, air supply rate, and water supply flow rate, respectively.

[0031] Secondly, the main steam pressure and temperature predicted by the first-level sub-model are used as inputs to the second-level sub-model, while the actual power generation is used as the output. The calculation formula is as follows.

[0032] .

[0033] In the formula, It is a predicted value of actual power generation, which reflects the grid's ability to regulate automatic power generation control commands.

[0034] S4. Based on the logistic regression function, the historical charge and discharge power of individual flywheel units is linearly fitted, and the maximum charge and discharge power of individual flywheel units is predicted.

[0035] Furthermore, the linear fitting method is least squares fitting.

[0036] Furthermore, the calculation formula for the maximum charging and discharging power of the flywheel unit is as follows.

[0037] .

[0038] In the formula, This represents the maximum charging power of a single flywheel unit. This represents the maximum discharge power of a single flywheel unit. and These represent the minimum and maximum values ​​of SOC, respectively. and These are the rated charging power and discharging power of the flywheel unit, respectively; is the scaling factor, and n is the rate factor, used to control the degree of change of the exponential term.

[0039] In practical applications, in order to predict the frequency regulation capability of the FESA and prevent individual FESUs from overcharging and discharging, this application designs power limit values ​​based on the cell's SOC and logistic regression function.

[0040] in, and These represent the minimum and maximum values ​​of SOC, respectively. and These represent the rated charging power and discharging power of the FESU, respectively.

[0041] For example, this application smooths and fits the experimental data of the maximum charging and discharging power of a 500 kW flywheel energy storage cell that has been put into engineering application in a certain location. , , , , and The values ​​are 0.1, 0.35, 0.65, 0.9, 0.5, and 15. The maximum output limit of the flywheel unit is as follows: Figure 2 As shown.

[0042] S5. Construct three target layers, including the grid AGC frequency regulation performance index, the secondary frequency regulation cost of the joint frequency regulation system, and the flywheel energy storage system loss. Quantify the achievement degree of each target layer through the satisfaction fuzzy membership function, and dynamically adjust the target weights using an adaptive weight coefficient matrix to construct a target layer-comprehensive satisfaction model.

[0043] Furthermore, the calculation formula for the frequency regulation performance index of the power grid AGC is as follows:

[0044] .

[0045] In the formula, For standard adjustment rate, The adjustment rate of the unit. For the deviation of the unit, To allow for adjustment deviation, The response time for a unit to exit the dead zone. This is the standard response time.

[0046] 2) The formula for calculating the secondary frequency modulation cost of the combined frequency modulation system is as follows.

[0047] .

[0048] In the formula, The cost of secondary frequency modulation for a combined frequency modulation system; Let t be the marginal cost of the thermal power unit. This represents the real-time power output of the thermal power unit. The levelized cost of electricity (LCOE) of the flywheel energy storage system at time t; Let be the energy storage output power at time t; For energy storage investment costs; For energy storage operation and maintenance costs; The electricity price for the power station during charging; This refers to the rated capacity of the energy storage. To improve the charging and discharging efficiency of energy storage systems; Total time; For real-time grid connection electricity price; Cost per unit capacity; Rated power; For system lifespan; For the first Annual maintenance costs; The discount rate is 6%. Cost of charging energy storage; Discharge duration at rated power.

[0049] In practical applications, the construction process of the target-level comprehensive satisfaction model is as follows.

[0050] (1) Construct the target layer.

[0051] Objective 1: Frequency regulation performance indicators of power grid AGC.

[0052] With the gradual opening and maturation of the AGC (Automatic Generation Control) ancillary services market, the compensation method for frequency regulation units by the power grid has evolved into market-based compensation oriented towards frequency regulation performance. Higher compensation fees are given to frequency regulation resources with timely response, fast adjustment rate, and high adjustment accuracy, while units with poor frequency regulation performance are subject to separate, itemized power assessments, such as... Figure 1 As shown, based on the frequency modulation start-up period, ramp-up period, or stabilization period, the response time, regulation rate, and regulation accuracy are ranked in that order, representing the comprehensive regulation performance indicators. The calculation formula is as follows.

[0053] .

[0054] The calculation methods for the three specific frequency regulation performance indicators are as follows.

[0055] .

[0056] in, It is the system's standard adjustment rate. It is the adjustment rate of the generator unit. This is the allowable adjustment deviation. It is the deviation of the unit. This is the standard response time. It is the response time for a unit to exit the dead zone.

[0057] Objective 2: Secondary frequency modulation cost of the joint frequency modulation system.

[0058] The cost of flywheel-thermal power combined secondary frequency regulation mainly includes the marginal cost of thermal power and the total cost of the flywheel system, and the calculation formula is as follows.

[0059] .

[0060] in, For the marginal cost of thermal power units, To achieve the levelized cost of electricity (LCOE) for flywheel energy storage systems.

[0061] The marginal cost of a thermal power unit refers to the additional cost incurred when it generates an additional 1 kWh of electricity during frequency regulation. This mainly includes additional fuel costs, additional maintenance costs, and equipment wear and tear costs caused by frequency regulation operations. As the cost per kilowatt-hour changes dynamically, the marginal cost of a 600MW reheat thermal power unit under typical operating conditions is calculated using the following formula.

[0062] .

[0063] The levelized cost of electricity (LCOE) for flywheel energy storage systems is the present value of costs over its lifecycle divided by the present value of electricity generated over its lifecycle. It reflects the internal average electricity price when the net present value (NPV) is zero, i.e., the break-even point for this investment. Energy storage costs include investment costs, operation and maintenance (O&M), and charging costs.

[0064] Investment costs mainly include capacity costs and power costs. Capacity costs refer to the equipment and construction costs related to the energy storage capacity in the energy storage system, while power costs refer to the equipment and construction costs related to the power in the energy storage system. The calculation formula is as follows.

[0065] .

[0066] in, The unit capacity cost (5000 yuan / kWh) Cost per unit power (2000 yuan / kW), For the rated capacity of energy storage, This is the rated power.

[0067] The operation and maintenance costs of flywheel energy storage mainly include labor, fuel and power, and component replacement costs, and the calculation formula is as follows.

[0068] .

[0069] in, For the first Annual maintenance costs The discount rate is 6%. The charging cost over the system's lifespan (20 years) is an important factor in calculating the cost per kilowatt-hour. The formula for calculating the charging cost based on the unified on-grid tariff of the Northwest Regional Power Grid System is as follows.

[0070] .

[0071] in, For the cost of energy storage charging, The electricity price for the power station during charging is 0.35 yuan / kWh. The charging and discharging efficiency of the energy storage system is taken as 92%. The levelized cost of electricity (LCOE) of the flywheel energy storage system can be calculated using equations 14, 15, and 16 as follows.

[0072] .

[0073] in, Discharge duration (h) at rated power.

[0074] Objective 3: Losses in the flywheel energy storage system.

[0075] A flywheel energy storage system consists of a flywheel body, a motor (electric motor / generator), a power electronic converter, a control system, and bearings. During flywheel operation, all components experience power losses: mechanical parts suffer from wind resistance and bearing friction losses; the motor experiences copper losses, hysteresis losses, and eddy current losses; and the power electronic converter experiences conduction and switching losses. The flywheel system's power command is... Then its actual output power The calculation formula is as follows.

[0076] .

[0077] Flywheel system losses and SOC Based on the functional relationship, the calculation formula for the power loss model of the flywheel system adopted in this application is as follows.

[0078] .

[0079] The formula for calculating the cumulative energy loss of a flywheel energy storage system is as follows.

[0080] .

[0081] The loss model parameters of a flywheel energy storage unit with a rated power of 500kW and a rated capacity of 125kW·h were obtained by least squares method: =0.05, =0.03, =16.25, =15.01.

[0082] (2) Satisfaction fuzzy membership function.

[0083] This application uses a multi-objective genetic algorithm to solve the Pareto optimal solution. The satisfaction level corresponding to each objective function of each Pareto solution is represented by a fuzzy membership function, and its calculation formula is as follows.

[0084] .

[0085] in, For the first Generations of populations The satisfaction value of the objective function at time j; For the first Generations of populations The first moment One objective function value; , For the first Generations of populations The first moment The maximum and minimum values ​​of the objective function are considered, and the closer the objective function is to 0, the higher the satisfaction level.

[0086] From the above, we can obtain The objective function of each Pareto solution at time t is satisfied by the matrix, and its calculation formula is as follows.

[0087] .

[0088] (3) Adaptive weight coefficient matrix.

[0089] Taking into account the urgency, economy, and equipment losses of power system frequency regulation, the adaptive weight matrix calculation formula for the objective function is as follows.

[0090] .

[0091] in, , , They are respectively Weighting coefficients for frequency regulation performance, frequency regulation cost, and energy storage system losses at all times.

[0092] Frequency deviation directly affects the stability of the power grid and the quality of power supply. As the basis for the dispatch center to initiate secondary frequency regulation, it represents the urgency of the power system's frequency adjustment needs. The larger the frequency deviation, the more urgent the frequency regulation, and the more important it is to pay attention to the system's frequency regulation performance. Therefore, this application determines the frequency deviation and frequency regulation performance based on their relationship. The calculation formula is as follows.

[0093] .

[0094] in, For power grid frequency deviation, This is the frequency deviation dead zone; This is the critical value for frequency deviation. The calculation formula is as follows.

[0095] .

[0096] in, for The power grid frequency at any time, This is the standard frequency for the power grid, typically 50Hz.

[0097] Similarly, and These are determined based on the ACE signal and SOC, respectively. Therefore, the comprehensive satisfaction weight matrix at time t can be obtained, calculated as follows.

[0098] .

[0099] S6. The frequency regulation power curve of the joint frequency regulation system, the frequency regulation capability of the thermal power unit, and the maximum charging and discharging power of the flywheel unit are used as boundary conditions for the output control of the joint system. The target layer-integrated satisfaction model is used as the judgment criterion. An improved multi-objective genetic algorithm is used to optimize the load allocation scheme of thermal power and flywheel output. By introducing dynamic iteration number and adaptive crossover probability, the computational efficiency is optimized while ensuring the optimization accuracy. The best thermal power and flywheel output allocation scheme is selected from the Pareto solution set.

[0100] Furthermore, the expression for the crossover probability of the improved multi-objective genetic algorithm is as follows.

[0101] .

[0102] In the formula, The crossover probability; and Don't specify the maximum and minimum crossover probabilities. This is the current fitness function value; and These are the average and minimum values ​​of the current fitness function, respectively.

[0103] Optionally, when the particle swarm gets trapped in a local extremum (i.e. It can increase Enhance global detection and reduce particle dispersion. Emphasis on localized mining This improves the quality and speed of frequency regulation responsibility allocation and increases optimization efficiency.

[0104] Furthermore, an improved multi-objective genetic algorithm is used to optimize the load allocation scheme of thermal power and flywheel output. By introducing dynamic iteration count and adaptive crossover probability, computational efficiency is optimized while ensuring optimization accuracy. The optimal thermal power and flywheel output allocation scheme is selected from the Pareto solution set, specifically including: 1) Based on the current AGC instructions, target time period, thermal power unit status and flywheel energy storage status, generate an initial population of an improved multi-objective genetic algorithm with thermal power and flywheel output weight coefficients as individuals, and select the best initial load allocation.

[0105] 2) Under the boundary conditions that satisfy the output control quantity of the joint system, the constraint layer restricts the boundary and rate of change of each individual in the population as an output scheme.

[0106] 3) Based on the target layer function, perform non-dominated sorting and crowding calculation on the individuals in the population, and select the top half of the high-quality individuals as the parents to guide the direction of population evolution.

[0107] 4) Perform adaptive crossover and mutation genetic operations on high-quality individuals to generate a new offspring population.

[0108] 5) Merge the parent and child generations and repeat the iteration until the maximum number of iterations is reached. Then, sort the Pareto optimal solution set by comprehensive satisfaction and select the best thermal power and flywheel output allocation scheme.

[0109] In practical applications, step S6 uses a multi-objective genetic algorithm to solve for the comprehensive optimal strategy for load allocation in the flywheel-thermal power system, and adaptively adjusts the key parameters of the algorithm based on actual frequency regulation requirements to balance accuracy and speed. The specific process is as follows.

[0110] (1) Number of dynamic iterations.

[0111] The higher the number of iterations in the core optimization factor of a genetic algorithm, the closer the optimization result is to the optimal value, but the longer the optimization time. Conversely, the more drastic the changes in AGC commands and the greater the frequency deviation, the greater the optimization difficulty. When both are relatively small, the output allocation can remain unchanged or change very little. In this application, the number of optimization iterations in the NSGA-II design increases with the rate of change of AGC commands and the frequency deviation, and vice versa. The formula for calculating the number of optimization iterations is as follows.

[0112] .

[0113] In the formula, To optimize the number of iterations; These are the initial iteration values; , Impact factor; , It is a constant.

[0114] (2) Improved crossover probability.

[0115] Crossover, as a key component of population genetics, plays a vital role in introducing diversity into the optimization and iteration process of algorithms, information sharing and propagation, and accelerating convergence. The crossover probability is the core parameter of the crossover operation. A lower crossover probability reduces population diversity and makes it easier to get trapped in local optima; conversely, an excessively high crossover probability can lead to excessive loss of diversity and disrupt existing favorable patterns.

[0116] First, based on the current AGC instructions, frequency regulation period, thermal power unit power, and flywheel energy storage system... Equal generation scale Individual vectors are [ , The initial population, of which This refers to the output weighting coefficient of the flywheel energy storage system operating in power regulation mode. The output ratio coefficient of thermal power units is used, and the optimal initial population individuals are used as the initial load allocation. Then, the boundaries and change rates of the population individuals are restricted by the constraint layer. Under the objective layer function, after individual non-dominated ranking and crowding degree calculation, the top [entities] are selected. The offspring generated by performing improved crossover and mutation on individuals of size 2 are merged into new parents. This process is iterated until the set conditions are met, resulting in a set of Pareto optimal solutions. Finally, the solution set is sorted by overall satisfaction, and the optimal solution is selected as the final instruction for the upper-level load allocation strategy. The optimal load allocation process is as follows: Figure 3 As shown.

[0117] S7. Based on the optimal power distribution scheme between thermal power and flywheel, issue power commands to the thermal power unit coordination control system and the flywheel energy storage array energy management system, and execute power response.

[0118] Furthermore, the power response specifically includes: adjusting the fuel quantity and valve opening according to the boiler power command and turbine power command in the power command, respectively; and controlling the output response of each individual flywheel energy storage array based on the state of each individual flywheel in the array.

[0119] As an optional implementation method, to verify the effectiveness of the proposed periodic dynamic weight target and lower-level array consistency control, a comparison strategy is designed based on the above: Comparison Strategy 1, the target layer only targets frequency regulation performance; Comparison Strategy 2, the weights of the three target layers remain fixed at 0.5, 0.3, and 0.2 respectively. Outlier processing and K-means clustering analysis were performed on the flywheel-thermal power plant's combined AGC frequency regulation data from July 2023 to March 2024. Simulations were conducted on a 60-minute high-frequency band and a typical 4-hour period. The results are as follows:Figures 4-5 As shown.

[0120] Depend on Figure 4 (a) Figure 5 (a) It can be seen that the joint frequency modulation system can effectively keep the maximum frequency deviation within the stable threshold range. Among them, the frequency modulation effect of strategy 1, which only focuses on frequency modulation performance, is relatively the best, followed by the frequency modulation effect of the strategy in this application, while strategy 2, which has a fixed target layer weight, has a relatively longer maximum frequency deviation and recovery time under larger load disturbances.

[0121] Depend on Figure 4 (b) Figure 4 (c) Figure 5 (b) Figure 5 (c) It can be seen that the joint frequency regulation system can effectively keep the maximum frequency deviation within the stable threshold range. When the target layer only has frequency regulation performance indicators, the thermal power unit and the flywheel energy storage system will try to improve the performance indicators of the period by outputting as much power as possible in different response stages. That is, the frequency regulation effect under strategy 1 is relatively the best; the frequency regulation effect of the strategy in this application is second. When the target layer weight is unchanged, the flywheel energy storage system tends to have a larger output during the start-up period and a smaller output during the ramp-up period. At the same time, the output is relatively low under high frequency regulation demand, resulting in a relatively longer maximum frequency deviation and recovery time under large load disturbance.

[0122] Table 1 shows the statistical results of each strategy at the target level in three periods, where the statistical values ​​for each period are the weighted average of the frequency modulation period. It can be seen that compared to the control strategies of this application, Strategy 1 improves various frequency modulation performance indicators, but the frequency modulation cost and flywheel system losses increase significantly, especially during the ramp-up period in the typical 4-hour period, requiring an increase of 25.03% in frequency modulation cost and 36.12% in system losses to improve performance indicators by only 2.74%; Strategy 2… The values ​​were relatively low, decreasing by 5.70% and 6.13% in the high-frequency range and typical period, respectively.

[0123] Table 1 Statistical results of the target layer at different time periods in Experiment 1

[0124] Table 2 shows the statistical results for a typical 4-hour period. It can be seen that Strategy 1 improved the frequency modulation performance index by 2.54%. This also increases frequency modulation costs by 22.42% and flywheel system losses by 36.43% under Strategy 2. The cost was reduced by 3.23%, while the frequency regulation cost and loss were reduced by only 4.48% and 6.34%, respectively. In comparison, the strategy proposed in this application is the best in terms of comprehensively improving grid frequency stability, cost control, and system energy efficiency. Moreover, it reduces the number of times thermal power units are assessed, while the net benefit from grid frequency regulation compensation is the largest, which is 43.59% and 15.37% higher than strategies 1 and 2, respectively.

[0125] Table 2 Statistical results of target layer throughout the entire time period

[0126] As another possible approach, Comparison Strategy 1 is adopted: a two-layer fuzzy control strategy for coordinated response of frequency regulation units to AGC commands under dual-signal distribution mode. The first-layer fuzzy controller targets the grid frequency regulation demand, determining the participation factor of energy storage based on the relationship between total power response and grid frequency difference demand under both ARR and ACE modes. The second-layer fuzzy controller considers the output of the first-layer fuzzy control and the real-time state of charge (SOC) of the flywheel group as inputs, deriving the relationship between the real-time SOC of energy storage and the participation factor, and determining the actual depth of output for each group. Comparison Strategy 2 is a secondary frequency regulation load distribution control strategy for energy storage coordinated with thermal power units based on successive variational mode decomposition (SVMD). This strategy uses SVMD to decompose frequency regulation commands into signals of different frequency bands, and performs high- and low-frequency segmentation and reconstruction of the mode functions. With economic efficiency as the goal, short-term high-frequency signals are sent to the flywheel energy storage system for response, while long-term low-frequency signals are sent to the thermal power unit for response. The allocation of groups and individual units within the flywheel array follows the SOC ratio principle. Outlier handling and K-means clustering analysis were performed on the flywheel-thermal power plant's combined AGC frequency regulation data from July 2023 to March 2024. A typical 4-hour period was selected for simulation, and the results are as follows: Figure 6 As shown.

[0127] Depend on Figure 6 It is known that adding flywheel energy storage can significantly improve the system frequency regulation effect. The strategy of this application can dynamically adjust the output power of each period according to the grid assessment requirements and target level, and charge and discharge the flywheel energy storage during non-assessment periods to restore its own power to the optimal range, effectively reducing the maximum frequency deviation and shortening the frequency regulation time. Compared with Strategy 1, which combines the advantages of ARR and ACE modes, it flexibly adjusts the output power of the energy storage group according to the frequency deviation, and adjusts the energy storage absorption or release power when the thermal power unit overshoots, which can better suppress frequency degradation and promote its recovery speed, but its effect of suppressing frequency difference is slightly inferior. Compared with Strategy 2, which sends high and low frequency signals to the two systems respectively, the flywheel energy storage has a large response power in the early stage of frequency regulation and then gradually decreases to 0, with large frequency regulation deviation and time.

[0128] Table 1 shows the statistics of frequency modulation results of the experimental system, in which... , In order, the maximum frequency deviation and the standard deviation of the frequency deviation are: , , , In order, the values ​​are: average start-up time, average gradeability, and average overshoot of thermal power units. , The values ​​are, in order, the maximum main steam pressure difference and its standard deviation.

[0129] As shown in Table 3, the joint frequency modulation system under the control allocation strategy of this application has the best overall frequency modulation effect. During high-frequency and typical periods, compared to Comparative Strategy 1, Comparative Strategy 2, and individual thermal power units, the efficiency was reduced by 10.63%, 20.88%, and 36.41%, and by 6.61%, 21.41%, and 40.40%, respectively. Simultaneously, the start-up time, ramp rate, and overshoot of the thermal power units were significantly reduced. In particular, the average ramp rate under the strategy of this application was reduced by 18.99%, 29.66%, and 46.18%, and by 19.67%, 32.85%, and 44.51%, respectively, compared to Comparative Strategy 1, Comparative Strategy 2, and individual thermal power units. The percentages decreased by 5.41%, 28.17%, 34.07% and 5.98%, 28.80%, 34.83% respectively. This is mainly because, under the control strategy of this application, the flywheel energy storage can retain energy during the start-up period to improve the frequency regulation performance index by rapidly charging and discharging at high power during the ramp-up period.

[0130] Table 3 Statistics on System Frequency Modulation

[0131] Figure 7 The statistical results for the target layer during a typical 4-hour period show that, during the startup phase, both comparative strategies 1 and 2 respond to frequency regulation with high output power. The response time index values ​​of comparative strategies 1 and 2 are relatively high, reaching above 1.5 in most frequency regulation periods, with the highest value reaching above 1.7. The response time index value under the strategy of this application is slightly lower, mainly distributed in the range of 1.4 to 1.6. During the ramp-up phase, the regulation rate index value under the strategy of this application is mainly distributed at 1.4 and above, while that of comparative strategies 1 and 2 is relatively low, especially comparative strategy 2, where 25% of the statistical values ​​are below 1.2. During the adjustment phase, based on the output characteristics analysis of thermal power under different operating conditions, this application utilizes flywheel energy storage to suppress the overshoot of thermal power output, which can effectively improve the regulation accuracy index value to around 1.6. In contrast, comparative strategies 2 and 3 mainly adjust based on factors such as frequency difference and SOC, resulting in relatively low regulation accuracy index values, mainly distributed around 1.35. Thermal power units, due to their large inertia, have very low values ​​for all three performance indicators and are subject to more frequent assessments.

[0132] Table 2 shows the target layer and benefit statistics for Experiment 2 throughout the entire time period. Overall, this application improves performance at the cost of relatively small increases in frequency modulation costs and flywheel system losses. The value reached above 1.45, representing an improvement of 5.65%, 7.41%, and 26.76% compared to Comparison Strategies 1 and 2, and 5.43%, 7.30%, and 26.52% for standalone thermal power plants in the two time periods. According to the two detailed rules of the power grid, in the typical 4-hour period, the grid compensation revenue under this application strategy reached 29,638.35 CNY, with a net profit of 7,752.761 CNY, representing an improvement of 38.007%, 32.86%, and 53.11% compared to Comparison Strategies 1 and 2, and standalone thermal power plant frequency regulation, respectively. Simultaneously, the number of times the AGC performance assessment failed was significantly reduced, greatly minimizing the economic penalties for power plants failing AGC assessments.

[0133] Table 4 Evaluation System for System Objectives

[0134] To promote the grid integration and consumption of renewable energy and improve the frequency regulation performance and economy of thermal power units, this application proposes a two-layer load distribution control strategy combining flywheel energy storage with secondary frequency regulation of thermal power units, based on the periodic assessment and compensation details of the power grid secondary frequency regulation, and considering the characteristics of a 660MW subcritical drum boiler unit and a large-scale flywheel array, using multi-objective genetic algorithms and consensus algorithms.

[0135] (1) The strategy of this application utilizes the fast response characteristics of the flywheel energy storage system and flexibly charges and discharges at different times based on the conditions such as the load reserve of thermal power units. This effectively reduces the frequency deviation of the power grid and the main steam pressure fluctuation and ramp rate of the thermal power units, enhances the stability of the power grid frequency, and reduces the frequency regulation pressure and loss of the thermal power units.

[0136] (2) The power optimization layer dynamically adjusts the weights of the three-layer objectives based on the grid frequency deviation, frequency regulation power demand, and the remaining energy of the flywheel array, thereby flexibly adjusting the power output of the two systems at different times. Overall, this strategy can improve the average frequency regulation performance index. The cost of frequency regulation is 2.7 times that of a standalone thermal power unit, while the cost of frequency regulation is reduced to less than 1.7 times and the flywheel system loss to less than 1.4 times. According to the grid assessment and compensation details, the net benefit of the joint frequency regulation unit from grid compensation has also increased by nearly 95%.

[0137] The technical effects of this application are as follows.

[0138] This application, based on the power grid frequency regulation specification (detailed assessment rules for power grid frequency regulation in different periods), aims to achieve optimal frequency regulation performance and efficient system operation with minimal resource consumption while ensuring power grid frequency stability, thereby improving the overall system performance and economy. During power optimization, the application analyzes the periodic assessment targets for power grid frequency regulation demand of thermal power units; uses LSTM prediction and flywheel traction experiments to estimate the frequency regulation capabilities of thermal power units and flywheel energy storage; establishes a frequency regulation cost model and a flywheel system loss model for the joint frequency regulation system; and establishes a comprehensive satisfaction model by using corresponding power grid assessment indicators, frequency regulation costs, and flywheel losses as three-tiered objectives at different periods, setting dynamic weights for the objectives based on indicators such as power grid frequency deviation, flywheel system SOE, and power demand; finally, an improved multi-objective genetic algorithm is used to solve for the output ratio coefficients of flywheel energy storage and thermal power units, and the results are distributed to the thermal power and flywheel system for power response execution. This application simultaneously improves the safety, efficiency, and economy of flywheel energy storage-thermal power combined secondary frequency regulation. Combined with the internal energy management of large-scale flywheel arrays, it reduces frequency regulation costs and flywheel system losses while ensuring frequency regulation performance. It can be widely applied in engineering projects of large-scale flywheel energy storage arrays in power generation units.

[0139] Example 2: This application also provides a computer device, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 9 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores and processes data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements the methods described above.

[0140] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0141] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0142] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0143] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method for grid frequency modulation specifications, characterized in that, The method comprises: According to the grid frequency modulation specification, the frequency modulation process of the complete response primary AGC of the thermal power is divided into multiple target periods based on the frequency modulation performance index; the target period includes: start-up period, climbing period, adjustment period and non-assessment period; Each period is assessed and compensated based on a preset performance index, and a frequency modulation power curve of the joint frequency modulation system is set according to the target value of the performance index; the preset performance index includes: response time, adjustment rate and adjustment accuracy; The LSTM model is used to predict the frequency modulation capacity of the thermal power unit; The historical flywheel monomer charging and discharging power is linearly fitted based on a logistic regression function, and the maximum charging and discharging power of the flywheel monomer is predicted; Three target layers including the grid AGC frequency modulation performance index, the secondary frequency modulation cost of the joint frequency modulation system and the loss of the flywheel energy storage system are constructed, the degree of achievement of each target layer is quantified by a satisfaction fuzzy membership function, and the target weight is dynamically adjusted by using an adaptive weight coefficient matrix to construct a target layer-comprehensive satisfaction model; The frequency modulation power curve of the joint frequency modulation system, the frequency modulation capacity of the thermal power unit and the maximum charging and discharging power of the flywheel monomer are taken as the boundary conditions of the output control amount of the joint system, the target layer-comprehensive satisfaction model is taken as the judgment criterion, an improved multi-objective genetic algorithm is used to optimize the load distribution of the thermal power and the flywheel output distribution scheme, the dynamic iteration number and the adaptive crossover probability are introduced to optimize the calculation efficiency while ensuring the optimization accuracy, and the best thermal power and flywheel output distribution scheme is selected from the Pareto solution set; According to the best thermal power and flywheel output distribution scheme, power instructions are issued to the thermal power unit coordinated control system and the flywheel energy storage array energy management system, and power response is performed.

2. The grid frequency regulation oriented flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The LSTM model comprises a primary sub-model and a secondary sub-model, the primary model predicts the main steam pressure and temperature based on historical working condition parameter data; The secondary model predicts the actual power generation output of the thermal power unit based on the main steam pressure and temperature; The historical working condition parameter data includes: coal quantity, air supply quantity, feed water flow, main steam pressure and main steam temperature.

3. The grid frequency regulation oriented flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The calculation formula of the maximum charging and discharging power of the flywheel monomer is as follows: ； wherein Pmax is the maximum charging power of the flywheel unit; Pmin is the maximum discharging power of the flywheel unit; and SOCmin and SOCmax represent the minimum and maximum values of SOC, respectively; and Pnom and Pnom represent the nominal charging and discharging power of the flywheel unit, respectively; is a proportionality factor; n is a rate factor.

4. The flywheel energy storage-coal power plant combined secondary frequency modulation multi-objective control method for grid frequency modulation specifications according to claim 1, characterized in that, The calculation formula of the grid AGC frequency modulation performance index is as follows: ； wherein is the standard adjustment rate, is the adjustment rate of the unit, is the unit's deviation, is the allowed adjustment deviation, is the response time of the unit to exit the dead zone, is the standard response time.

5. The grid frequency regulation oriented flywheel energy storage-coal power plant combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The calculation formula of the secondary frequency modulation cost of the joint frequency modulation system is as follows: ； In the formula, is the quadratic frequency cost of the joint frequency modulation system; is the marginal cost of the thermal power unit at time t; is the real-time power of the thermal power unit; is the flat-rate electricity cost of the flywheel energy storage system at time t; is the output power of the energy storage at time t; is the investment cost of the energy storage; is the operation and maintenance cost of the energy storage; is the power station electricity price during charging; is the rated capacity of the energy storage; is the charging and discharging efficiency of the energy storage system; is the total time; is the real-time on-grid electricity price; is the unit capacity cost; is the rated power; is the system life; is the operation and maintenance cost of the energy storage in the year; is the discount rate; is the charging cost of the energy storage; is the discharging duration at the rated power.

6. The grid frequency regulation oriented flywheel energy storage-coal power plant combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The expression of the crossover probability of the improved multi-objective genetic algorithm is as follows: ； where is the cross probability; and are the maximum and minimum values of the cross probability, respectively, is the current fitness function value; and are the average and minimum values of the current fitness function, respectively.

7. The grid frequency regulation oriented flywheel energy storage-coal power plant combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The improved multi-objective genetic algorithm is used to optimize the load distribution of the thermal power and the flywheel output distribution scheme, the dynamic iteration number and the adaptive crossover probability are introduced to optimize the calculation efficiency while ensuring the optimization accuracy, and the best thermal power and flywheel output distribution scheme is selected from the Pareto solution set, which specifically comprises: According to the current AGC instruction, the target period, the state of the thermal power unit and the state of the flywheel energy storage, an initial population of the improved multi-objective genetic algorithm is generated with the weight coefficient of the thermal power and the flywheel output as the individual, and the initial load distribution is optimally set; Under the boundary conditions of the joint system output control amount, each individual in the constraint layer is taken as the boundary and the change speed of an output scheme; The target layer function is used to sort and calculate the crowding degree of the population individuals, and the top half of the high-quality individuals are selected as the parents to guide the evolution direction of the population; The high-quality individuals are subjected to adaptive crossover and mutation genetic operations to generate a new population of offspring; The parent and offspring populations are combined and iterated repeatedly until the maximum number of iterations is reached, and the obtained Pareto optimal solution set is subjected to comprehensive satisfaction sorting to select the best thermal power and flywheel output allocation scheme.

8. The grid frequency regulation oriented flywheel energy storage-coal power plant combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The power response specifically includes: based on the boiler power instruction and the turbine power instruction in the power instruction, adjusting the fuel quantity and the valve opening degree respectively, and the energy management system of the flywheel energy storage array controls the output of each flywheel unit based on the state of each flywheel unit in the array.

9. The grid frequency regulation oriented flywheel energy storage-coal power plant combined secondary frequency modulation multi-objective control method according to claim 1, characterized in that, The linear fitting is least square fitting.

10. A computer device comprising: The memory, the processor, and the computer program stored in the memory and executable on the processor are characterized in that the processor executes the computer program to implement the flywheel energy storage-thermal power combined secondary frequency modulation multi-objective control method for grid frequency modulation specification of any one of claims 1-9.

Images