Control method and apparatus for assisting frequency regulation of thermal power generating units with molten salt energy storage

Abstract

A control method and apparatus is provided for assisting frequency regulation of a thermal power generating unit with molten salt energy storage. The method includes the steps of obtaining a real-time temperature of the molten salt at the inlet of an electric heater, inputting the real-time temperature of the molten salt at the inlet of the electric heater, the set temperature of the molten salt at the outlet of the electric heater, and the electric heater power into a trained neural network model of the molten salt flow rate to obtain a desired target molten salt flow rate, and controlling the molten salt flow rate for supporting the frequency regulation of a thermal power generation unit based on the target molten salt flow rate, thereby effectively realizing fast and accurate control of the molten salt flow rate in the process of supporting the frequency regulation of a thermal power generation unit and avoiding large changes in the molten salt temperature at the outlet of the electric heater and overheating decomposition caused by delays in molten salt flow rate control and delayed control response, thereby meeting the demands of practical application.

Description

[Technical Field]

[0001] The present disclosure relates to the technical field of frequency regulation of thermal power generating units, and more particularly to a control method and apparatus for assisting frequency regulation of thermal power generating units with molten salt energy storage. [Background technology]

[0002] In recent years, molten salt energy storage technology has attracted widespread attention in the field of renewable energy as an efficient thermal energy storage method, and its core principle is to use molten salts as a thermal energy carrier to achieve energy storage and conversion by absorbing and releasing heat. This technology has been applied for the first time to solar thermal power generation systems to solve the problems of solar power generation being limited by weather and time, and to ensure the stability and continuity of power supply.

[0003] The working process of molten salt energy storage mainly consists of two stages: energy storage and heat release. In the heat storage stage, molten salt is heated to a high temperature by a heat source such as solar energy or an electric heater, and the thermal energy of the molten salt is stored. In the heat release stage, when the thermal energy needs to be used, the molten salt transfers the thermal energy to a working medium such as steam through a heat exchanger, allowing the thermal energy to be used for power generation, heating industrial processes, etc.

[0004] Molten salt has significant advantages in energy storage. First, it has a high specific heat capacity and thermal stability, allowing it to store large amounts of thermal energy at high temperatures and is resistant to decomposition. Second, the efficiency of molten salt energy storage systems is high, at over 95%, with little energy loss. Furthermore, molten salt energy storage systems have low operating costs, are easy to maintain, and have a long service life of several decades. Finally, molten salt energy storage can effectively smooth out the intermittent output of renewable energy and improve the stability and reliability of the power grid, which is of great significance for the construction of a new type of power system based on new energy sources.

[0005] However, molten salt energy storage technology also faces many challenges, such as the selection and optimization of molten salt materials, durability issues of high-temperature equipment, and the complexity of system integration and control technology. With the continuous progress in materials science, thermal engineering, and control system technology, molten salt energy storage is expected to be applied in a wider range of fields and become an important force promoting energy transition and sustainable development.

[0006] A typical method for applying a molten salt energy storage system to a thermal power generation unit is as follows: electricity is taken in at the generator outlet of the generator unit or at a booster station, and the electrical energy is converted into thermal energy through an electric heater and stored; in the heat dissipation stage, the heat of the molten salt is transferred to steam through a heat exchanger, and the steam is supplied to the outside or returned to the unit's thermal system; in the energy storage stage, the heating power of the electric heater is adjusted, thereby changing the power transported from the generator unit to the power grid, allowing the generator unit to participate in the frequency regulation service of the power grid and achieving a significant frequency regulation effect.

[0007] The key equipment in the energy storage stage of a molten salt energy storage system is the molten salt electric heater, which converts electrical energy into thermal energy. The molten salt flows into the inlet of the electric heater, is heated, and then flows out of the outlet of the electric heater. Therefore, to achieve maximum heat storage, the temperature of the molten salt at the outlet of the electric heater must be stable. Furthermore, to ensure the safety and stability of the molten salt energy storage system, the temperature of the molten salt must not exceed its decomposition temperature.

[0008] The outlet molten salt temperature of an electric heater is primarily related to the inlet molten salt temperature, the heater's power, and the molten salt flow rate. When a molten salt energy storage system is used to regulate the frequency of a thermal power generation unit, the unit's automatic generation control (AGC) command is a nearly square wave. To keep up with this change in input power, the molten salt energy storage system's power also changes rapidly over a short period of time. This necessitates improving the unit's AGC response speed and accuracy. In this case, conventional technologies can only ensure stable outlet molten salt temperature by adjusting the rapid changes in the molten salt flow rate. However, due to the large specific heat capacity and slow flow rate of molten salt, existing control methods such as proportional-integral-derivative (PID) control suffer from control delays and slow control response, resulting in large fluctuations in the outlet molten salt temperature and a tendency for overheating and decomposition. Summary of the Invention

[0009] The present disclosure aims to solve at least one of the problems in the prior art by providing a control method and apparatus for assisting frequency regulation of a thermal power generation unit using molten salt energy storage.

[0010] In one aspect of the present disclosure, there is provided a control method for assisting frequency regulation of a thermal power generation unit with molten salt energy storage, the control method comprising: Obtaining a real-time temperature of the molten salt at the inlet of the electric heater; Inputting the real-time temperature of the molten salt at the inlet of the electric heater, the set temperature of the molten salt at the outlet of the electric heater, and the electric power of the electric heater into a trained neural network model of the molten salt flow rate to obtain a desired target molten salt flow rate; and controlling the molten salt flow rate to assist in frequency regulation of the thermal power generation unit based on the target molten salt flow rate.

[0011] Optionally, training the neural network model of the molten salt flow rate comprises: Establishing a neural network model of the molten salt flow rate using the molten salt temperature at the outlet of the electric heater, the molten salt temperature at the inlet of the electric heater, and the electric power of the electric heater as input parameters and the molten salt flow rate as an output parameter; obtaining sample data corresponding to each of the input parameters and the output parameters under different steady-state operating conditions to construct a training data set; and training a neural network model of the molten salt flow rate using the training data set.

[0012] Optionally, the neural network model of the molten salt flow rate comprises a radial basis function network model, the radial basis function network model comprising an input layer, a hidden layer and an output layer; The output data of the hidden layer can be expressed as:

number

number

number

number

number

number

number

[0013] Optionally, the step of training a neural network model of the molten salt flow rate using the training data set comprises: and performing iterative calculations on the Gaussian function center vector, width vector, and weighting coefficient of each hidden layer node using the training data set and corresponding iterative formula based on corresponding initial values ​​until the root mean square error output from the radial basis function network model is smaller than a preset accuracy.

[0014] Optionally, the initial value corresponding to the center vector of said Gaussian function is expressed as:

number

number

number

number

number

number

number

number

number

number

number

number

[0015] Optionally, the iterative formula corresponding to said weighting factors is expressed as:

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

[0016] Optionally, the root mean square error output from the radial basis function network model may be expressed as RMS, where RMS is:

number

[0017] Optionally, the step of training a neural network model of the molten salt flow rate using the training data set comprises: The method further includes normalizing input samples included in the training data set before the step of performing iterative calculations on the Gaussian function center vector, width vector, and weighting coefficient of each hidden layer node using the training data set and corresponding iterative formula based on corresponding initial values ​​until the root mean square error output from the radial basis function network model is smaller than a predetermined accuracy.

[0018] Optionally, the step of controlling the molten salt flow rate to assist in frequency regulation of a thermal power generation unit based on the target molten salt flow rate comprises: Using a molten salt flow meter in combination with a variable frequency molten salt pump and adjusting the frequency of the variable frequency molten salt pump to adjust the molten salt flow rate to the target molten salt flow rate; or The method includes using the molten salt flow meter in combination with an adjustment valve and adjusting the opening degree of the adjustment valve to adjust the flow rate of the molten salt to the target flow rate of the molten salt.

[0019]

[0013] In another aspect of the present disclosure, there is provided a controller for assisting frequency regulation of a thermal power generating unit with molten salt energy storage, the controller comprising: an acquisition module used to acquire the inlet molten salt real-time temperature of the electric heater; an input module used to input the inlet molten salt real-time temperature of the electric heater, the set outlet molten salt temperature of the electric heater, and the electric heater power into the trained neural network model of molten salt flow rate to obtain a desired target molten salt flow rate; and a control module used to control the molten salt flow rate to assist in frequency regulation of the thermal power generation unit based on the target molten salt flow rate.

[0020] Compared to the prior art, the present disclosure uses a trained neural network model of molten salt flow rate to obtain a desired target molten salt flow rate based on the real-time molten salt temperature at the inlet of the electric heater, the molten salt temperature at the outlet of the installed electric heater, and the power of the electric heater, and controls the molten salt flow rate that assists in the frequency adjustment of the thermal power generation unit based on the target molten salt flow rate. This effectively achieves fast and accurate control of the molten salt flow rate in the process of assisting in the frequency adjustment of the thermal power generation unit, avoids large changes in the molten salt temperature at the outlet of the electric heater and overheating decomposition caused by delays in molten salt flow rate control and delays in control response, and meets the demands for practical application. [Brief explanation of the drawings]

[0021] One or more embodiments are illustratively described by the accompanying drawings, which are not intended to be limiting of the embodiments, in which elements having the same reference numerals in the drawings are similar elements, and in which the drawings are not intended to be drawn to scale unless otherwise specified. [Figure 1] 1 is a flowchart of a control method for assisting frequency regulation of a thermal power generation unit with molten salt energy storage provided in one embodiment of the present disclosure. [Figure 2] FIG. 1 is a schematic diagram illustrating the structure of a molten salt energy storage system provided in another embodiment of the present disclosure. [Figure 3] FIG. 10 is a schematic diagram illustrating the training process of a neural network model of molten salt flow rate provided in another embodiment of the present disclosure. [Figure 4] FIG. 10 is a schematic diagram illustrating a neural network model of molten salt flow rate and a molten salt flow rate control process provided in another embodiment of the present disclosure. [Figure 5] FIG. 10 is a schematic diagram illustrating the structure of a control device that assists frequency regulation of a thermal power generation unit by molten salt energy storage, provided in another embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0022] In order to clarify the objectives, technical solutions, and advantages of the embodiments of the present disclosure, the following detailed description of each embodiment is provided in conjunction with the drawings. However, those skilled in the art will understand that many technical details are provided in each embodiment of the present disclosure to help readers better understand the present disclosure. However, the technical solutions claimed to be protected by the present disclosure can be achieved without these technical details and various changes and modifications based on the following embodiments. The division of the following embodiments is for the convenience of explanation and does not limit the specific implementation of the present disclosure. The embodiments may be combined with and referenced to each other, provided that they are not mutually inconsistent.

[0023] An embodiment of the present disclosure relates to a control method for assisting frequency regulation of a thermal power generation unit by molten salt energy storage, and the process includes the following steps S110, S120 and S130, as shown in FIG.

[0024] In step S110, the real-time temperature of the molten salt at the inlet of the electric heater is obtained.

[0025] Specifically, when the molten salt energy storage system is operating normally, in step S110, the molten salt real-time temperature of the electric heater in the molten salt energy storage system may be directly read.

[0026] As shown in Figure 2, a molten salt energy storage system typically includes a molten salt pipe, a molten salt pump arranged in sequence on the molten salt pipe, a regulating valve, and an electric heater. The molten salt pump is combined with a frequency converter to form a variable-frequency molten salt pump. By adjusting the frequency of the frequency converter, the molten salt pump can be used to regulate the flow rate of the molten salt in the molten salt pipe. Furthermore, by adjusting the aperture of the regulating valve with an electric actuator, the flow rate of the molten salt in the molten salt pipe can also be adjusted using the aperture of the regulating valve.

[0027] In step S120, the real-time molten salt temperature at the inlet of the electric heater, the set molten salt temperature at the outlet of the electric heater, and the electric heater power are input into the trained neural network model of molten salt flow rate to obtain the desired target molten salt flow rate.

[0028] Specifically, in conjunction with FIG. 2, during normal operation of the molten salt energy storage system, in step S120, the inlet temperature T of the electric heater obtained in real time from the molten salt energy storage system in The real-time temperature of the molten salt at the inlet of the electric heater may be used as the real-time temperature of the molten salt at the outlet of the electric heater, i.e., the outlet temperature T of the electric heater in FIG. in , and the power of the electric heater can be set according to actual requirements.

[0029] Illustratively, as shown in FIG. 3, the training process of the neural network model of molten salt flow rate includes the following steps S310, S320, and S330.

[0030] In step S310, a neural network model of the molten salt flow rate is established using the molten salt temperature at the outlet of the electric heater, the molten salt temperature at the inlet of the electric heater, and the electric heater power as input parameters, and the molten salt flow rate as an output parameter.

[0031] The neural network model of the molten salt flow rate may be a multi-input, single-output neural network model. Exemplarily, the neural network model of the molten salt flow rate includes a radial basis function (RBF) network model. In conjunction with FIG. 4, the RBF network model includes an input layer, a hidden layer, and an output layer. Here, the input layer may include three nodes, each corresponding to an input parameter. For example, as shown in FIG. 4, if the input parameters corresponding to each node in the input layer are denoted as x1, x2, and x3, respectively, x1, x2, and x3 may correspond to the molten salt temperature at the outlet of the electric heater, the molten salt temperature at the inlet of the electric heater, and the power of the electric heater, respectively.

[0032] The hidden layer is used to realize nonlinear mapping. In conjunction with Figure 4, the output data of each node in the hidden layer is

number

number

number

number

number

number

[0033] Here, j represents the node number in the hidden layer, j=1, 2, ..., q.

number

number

number

number

number

number

number

number

number

number

[0034] The output data of the output layer is expressed as follows:

number

[0035] where y represents the molten salt flow rate,

number

[0036] Using the output data corresponding to the hidden layer and the output layer, the mapping between the input parameters, i.e., the outlet molten salt temperature of the electric heater, the inlet molten salt temperature of the electric heater, and the electric heater power, and the output parameter, i.e., the molten salt flow rate, is established. In the training stage, the RBF network model is trained based on a large number of actual operating data samples including the input parameters and output parameters, thereby:

number

number

number

[0037] In step S320, sample data corresponding to each of the input parameters and output parameters in different steady-state operating states are obtained to construct a training data set.

[0038] Specifically, since data in the steady operating state of the molten salt energy storage system can accurately and realistically reflect the molten salt energy storage situation, the sample data for constructing the training dataset should be selected from the datasets of the molten salt temperature at the outlet of the electric heater, the molten salt temperature at the inlet of the electric heater, the electric heater power, and the molten salt flow rate in different steady operating states of the molten salt energy storage system, and should not be selected from data in the variable operating states of the molten salt energy storage system.

[0039] In step S330, the training data set is used to train a neural network model of molten salt flow rate.

[0040] Illustratively, step S330 includes: using a training data set and a corresponding iterative formula based on corresponding initial values, iteratively calculating the Gaussian function center vector, width vector, and weighting coefficient of each hidden layer node until the root mean square error output from the radial basis function network model is smaller than a preset accuracy.

[0041] Illustratively, the initial value corresponding to the center vector of the Gaussian function is expressed as follows:

number

[0042] Here, i represents the node number of the input layer, i=1, 2, ..., p, p represents the number of nodes in the input layer,

number

number

number

[0043] The initial value corresponding to the width vector is expressed as follows:

number

[0044] where:

number

number

number

number

[0045] The initial values ​​corresponding to the weighting coefficients are expressed as follows:

number

[0046] where:

number

number

[0047] Based on the initial values ​​of the Gaussian function center vector, width vector, and weighting coefficients, during the model training stage, the output data of the output layer can be expressed as:

number

number

number

number

number

number

number

[0048] Specifically, the iterative formula corresponding to the weighting coefficients is expressed as follows:

number

[0049] Here, t represents the number of iterations.

number

number

number

number

number

number

number

[0050] where:

number

number

number

[0051] The iteration formula corresponding to the width vector is expressed as follows:

number

[0052] where:

number

number

number

[0053] In each of the above iterative equations, E represents the output bias evaluation function, which is expressed as follows:

number

[0054] Here, n represents the sample number, n=1, 2, 3, ... , N. N represents the total number of samples.

number

number

[0055] For example, the root mean square error output from the RBF network model is

number

number

number

[0056] Specifically, the preset accuracy

number

number

[0057] After the training of the RBF network model is completed and a trained neural network model of molten salt flow rate is obtained, in step S120, the real-time molten salt temperature at the inlet of the electric heater, the set molten salt temperature at the outlet of the electric heater, and the electric heater power can be used as input parameters of the neural network model of molten salt flow rate, i.e., the corresponding input values ​​of the molten salt temperature at the inlet of the electric heater, the molten salt temperature at the outlet of the electric heater, and the electric heater power, respectively, so that the desired target molten salt flow rate can be obtained using the trained neural network model of molten salt flow rate.

[0058] In particular, step S330 further includes: normalizing the input samples included in the training data set before the step of performing iterative calculations on the Gaussian function center vectors, width vectors, and weighting coefficients of each hidden layer node using the training data set and the corresponding iterative formula based on the corresponding initial values, until the root mean square error output from the radial basis function network model is smaller than a preset accuracy.

[0059] Specifically, since the input samples corresponding to each input parameter typically correspond to different units and have large differences in magnitude, it is necessary to normalize the input samples to unify their formats to eliminate the differences in units and magnitudes.

[0060] The actual operating value of the input sample corresponding to the i-th input layer node is

number

number

number

number

number

number

number

number

number

[0061] In step S130, the flow rate of the molten salt that assists in frequency adjustment of the thermal power generation unit is controlled based on the target flow rate of the molten salt.

[0062] Specifically, step S130 can be controlled by various control methods. Exemplarily, step S130 may include using a molten salt flow meter in combination with a variable frequency molten salt pump and adjusting the frequency of the variable frequency molten salt pump to adjust the molten salt flow rate to the target molten salt flow rate. For example, referring to Fig. 4, in step S130, a molten salt flow meter may be used in combination with a variable frequency molten salt pump (i.e., the variable frequency pump of Fig. 4) to adjust the molten salt flow rate. Specifically, a flow rate signal may be provided by PID control, and the frequency of a frequency converter of the variable frequency molten salt pump may be adjusted based on the flow rate signal to adjust the molten salt flow rate to the desired target molten salt flow rate.

[0063] Illustratively, step S130 may further include using a molten salt flowmeter in combination with an adjusting valve and adjusting the aperture of the adjusting valve to adjust the molten salt flow rate to a target molten salt flow rate. Referring to Fig. 4, for example, in step S130, the molten salt flow rate may be adjusted using a molten salt flowmeter in combination with an adjusting valve. Specifically, a flow rate signal may be provided using a PID control method, parameter settings of an electric actuator may be adjusted based on the flow rate signal, and the set electric actuator may be used to adjust the aperture of the adjusting valve to adjust the molten salt flow rate to the desired target molten salt flow rate.

[0064] The flow rate of the molten salt responds quickly to the adjustment of the variable frequency molten salt pump or the valve opening of the regulating valve. The frequency of the variable frequency molten salt pump can be quickly adjusted by adjusting the corresponding frequency converter, and the valve opening of the regulating valve can be quickly adjusted by setting the parameters of the electric actuator. Therefore, the flow rate of the molten salt can be quickly and accurately controlled using conventional PID control, which meets the requirements of practical application.

[0065] In the control method for assisting the frequency adjustment of a thermal power generation unit using molten salt energy according to an embodiment of the present disclosure, compared to the prior art, a desired target molten salt flow rate is obtained using a trained molten salt flow rate neural network model based on the real-time molten salt temperature at the inlet of the electric heater, the molten salt temperature at the outlet of the installed electric heater, and the power of the electric heater, and the molten salt flow rate for assisting the frequency adjustment of the thermal power generation unit is controlled based on the target molten salt flow rate, thereby effectively realizing fast and accurate control of the molten salt flow rate in the process of assisting the frequency adjustment of the thermal power generation unit, avoiding large changes in the molten salt temperature at the outlet of the electric heater and overheating decomposition caused by delays in molten salt flow rate control and delays in control response, and meeting the demands for practical application.

[0066] Another embodiment of the present disclosure relates to a control device for assisting frequency regulation of a thermal power generation unit with molten salt energy storage, as shown in FIG. 5 , an acquisition module 510 used to acquire the inlet molten salt real-time temperature of the electric heater; an input module 520 used to input the electric heater inlet molten salt real-time temperature, the electric heater outlet molten salt temperature and the electric heater power into the trained neural network model of molten salt flow rate to obtain a desired target molten salt flow rate; and a control module 530 used to control the molten salt flow rate to assist in frequency regulation of the thermal power generation unit based on the target molten salt flow rate.

[0067] For a specific implementation method of a control device that assists in frequency regulation of a thermal power generation unit using molten salt energy storage according to an embodiment of the present disclosure, please refer to the control method that assists in frequency regulation of a thermal power generation unit using molten salt energy storage according to an embodiment of the present disclosure, and will not be repeated here.

[0068] In a control device that uses molten salt energy storage to assist in frequency adjustment of a thermal power generation unit according to an embodiment of the present disclosure, compared to the prior art, a desired target molten salt flow rate is obtained using a trained neural network model of molten salt flow rate based on the real-time molten salt temperature at the inlet of the electric heater, the molten salt temperature at the outlet of the installed electric heater, and the power of the electric heater, and the molten salt flow rate that assists in frequency adjustment of the thermal power generation unit is controlled based on the target molten salt flow rate, thereby effectively realizing fast and accurate control of the molten salt flow rate in the process of assisting in frequency adjustment of the thermal power generation unit, avoiding large changes in the molten salt temperature at the outlet of the electric heater and overheating decomposition caused by delays in molten salt flow rate control and delays in control response, and meeting the demands for practical application.

[0069] The above-described embodiments are specific embodiments for realizing the present disclosure, and it will be understood by those skilled in the art that various changes in form and details may be made in practical application without departing from the spirit and scope of the present disclosure.

Claims

1. 1. A control method for assisting frequency regulation of a thermal power generation unit with molten salt energy storage, comprising: Obtaining a real-time temperature of the molten salt at the inlet of the electric heater; Inputting the real-time temperature of the molten salt at the inlet of the electric heater, the set temperature of the molten salt at the outlet of the electric heater, and the electric power of the electric heater into a trained neural network model of the molten salt flow rate to obtain a desired target molten salt flow rate; and controlling a molten salt flow rate that assists in frequency regulation of a thermal power generation unit based on the target molten salt flow rate.

2. The step of training the neural network model of the molten salt flow rate includes: Establishing a neural network model of the molten salt flow rate using the molten salt temperature at the outlet of the electric heater, the molten salt temperature at the inlet of the electric heater, and the electric power of the electric heater as input parameters and the molten salt flow rate as an output parameter; obtaining sample data corresponding to each of the input parameters and the output parameters under different steady-state operating conditions to construct a training data set; and training a neural network model of the molten salt flow rate using the training data set.

3. the neural network model of the molten salt flow rate includes a radial basis function network model, the radial basis function network model including an input layer, a hidden layer, and an output layer; The output data of the hidden layer can be expressed as follows: [Equation 1] where j represents the node number of the hidden layer, j=1, 2, . . . , q, q represents the number of nodes in the hidden layer, [Equation 2] is the input sample vector [Equation 3] represents the output data of the j-th hidden layer node in the case [Equation 4] represents the Gaussian function center vector of the j-th hidden layer node, [Equation 5] represents the width vector of the j-th hidden layer node, T represents the transpose, and exp represents the natural exponential function. The output data of the output layer is expressed as follows: [Equation 6] where y represents the molten salt flow rate, [Equation 7] 3. The control method of claim 2, wherein j represents the weighting coefficient of the j-th hidden layer node.

4. training a neural network model of the molten salt flow rate using the training data set; 4. The control method of claim 3, further comprising: using the training data set and a corresponding iterative formula based on corresponding initial values, iteratively calculating the Gaussian function center vector, width vector, and weighting coefficient of each hidden layer node until the root mean square error output from the radial basis function network model is smaller than a preset accuracy.

5. The initial value corresponding to the center vector of the Gaussian function is expressed as follows: [Equation 8] Here, i represents the node number of the input layer, i=1, 2, . . . , p, p represents the number of nodes in the input layer, [Equation 9] represents the Gaussian function center vector of the jth hidden layer node based on the ith input layer node, [Equation 10] represents the minimum value of the input sample corresponding to the i-th input layer node, [0011] represents the maximum value of the input sample corresponding to the i-th input layer node, The initial value corresponding to the width vector is expressed as: [0012] where: [0013] represents the width vector of the jth hidden layer node based on the ith input layer node, [0014] represents the adjustment factor, [Equation 15] 、 [0016] represents the input sample corresponding to the i-th input layer node, The initial values ​​corresponding to the weighting factors are expressed as follows: [Equation 17] where: [Equation 18] represents the minimum value of the output sample, [Equation 19] 5. The control method of claim 4, wherein σ represents the maximum value of the output samples.

6. The iterative equation corresponding to the weighting factors is expressed as follows: [Equation 20] where t represents the number of iterations, [0000] represents the weighting coefficient of the j-th hidden layer node at the t+1-th iteration, [Equation 22] represents the weighting coefficient of the j-th hidden layer node at the t-th iteration, [Equation 23] represents the weighting coefficient of the j-th hidden layer node at the t-1-th iteration, [0000] represents the learning factor, [Equation 25] represents the repetition factor, and 0< [Equation 26] <1, E represents the output bias evaluation function, [0000] , n represents the sample number, n=1, 2, 3, . . . , N, N represents the total number of samples, [0000] represents the output sample value corresponding to the nth input sample, [0000] represents the output value of the radial basis function network model corresponding to the n-th input sample, The iterative equation corresponding to the center vector of the Gaussian function is expressed as follows: [Equation 30] where: [Equation 31] represents the Gaussian function center vector of the jth hidden layer node based on the ith input layer node at the t+1th iteration, [Equation 32] represents the Gaussian function center vector of the jth hidden layer node based on the ith input layer node at the tth iteration, [Equation 33] represents the Gaussian function center vector of the jth hidden layer node based on the ith input layer node at the t-1th iteration, The iterative formula corresponding to the width vector is expressed as follows: [Equation 34] where: [Equation 35] represents the width vector of the j-th hidden layer node based on the i-th input layer node at the t+1th iteration, [Equation 36] represents the width vector of the j-th hidden layer node based on the i-th input layer node at the t-th time, [Equation 37] The control method according to claim 5, characterized in that: represents the width vector of the j-th hidden layer node based on the i-th input layer node at the t-1th time.

7. When the root mean square error output from the radial basis function network model is expressed as RMS, the RMS is: [Number 38] The control method according to claim 6, characterized in that it is expressed as follows:

8. training a neural network model of the molten salt flow rate using the training data set; 5. The control method of claim 4, further comprising: normalizing input samples included in the training data set before the step of iteratively calculating the Gaussian function center vector, width vector, and weighting coefficient of each hidden layer node using the training data set and corresponding iterative formula based on corresponding initial values ​​until the root mean square error output from the radial basis function network model is smaller than a preset accuracy.

9. The step of controlling the molten salt flow rate for supporting frequency adjustment of a thermal power generation unit based on the target molten salt flow rate includes: Using a molten salt flow meter in combination with a variable frequency molten salt pump and adjusting the frequency of the variable frequency molten salt pump to adjust the molten salt flow rate to the target molten salt flow rate; or 2. The control method according to claim 1, further comprising: using the molten salt flow meter in combination with an adjustment valve and adjusting the opening degree of the adjustment valve to adjust the molten salt flow rate to the target molten salt flow rate.

10. 1. A control device for assisting frequency regulation of a thermal power generation unit using molten salt energy storage, comprising: an acquisition module used to acquire the inlet molten salt real-time temperature of the electric heater; an input module used to input the inlet molten salt real-time temperature of the electric heater, the set outlet molten salt temperature of the electric heater, and the electric heater power into the trained neural network model of molten salt flow rate to obtain a desired target molten salt flow rate; a control module used to control a molten salt flow rate that assists in frequency regulation of a thermal power generation unit based on the target molten salt flow rate.

Images