Modular molten salt energy storage vapor heat release control method and system

Abstract

To provide a modular molten salt energy storage system that optimizes the heat exchange process between molten salt and water vapor, significantly improving the overall efficiency of the energy storage system and reducing energy loss. [Solution] The method includes the steps of establishing a BP neural network model of a modular molten salt energy storage and heat dissipation control system and training the BP neural network model; defining an output error after normalization and analyzing the BP neural network model based on the output error; processing the analyzed BP neural network model with fluctuation rate adjustment technology, establishing a corresponding real mathematical model, and completing the steam heat dissipation control of the modular molten salt energy storage.

Description

[Technical Field]

[0001] The present invention relates to the technical field of molten salt energy storage, and in particular to a vapor heat release control method and system for modular molten salt energy storage. [Background technology]

[0002] Currently, molten salt has become an ideal thermal energy storage medium due to its high specific heat capacity, high thermal stability, and good thermal conductivity at high temperatures. The emergence of molten salt energy storage technology has resolved the conflict between intermittent energy supply and continuous demand. While existing solar thermal power generation systems cannot collect sufficient solar radiation at night or on cloudy days, molten salt energy storage can store excess thermal energy during the day and release it when needed, ensuring continuous power supply. The current development trend is to combine molten salt energy storage systems with thermal power generation units to achieve thermal-electrolytic coupling auxiliary peak frequency regulation of the units.

[0003] Currently, there are two types of commonly used molten salt energy storage systems: double tank and single tank. These systems are usually bulky, have high investment costs, and are not easy to adjust, limiting their wide application. Therefore, modular molten salt energy storage technology is an advanced thermal energy storage solution that combines the high efficiency of molten salt energy storage with the flexibility of modular design, providing a new path for the integration of renewable energy and the optimization of traditional thermal power generation. The modular design further improves the scalability, economy, and adaptability of the system.

[0004] The modular molten salt energy storage system solves the above problems by dividing a large energy storage unit into multiple smaller, independently operable modules, each of which can be manufactured, tested, transported, and installed independently, significantly reducing the complexity and cost of on-site construction. Furthermore, the modular design allows the system to be flexibly expanded according to actual demand, allowing it to quickly adapt to different load requirements, both during initial construction and subsequent expansion. Summary of the Invention

[0005] In view of the problems that exist in existing methods and systems for controlling vapor heat release in modular molten salt energy storage, the present invention is proposed.

[0006] Therefore, the objective of the present invention is to provide a vapor heat release control method and system for modular molten salt energy storage, which combines BP neural network and regulation technology to solve the problem of insufficient control precision of existing molten salt energy storage systems.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present invention provides a method for controlling steam heat release of modular molten salt energy storage, including the steps of establishing a BP neural network model of a modular molten salt energy storage and heat release control system, training the BP neural network model, and obtaining node parameters of the model; normalizing the node parameters, defining an output error after normalization, and analyzing the BP neural network model based on the output error; processing the analyzed BP neural network model with a fluctuation rate adjustment technique; and establishing a corresponding real mathematical model based on the processed BP neural network model, thereby completing steam heat release control of the modular molten salt energy storage.

[0009] In a preferred embodiment of the method for controlling steam heat release of modular molten salt energy storage described in the present invention, the step of establishing the BP neural network model includes: y1 , T y2 , T y3 and T y4 as a medium temperature of the molten salt; The input includes inputting the N energy storage and heat dissipation parameters through an input layer and outputting them through a hidden layer, and the hidden layer obtains a number of hidden layer nodes corresponding to the number of nodes in the input layer, and the number of nodes in the input layer includes x1 to x7, where x1→fy, x2→Lso, x3→Tsin, and x4→T y1 , x5→T y2 , x6→T y3 , x7→T y4 , in this case, x4→T y1 represents a hidden layer, the number of nodes in the hidden layer is L=h, and the output equation of the j-th neuron in the hidden layer is as follows:

number

[0010] In a preferred embodiment of the method for controlling steam heat dissipation of modular molten salt energy storage described in the present invention, the step of outputting through the hidden layer includes introducing a node activation function to preprocess the j-th neuron of the hidden layer, so as to accelerate the learning speed of the BP neural network model, and calculating the j-th neuron of the output layer, and the specific calculation formula is as follows:

number

[0011] In a preferred embodiment of the method for controlling steam heat release of modular molten salt energy storage described in the present invention, the step of training a BP neural network model includes: setting weights from an input layer to a hidden layer and weights from an output layer to a hidden layer; When the weight from the input layer to the hidden layer is greater than the threshold of the hidden layer node of the jth neuron, it enters the saturation region of the BP neural network model, at this time, the BP neural network model converges and enters the saturation region to set the weight from the input layer to the hidden layer and the weight from the output layer to the hidden layer, and randomly takes non-zero values ​​as the weight from the input layer to the hidden layer and the weight from the output layer to the hidden layer.

[0012] In a preferred embodiment of the method for controlling steam heat release of modular molten salt energy storage described in the present invention, the step of normalizing node parameters includes unifying all node parameters using an activation function, and the specific normalization formula of the activation function is as follows:

number

number

number

[0013] In a preferred embodiment of the vapor heat release control method for modular molten salt energy storage described in the present invention, the step of analyzing the BP neural network model based on the output error is:

number

number

number

number

number

number

number

[0014] In a preferred embodiment of the method for controlling vapor heat release of modular molten salt energy storage described in the present invention, the step of processing the analyzed BP neural network model includes adjusting the learning rate using a variation rate adjustment technique.

number

number

number

number

number

number

number

[0015] In a second aspect, an embodiment of the present invention provides a vapor heat release control system for modular molten salt energy storage, including: a training module for establishing a BP neural network model of the modular molten salt energy storage and heat release control system and training the BP neural network model; an analysis module for defining a normalized output error and analyzing the BP neural network model based on the output error; and a control module for processing the analyzed BP neural network model with a fluctuation rate adjustment technology, establishing a corresponding real mathematical model, and completing the vapor heat release control of the modular molten salt energy storage.

[0016] In a third aspect, an embodiment of the present invention provides a computer system including a memory having a computer program stored therein and a processor, the computer system implementing any of the steps of the method for controlling vapor heat release for modular molten salt energy storage described above when the processor executes the computer program.

[0017] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having stored thereon a computer program that, when executed by a processor, implements any of the steps of the method for controlling vapor heat release for modular molten salt energy storage described above.

[0018] The beneficial effects of this invention are as follows: By introducing BP neural network and regulation technology, the invention effectively improves the steam heat release control accuracy and response speed of the modular molten salt energy storage system, normalizes node parameters and precisely analyzes output errors to ensure the stability and reliability of the control system, and adjusts the learning rate and introduces an activation function to enable the neural network model to converge quickly, thereby achieving precise control of the molten salt pump frequency and optimizing the heat exchange process between the molten salt and steam, significantly improving the overall efficiency of the energy storage system and reducing energy losses. [Brief explanation of the drawings]

[0019] In order to more clearly describe the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings used in the description of the embodiments, and it should be understood that the drawings in the following description are only some embodiments of the present invention, and those skilled in the art can derive other drawings based on these drawings without any creative efforts. [Figure 1] FIG. 1 is a diagram showing a specific flow chart of a vapor heat release control method and system for modular molten salt energy storage provided by an embodiment of the present invention. [Figure 2] FIG. 1 is a diagram showing the process flow of a modular molten salt energy storage system structure according to an embodiment of the modular molten salt energy storage vapor heat release control method and system provided by the present invention. [Figure 3] FIG. 1 is a diagram showing the BP neural network structure of the vapor heat release control method and system for modular molten salt energy storage provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention clearer and easier to understand, the following will describe in detail specific embodiments of the present invention in conjunction with the accompanying drawings of the specification, but it is clear that the described embodiments are only a part of the embodiments of the present invention, and are not all of them. All other embodiments that can be obtained by those skilled in the art based on the embodiments of the present invention without requiring creative efforts shall fall within the protection scope of the present invention.

[0021] Numerous specific details are set forth below to facilitate a thorough understanding of the present invention; however, the present invention may be practiced in other ways than those described herein, and similar extensions may be made by those skilled in the art without departing from the spirit of the present invention, and therefore the present invention is not limited by the specific examples disclosed below.

[0022] Also, as used herein, "one embodiment" or "embodiment" refers to a particular feature, structure, or characteristic that may be included in at least one embodiment of the present invention. The appearances of the phrase "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they mutually exclusive embodiments separate or alternative from other embodiments.

[0023] The present invention will be described in detail in conjunction with schematic diagrams, and when describing the embodiments of the present invention in detail, for ease of explanation, the cross-sectional views showing the structure of the device are partially enlarged and not according to the normal scale, and the schematic diagrams are merely illustrative and do not limit the scope of protection of the present invention, and should include three-dimensional spatial dimensions of length, width and depth during actual production.

[0024] Also, in the description of the present invention, as explained, terms such as "upper, lower, inner, outer" indicate orientations or positional relationships based on those shown in the accompanying drawings and are intended only to facilitate and simplify the description of the present invention, and do not direct or suggest that the referenced devices or elements must have a particular orientation or be constructed and operated in a particular orientation, and therefore should not be understood as a limitation of the present invention. Furthermore, the terms "first, second, or third" are for descriptive purposes only and should not be understood as directing or suggesting relative importance.

[0025] In the present invention, the terms "attach, connect, couple" should be understood in a broad sense unless otherwise specified or limited, and may mean, for example, fixedly connected, detachably connected, or integrally connected, mechanically connected, electrically connected, directly connected, indirectly connected via an intermediate medium, or internal communication between the two members. Those skilled in the art can understand the specific meaning of the above terms in the present application according to the actual situation.

[0026] Example 1 Referring to FIGS. 1 to 3, a first embodiment of the present invention provides a vapor heat dissipation control method for modular molten salt energy storage, which includes the following steps S1 to S3.

[0027] S1: Establish a BP neural network model of the modular molten salt energy storage and heat dissipation control system, train the BP neural network model, and obtain the node parameters of the model.

[0028] Here, the step of establishing the BP neural network model is to calculate T based on the input N energy storage and heat dissipation parameters. y1 , T y2 , T y3 and T y4 as a medium temperature of the molten salt; The input includes inputting N energy storage and heat dissipation parameters through the input layer and outputting them through the hidden layer, and the number of nodes in the hidden layer corresponds to the number of nodes in the input layer. The number of nodes in the input layer includes x1 to x7, where x1 → fy, x2 → Lso, x3 → Tsin, and x4 → T y1 , x5→T y2 , x6→T y3 , x7→T y4 , in this case, x4→T y1 represents the hidden layer, the number of nodes in the hidden layer is L=h, and the output formula of the j-th neuron in the hidden layer is as follows:

number

[0029] Furthermore, this invention establishes a BP neural network model for a modular molten salt energy storage and heat dissipation control system to accurately control the steam heat dissipation process. The model is constructed based on N energy storage and heat dissipation parameters, mainly including key indicators such as molten salt temperature, pressure, and flow rate. For example, assume that the input layer contains 12 nodes corresponding to 12 key energy storage parameters. These parameters are input to the input layer and then processed by hidden layer calculations. The number of hidden layer nodes is determined by matching the number of nodes in the input layer. Assume that the hidden layer contains 20 nodes. Each hidden layer node is processed by an activation function based on the weight and threshold of the input layer node and then outputs a result. These output results are then used to extract the molten salt medium temperature to reflect the current thermal state of the system. This neural network model can accurately simulate the dynamic behavior of the molten salt energy storage system and lay the foundation for subsequent accurate control. In this way, by processing large amounts of input data and generating accurate output node parameters, the system is provided with efficient and real-time control capabilities, effectively improving the overall efficiency and response speed of energy storage and heat dissipation.

[0030] S1.1: The step of outputting through the hidden layer includes introducing a node activation function to preprocess the jth neuron of the hidden layer to accelerate the learning speed of the BP neural network model, and calculating the jth neuron of the output layer. The specific calculation formula is as follows:

number

number

[0031] Furthermore, to improve the learning efficiency of the BP neural network model, the present invention preprocesses each neuron in the hidden layer using a node activation function before outputting it through the hidden layer. Specifically, assuming the hidden layer contains 20 neurons, an activation function is applied to each neuron, accelerating the convergence of the neural network. For example, when processing input data containing 1,000 data points, the activation function can be used to quickly filter parameters closely related to steam heat dissipation. These processed data are then transmitted to the output layer, which may contain five neurons. The connection weights between each neuron and the 20 neurons in the hidden layer are used to calculate the final output energy storage and heat dissipation parameters. Under different operating conditions, such as high load and low load, the neurons in the output layer generate different control signals to accurately adjust the system's steam heat dissipation. Through this neural network pre-model, the system can analyze and obtain various input and output energy storage and heat dissipation parameters in real time, providing reliable data support for subsequent control decisions, significantly improving the accuracy and efficiency of control.

[0032] S1.2: The step of training the BP neural network model includes setting weights from the input layer to the hidden layer and weights from the output layer to the hidden layer; When the weight from the input layer to the hidden layer is greater than the threshold of the hidden layer node of the jth neuron, it enters the saturation region of the BP neural network model, at this time, the BP neural network model converges and enters the saturation region for setting the weight from the input layer to the hidden layer and the weight from the output layer to the hidden layer, and randomly takes non-zero values ​​as the weight from the input layer to the hidden layer and the weight from the output layer to the hidden layer.

[0033] To ensure the effectiveness of the BP neural network model, the training process of the present invention requires accurate weighting between the input layer and the hidden layer, and between the hidden layer and the output layer. Assume that the input layer has 12 nodes, the hidden layer has 20 nodes, and the output layer has 5 nodes. The connection weights between these nodes are initially randomly assigned non-zero values. For example, the weights from the input layer to the hidden layer can be set to random values ​​between 0.01 and 0.1. As training progresses, these weights gradually increase until they exceed the node threshold of a certain neuron in the hidden layer, causing the network to enter the so-called saturation region. In this region, the BP neural network model begins to converge, i.e., the error gradually decreases and the network output effect is optimized. The system can then fix these weights within the saturation region and further adjust them to ensure that the model can effectively learn and adapt to new data inputs even in a saturated state. This avoids overtraining or undertraining the model, ensuring system stability and accuracy under different operating conditions and effectively improving the accuracy of steam heat release control. The neural network weight settings are shown in Table 1 below.

[0034] Table 1 Neural network weight setting data table [Table 1] The table shows the number of nodes in each layer of the BP neural network model and their weight settings. The input layer contains 12 nodes, the hidden layer contains 20 nodes, and the output layer contains 5 nodes. Furthermore, the table lists the range of weights from the input layer to the hidden layer as 0.01 to 0.1, but does not specify the specific range of weights from the hidden layer to the output layer. Overall, this table summarizes the initial weight settings of the model and the configuration between different layers.

[0035] S2: Normalize the node parameters, define the output error after normalization, and analyze the BP neural network model based on the output error.

[0036] Here, the step of normalizing the node parameters includes unifying all the node parameters using an activation function, and the specific normalization formula of the activation function is as follows:

number

number

number

number

number

[0037] Furthermore, in the present invention, normalization of node parameters is an important step in improving the accuracy and stability of the BP neural network model. Specifically, when dealing with the steam heat dissipation control of a molten salt energy storage system, multiple input parameters such as temperature, pressure, and flow rate must be processed uniformly. These parameters may have significantly different physical scales. In order to prevent these differences from affecting the learning effect of the neural network, these parameters are normalized using an activation function. If the temperature range is 200°C to 500°C and the pressure range is 2MPa to 10MPa, these parameters are converted to a uniform scale by normalization so that they fluctuate within the same range (e.g., 0 to 1).

[0038] After completing normalization, the system further defines output error as an indicator of the neural network model's performance and determines the model's accuracy by calculating the difference between the normalized output value and the actual value. For example, in a single steam heat dissipation operation, if the normalized output temperature is 0.8 and the corresponding normalized value is 0.75, the output error will reflect this difference. Based on this error data, the system analyzes and adjusts the BP neural network model to ensure accurate control under various operating conditions. This not only improves the model's robustness, but also enhances the system's ability to adapt to complex dynamic environments, helping to achieve more efficient energy storage and heat dissipation management.

[0039] S2.1: The step of analyzing the BP neural network model based on the output error is to determine the minimum error

number

number

number

number

number

number

number

[0040] In the present invention, analyzing the output error is an important step in optimizing the BP neural network model. First, a minimum error value is set to allow the system to determine whether the model training has reached an ideal state. If the output error is smaller than this minimum error value, for example, set to 0.001, it means that the model is accurate enough and the model training can be terminated. If the output error is larger than this minimum error value, it means that the model is out of alignment and needs to be further optimized. At this time, the system will use gradient descent to adjust the neural network weights.

[0041] In actual operation, assuming the hidden layer contains 20 neurons and the output layer contains 5 neurons, the system fine-tunes the weights of each neuron to reduce error. For example, if a calculation finds that the weights of the hidden layer neurons are too high, resulting in a large output error, it uses gradient descent to fine-tune the weights and reduce the overall error. Setting the learning rate to 0.01 means that each adjustment is relatively small, allowing the model to gradually converge to an optimal state. After repeated iterations, the system gradually reduces the error within the set range, finally completing the establishment of the BP neural network model. This fine-grained error analysis and weight adjustment method ensures the accuracy and reliability of the model and enables efficient control of complex energy storage and heat dissipation processes.

[0042] S3: The analyzed BP neural network model is processed using the fluctuation rate adjustment technology, and based on the processed BP neural network model, a corresponding real mathematical model is established to complete the steam heat release control of the modular molten salt energy storage.

[0043] Here, the step of processing the analyzed BP neural network model is to adjust the learning rate using a variance adjustment technique.

number

number

number

number

number

number

number

[0044] Preferably, in the present invention, a variable rate adjustment technique is used to optimize the learning rate of the BP neural network model. In the initial stage, the system uses a relatively large learning rate, such as 0.1, to ensure rapid convergence of the model. As the iterations progress, the learning rate is gradually reduced, sometimes to 0.01 or less for fine-tuning. After multiple iterations of optimization, the model calculates a molten salt pump frequency suitable for the current operating conditions, such as 1500 RPM, and sends this frequency to the molten salt pump frequency converter. The frequency converter then controls the molten salt flow rate based on the adjusted frequency, thereby accurately adjusting the heat exchange rate between the molten salt and steam and achieving effective control of the steam heat dissipation process. In this way, the system can maintain high response speed and control accuracy during actual operation.

[0045] In a preferred embodiment, the modular molten salt energy storage steam heat release control system includes: a training module for establishing a BP neural network model of the modular molten salt energy storage and heat release control system and training the BP neural network model; an analysis module for defining a normalized output error and analyzing the BP neural network model according to the output error; and a control module for processing the analyzed BP neural network model with a fluctuation rate adjustment technology, establishing a corresponding real mathematical model, and completing the modular molten salt energy storage steam heat release control.

[0046] Each of the above unit modules may be incorporated in a processor in a computer facility in the form of hardware, may be independent, or may be stored in a memory in a computer facility in the form of software so that the processor can easily perform operations corresponding to each of the above modules.

[0047] The computer device may be a terminal, and includes a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor of the computer device is used to provide calculation and control functions. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an operating environment for the operating system and computer programs stored in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal via a wired or wireless method, which may be achieved by Wi-Fi, an operator network, NFC (near field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display, and the input device of the computer device may be a touch layer covered by the display, buttons, a trackball, or a touchpad on the housing of the computer device, or an external keyboard, touchpad, or mouse.

[0048] In summary, by introducing the BP neural network and regulation technology, the present invention effectively improves the steam heat dissipation control accuracy and response speed of the modular molten salt energy storage system. The normalization of node parameters and precise output error analysis ensure the stability and reliability of the control system. By adjusting the learning rate and introducing an activation function, the neural network model can converge quickly, thereby achieving precise control of the molten salt pump frequency and optimizing the heat exchange process between the molten salt and steam. This significantly improves the overall efficiency of the energy storage system and reduces energy losses.

[0049] Example 2 Referring to Figures 1 to 3, a second embodiment of the present invention provides a vapor heat release control method for modular molten salt energy storage, and scientific demonstrations have been carried out through simulation experiments to verify the beneficial effects of the present invention.

[0050] The experimental process began with the establishment of a BP neural network model for the modular molten salt energy storage and heat dissipation control system. Initially, the input layer had 15 nodes, the hidden layer had 25 nodes, and the output layer had 6 nodes. To evaluate the model's performance, the weights from the input layer to the hidden layer were initially set between 0.02 and 0.08, and the weights from the hidden layer to the output layer ranged from 0.05 to 0.15. The training phase used 1,000 sets of experimental data, each containing parameters such as molten salt temperature and flow rate. The model was trained using a learning rate adjustment technique, which initially set the learning rate to 0.05 and gradually decreased it to 0.005. 5,000 training iterations were performed, gradually adjusting the weights to optimize model performance.

[0051] During the testing phase, new molten salt temperature data such as 300°C and 320°C and flow rate data such as 0.8L / s and 1.2L / s were input, and the molten salt pump frequencies output by the model were 1800RPM and 1900RPM, respectively. The frequency values ​​were transferred to the molten salt pump frequency converter to adjust the heat exchange rate between the molten salt and steam. Finally, the experimental results showed that the system could accurately adjust the steam heat dissipation process, improve control accuracy, and keep the error range within 5%, which verified the effectiveness of the model in practical use. The experimental data are summarized in Table 2 below.

[0052] Table 2 Summary of experimental data [Table 2] This table summarizes the main data obtained from the experimental process, such as the number of nodes of the BP neural network model, the weight range, and the specific parameters of training and testing. This table shows the number of nodes and their weight ranges in the input layer, hidden layer and output layer, the number of data sets in the training process, the learning rate and number of iterations, the temperature and flow rate data used in the testing stage, as well as the error range of the pump frequency and control accuracy output from the model. A comparison between the technical solution of the present invention and the prior art is shown in Table 3 below.

[0053] Table 3 Comparison table between the technical solution of the present invention and the prior art [Table 3] The table compares the main differences and advantages of the technical solution of the present invention with the prior art. The present invention adopts a BP neural network model, which can handle more complex nonlinear problems compared with traditional linear models, providing higher accuracy, more accurate weight initialization, and adaptive adjustment of the learning rate, making model training more stable and efficient. In terms of data processing, normalization improves the consistency of data and the convergence of the model, significantly improving control accuracy and reducing the error range to 5%. Utilizing more data and training iterations improves the generalization ability of the model. Finally, the fast adjustment ability of output feedback improves the timeliness of the control response. Overall, the present invention is superior to the prior art in every respect and provides a more accurate and stable control method.

[0054] It should be noted that the above-described embodiments are only for illustrating the technical solutions of the present invention, and are not intended to limit the same. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art may make modifications or equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and these modifications or substitutions should be included in the scope of the claims of the present invention.

Claims

1. 1. A method for controlling vapor heat release in modular molten salt energy storage, comprising: Establishing a BP neural network model of a modular molten salt energy storage and heat release control system, training the BP neural network model, and obtaining node parameters of the model; normalizing the node parameters, defining an output error after normalization, and analyzing the BP neural network model based on the output error; A method for controlling vapor heat release of modular molten salt energy storage, comprising the steps of: processing the analyzed BP neural network model with a fluctuation rate adjustment technology; and establishing a corresponding real mathematical model based on the processed BP neural network model, thereby completing the vapor heat release control of modular molten salt energy storage.

2. The step of establishing the BP neural network model includes: ｙ１ , T ｙ２ , T ｙ３ and T ｙ４ as a medium temperature of the molten salt; The input includes inputting the N energy storage and heat dissipation parameters through an input layer and outputting them through a hidden layer, and the hidden layer obtains a number of nodes in the hidden layer corresponding to a number of nodes in the input layer, and the number of nodes in the input layer is x １ ~x ７ where x １ → fy, x ２ →Lso, x ３ →Tsin, x ４ →T ｙ１ , x ５ →T ｙ２ , x ６ →T ｙ３ , x ７ →T ｙ４ , in this case, x ４ →T ｙ１ represents a hidden layer, the number of nodes in the hidden layer is L=h, and the output equation of the j-th neuron in the hidden layer is as follows: [Equation 1] Here, i = 1, 2, 3... n, j = 1, 2, 3... L, wh ｉｊ represents the weight from the input layer to the hidden layer, and a ｊ represents the threshold of the hidden layer node, and h ｊ The method for controlling vapor heat release of modular molten salt energy storage according to claim 1 , characterized in that: represents the j-th neuron of the hidden layer.

3. The step of outputting through the hidden layer includes introducing a node activation function to preprocess the j-th neuron of the hidden layer to accelerate the learning speed of the BP neural network model, and calculating the j-th neuron of the output layer, the specific calculation formula is as follows: [Equation 2] Here, y ｊ represents the j-th neuron in the output layer, and wy ｊ represents the weight from the output layer to the hidden layer, and b １ represents the threshold of the output layer node, and h ｊ represents the j-th neuron in the hidden layer, The method for controlling steam heat dissipation of modular molten salt energy storage as claimed in claim 2, characterized in that for the jth neuron of the hidden layer and the jth neuron of the output layer, a preliminary model of neural network is established, and all node parameters of input and output energy storage and heat dissipation are obtained based on the steam heat dissipation under different situations.

4. The step of training the BP neural network model includes setting weights from the input layer to the hidden layer and weights from the output layer to the hidden layer; 4. The method for controlling steam heat dissipation of modular molten salt energy storage according to claim 3, characterized in that when the weight from the input layer to the hidden layer is greater than the threshold of the hidden layer node of the jth neuron, the BP neural network model enters the saturation region, at this time, the BP neural network model converges and enters the saturation region for setting the weight from the input layer to the hidden layer and the weight from the output layer to the hidden layer, and the weight from the input layer to the hidden layer and the weight from the output layer to the hidden layer randomly take non-zero values.

5. The step of normalizing the node parameters includes unifying all the node parameters using an activation function, and the specific normalization formula of the activation function is as follows: [Equation 3] where x represents the normalized value, and x ｒ represents the physical quantity to be operated, [Equation 4] represents the maximum value of the activation function, [Equation 5] represents the minimum value of the activation function, The specific formula for the above step of defining the output error after normalization is as follows: [Equation 6] where: [Equation 7] represents the output value of the BP neural network model, and y １ The method for controlling steam heat dissipation of modular molten salt energy storage as claimed in claim 4, wherein: represents the actual value, and E represents the output error after normalization.

6. The step of analyzing the BP neural network model based on the output error comprises: [Equation 8] including setting [Equation 9] If , the establishment of the BP neural network model is completed, [Equation 10] In this case, the weight of the j-th neuron in the hidden layer and the j-th neuron in the output layer is adjusted by gradient descent. The specific adjustment formula is as follows: [0011] where: [0012] represents the learning rate, [0013] represents an operator, [0014] The method for controlling vapor heat release of modular molten salt energy storage according to claim 5, wherein represents the first derivative of the actual value.

7. The step of processing the analyzed BP neural network model includes a learning rate [Equation 15] Iteratively adjust the learning rate [0016] and outputting a BP neural network model of the molten salt pump frequency fy based on adjusting the iteration time of fy. The specific calculation formula of the regulation rate adjustment technique is as follows: [Equation 17] where: [Equation 18] represents the maximum learning rate, [Equation 19] represents the minimum learning rate, [Equation 20] represents the maximum number of iterations, t represents the current number of iterations, [Equation 21] represents the rate of change, The method for controlling steam heat release of modular molten salt energy storage according to claim 6, characterized in that the steam heat release control includes: substituting a node of an input layer into a BP neural network model that outputs a molten salt pump frequency fy, sending the output molten salt pump frequency fy to a molten salt pump frequency converter, and adjusting the heat exchange rate between the molten salt and steam according to changes in molten salt flow conditions, thereby completing the steam heat release control.

8. A modular molten salt energy storage vapor heat release control system, which is based on the modular molten salt energy storage vapor heat release control method of claim 1, a training module for establishing a BP neural network model of the modular molten salt energy storage and heat release control system and training the BP neural network model; an analysis module for defining an output error after normalization and analyzing the BP neural network model based on the output error; A control module that processes the analyzed BP neural network model with a fluctuation rate adjustment technology, establishes a corresponding real mathematical model, and completes the vapor heat release control of the modular molten salt energy storage.

9. A computer facility comprising a memory having a computer program stored therein and a processor, the computer facility implementing the steps of the method for controlling vapor heat release for modular molten salt energy storage according to claim 1 when the processor executes the computer program.

10. 10. A computer-readable storage medium having a computer program stored thereon, the computer program implementing the steps of the modular molten salt energy storage vapor heat release control method of claim 1 when executed by a processor.

Images