Operation control method and device of fused salt heat storage system and readable storage medium
By training a load prediction model and generating operation control rules in a molten salt thermal storage system, the problems of low efficiency and slow response of the system under load fluctuations are solved, achieving accurate load prediction and effective operation control.
Patent Information
- Application Number
- CN202511594700.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-03
- Publication Date
- 2026-03-27
AI Technical Summary
Existing molten salt thermal energy storage systems struggle to accurately capture dynamic load fluctuations under the coupling of multiple factors during operation and control, resulting in low thermal energy storage and release efficiency and slow dynamic response.
By acquiring historical operating data of the molten salt thermal storage system, the load prediction model is iteratively trained to generate the target load prediction model. Then, operating control rules are generated for each discrete time interval within the load prediction time domain. Considering multi-dimensional constraints, a multi-stage operating control decision model is established.
It enables accurate load demand identification and effective operation control of molten salt thermal storage systems, improving the system's heat storage and release efficiency and dynamic response capability.
Smart Images

Figure CN121739796A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of molten salt energy storage technology, and in particular to an operation control method, an operation control device, and a computer-readable storage medium for a molten salt thermal storage system. Background Technology
[0002] With the continuous increase in installed capacity of renewable energy power generation, the power grid has placed higher demands on the peak-shaving and frequency regulation capabilities of thermal power units. Traditional thermal power units, limited by their thermoelectric coupling characteristics, struggle to achieve deep peak-shaving operation. Therefore, to improve the flexibility of thermal power units, molten salt thermal energy storage technology is commonly used. Molten salt thermal energy storage technology refers to the ability to store excess heat energy from thermal power units in high-temperature molten salt during low-load periods and release the stored heat during high-load periods. This allows for deep peak-shaving while ensuring heating demand, and also reduces energy losses caused by frequent unit start-ups, shutdowns, or load changes.
[0003] Currently, in the operation and control of molten salt thermal storage systems in coal-fired power units, corresponding operation and control rules and scheduling strategies are usually set in advance, or intelligent control is carried out based on neural network models. However, due to the complexity of load fluctuation factors affecting the operation of molten salt thermal storage systems, even with the use of neural network models, it is difficult to accurately capture the dynamic load fluctuation patterns under the coupling of multiple factors. This can lead to low heat storage and release efficiency and slow dynamic response in molten salt thermal storage systems. Therefore, the current control effect of operating molten salt thermal storage systems is poor. Summary of the Invention
[0004] Based on this, it is necessary to provide an operation control method, an operation control device, and a computer-readable storage medium for molten salt thermal storage systems that improves the control effect of operation control of molten salt thermal storage systems in response to the above-mentioned technical problems.
[0005] In a first aspect, this application provides an operation control method for a molten salt thermal storage system, including:
[0006] Obtain historical operational data of the molten salt thermal storage system, wherein the historical operational data includes historical operational training data and historical operational test data;
[0007] Based on the historical operational training data, the initial load prediction model of the molten salt thermal storage system is iteratively trained to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system.
[0008] After the target load prediction model is tested and found to meet the standards based on the historical operation data, the real-time operation data is input into the target load prediction model, and the operation load prediction value for each discrete prediction time interval within the load prediction time domain is predicted by the target load prediction model.
[0009] Based on all predicted operating load values, generate corresponding operating control rules for each discrete prediction time interval;
[0010] Based on all operational control rules, the molten salt thermal storage system is operated and controlled within the load prediction time domain.
[0011] In one embodiment, generating corresponding operation control rules for each discrete prediction time interval based on all predicted operating load values includes:
[0012] Based on all predicted operating load values, different dimensions of operating constraints are constructed for each discrete prediction time interval;
[0013] Based on all operational constraints, a multi-stage operation control decision model for the molten salt thermal storage system is generated within the time domain of the load prediction.
[0014] The multi-stage operation control decision model outputs the operation control rules corresponding to each discrete prediction time interval.
[0015] In one embodiment, the operational constraints include the molten salt capacity constraint of the molten salt thermal storage system, the heat storage and release rate constraint of the molten salt thermal storage system, the thermal load constraint of the molten salt thermal storage system, and the electrical load constraint of the molten salt thermal storage system.
[0016] In one embodiment, the molten salt capacity constraint of the molten salt thermal storage system includes:
[0017] Minimum energy storage value of the first molten salt tank Real-time energy storage value The maximum energy storage value of the first molten salt tank and the minimum energy storage value of the second molten salt tank. The allowable energy dispatch value of the molten salt thermal storage system The maximum energy storage value of the second molten salt tank.
[0018] In one embodiment, the heat storage and release rate constraint of the molten salt thermal storage system includes:
[0019] 0 The real-time stored heat value of the molten salt thermal storage system The maximum heat storage value of the molten salt thermal storage system, and 0 The heat value of the steam generated by the molten salt thermal storage system The upper limit of real-time heat release of the molten salt thermal storage system.
[0020] In one embodiment, the heat load constraints of the molten salt thermal storage system include:
[0021] The steam heat fusion value is obtained by combining the heat output steam value of the molten salt thermal storage system and the heat output steam generated by the molten salt thermal storage system. The heat load requirement value of the molten salt thermal storage system.
[0022] In one embodiment, the electrical load constraints of the molten salt thermal storage system include:
[0023] Real-time steam extraction heat value of the molten salt thermal storage system The electrical load demand value of the molten salt thermal storage system.
[0024] Secondly, this application also provides an operation control device for a molten salt thermal storage system, comprising:
[0025] The acquisition module is used to acquire the historical basic operating data of the molten salt thermal storage system, wherein the historical basic operating data includes historical basic operating training data and historical basic operating test data;
[0026] The training module is used to iteratively train the initial load prediction model of the molten salt thermal storage system based on the historical operational training data to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system.
[0027] The prediction module is used to input real-time operational data into the target load prediction model after testing the target load prediction model based on the historical operational data, and to predict the operational load prediction value for each discrete prediction time interval within the load prediction time domain through the target load prediction model.
[0028] The generation module is used to generate corresponding operation control rules for each discrete prediction time interval based on all predicted operating load values.
[0029] The control module is used to control the operation of the molten salt thermal storage system within the load prediction time domain according to all operation control rules.
[0030] In one embodiment, the generation module is further configured to:
[0031] Based on all predicted operating load values, different dimensions of operating constraints are constructed for each discrete prediction time interval;
[0032] Based on all operational constraints, a multi-stage operation control decision model for the molten salt thermal storage system is generated within the time domain of the load prediction.
[0033] The multi-stage operation control decision model outputs the operation control rules corresponding to each discrete prediction time interval.
[0034] In one embodiment, the operating constraints include the molten salt capacity constraint of the molten salt thermal storage system, the heat storage and release rate constraint of the molten salt thermal storage system, the heat load constraint of the molten salt thermal storage system, and the electrical load constraint of the molten salt thermal storage system.
[0035] In one embodiment, the molten salt capacity constraint of the molten salt thermal storage system includes:
[0036] Minimum energy storage value of the first molten salt tank Real-time energy storage value The maximum energy storage value of the first molten salt tank and the minimum energy storage value of the second molten salt tank. The allowable energy dispatch value of the molten salt thermal storage system The maximum energy storage value of the second molten salt tank.
[0037] In one embodiment, the heat storage and release rate constraint of the molten salt thermal storage system includes:
[0038] 0 The real-time stored heat value of the molten salt thermal storage system The maximum heat storage value of the molten salt thermal storage system, and 0 The heat value of the steam generated by the molten salt thermal storage system The upper limit of real-time heat release of the molten salt thermal storage system.
[0039] In one embodiment, the heat load constraints of the molten salt thermal storage system include:
[0040] The steam heat fusion value is obtained by combining the heat output steam value of the molten salt thermal storage system and the heat output steam generated by the molten salt thermal storage system. The heat load requirement value of the molten salt thermal storage system.
[0041] In one embodiment, the electrical load constraints of the molten salt thermal storage system include:
[0042] Real-time steam extraction heat value of the molten salt thermal storage system The electrical load demand value of the molten salt thermal storage system.
[0043] In one embodiment, the generation module is further configured to:
[0044] Based on all the operational constraints, a multi-stage optimization objective function is constructed; based on the energy balance relationship of the molten salt thermal storage system, state transition equations for adjacent stages within the load prediction time domain are established; and based on the objective function, the state transition equations, and all the operational constraints, an initial multi-stage operation control decision model is generated; the initial multi-stage operation control decision model is iteratively optimized to obtain a multi-stage operation control decision model.
[0045] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:
[0046] Historical operational data of the molten salt thermal energy storage system is acquired, including historical operational training data and historical operational test data. Based on the historical operational training data, an initial load prediction model for the molten salt thermal energy storage system is iteratively trained to obtain a target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal energy storage system. After the target load prediction model is tested and found to meet the standards based on the historical operational test data, real-time operational data is input into the target load prediction model. The target load prediction model then predicts the operational load prediction value for each discrete prediction time interval within the load prediction time domain length. Based on all operational load prediction values, corresponding operational control rules are generated for each discrete prediction time interval. Based on all operational control rules, operational control of the molten salt thermal energy storage system is performed within the load prediction time domain length.
[0047] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:
[0048] Historical operational data of the molten salt thermal energy storage system is acquired, including historical operational training data and historical operational test data. Based on the historical operational training data, an initial load prediction model for the molten salt thermal energy storage system is iteratively trained to obtain a target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal energy storage system. After the target load prediction model is tested and found to meet the standards based on the historical operational test data, real-time operational data is input into the target load prediction model. The target load prediction model then predicts the operational load prediction value for each discrete prediction time interval within the load prediction time domain length. Based on all operational load prediction values, corresponding operational control rules are generated for each discrete prediction time interval. Based on all operational control rules, operational control of the molten salt thermal energy storage system is performed within the load prediction time domain length.
[0049] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:
[0050] Historical operational data of the molten salt thermal energy storage system is acquired, including historical operational training data and historical operational test data. Based on the historical operational training data, an initial load prediction model for the molten salt thermal energy storage system is iteratively trained to obtain a target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal energy storage system. After the target load prediction model is tested and found to meet the standards based on the historical operational test data, real-time operational data is input into the target load prediction model. The target load prediction model then predicts the operational load prediction value for each discrete prediction time interval within the load prediction time domain length. Based on all operational load prediction values, corresponding operational control rules are generated for each discrete prediction time interval. Based on all operational control rules, operational control of the molten salt thermal energy storage system is performed within the load prediction time domain length.
[0051] The aforementioned operation control method, operation control device, and computer-readable storage medium for the molten salt thermal storage system first acquire historical operational data of the system. This historical operational data includes historical operational training data and historical operational testing data, thus achieving the goal of dividing the historical operational data of the molten salt thermal storage system into historical operational training data and historical operational testing data. Then, using the historical operational training data, the initial load prediction model of the molten salt thermal storage system is iteratively trained to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time series length of the molten salt thermal storage system. After testing the target load model based on the historical operational testing data until it meets the standards, real-time operational data is input into the target load prediction model. This allows the target load prediction model to predict the operational load forecast value for each discrete prediction time interval within the load prediction time domain. This enables accurate identification of the future load forecast values of the molten salt thermal storage system using the compliant target load prediction model. The goal is to determine the trend and range of load demand changes within discrete time intervals. Then, by running composite predicted values, corresponding operation control rules are generated for each discrete prediction time interval. Finally, based on all operation control rules, the molten salt thermal storage system is operated and controlled within the load prediction time domain. Since the target load prediction model is trained and validated based on historical operational training data, the accuracy of load demand identification is ensured. Simultaneously, by running load predicted values, operation control rules perfectly adapted to load changes in each discrete prediction time interval can be generated, thus enabling effective control of the molten salt thermal storage system. Therefore, this overcomes the technical shortcomings of the complex load fluctuation factors affecting the molten salt thermal storage system during operation, which makes it difficult to accurately capture the dynamic load fluctuation patterns under multi-factor coupling, even with neural network models. This leads to low heat storage and release efficiency and slow dynamic response in the molten salt thermal storage system. Therefore, the control effect of operating the molten salt thermal storage system is improved. Attached Figure Description
[0052] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0053] Figure 1 This is a flowchart illustrating the operation control method of a molten salt thermal energy storage system in one embodiment;
[0054] Figure 2 This is a flowchart illustrating the operation control method of a molten salt thermal energy storage system in one embodiment;
[0055] Figure 3 This is a schematic flowchart of the operation control method for a molten salt thermal storage system in one embodiment.
[0056] Figure 4 This is a structural block diagram of the operation control device of a molten salt thermal energy storage system in one embodiment;
[0057] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation
[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0059] First, it should be understood that molten salt thermal energy storage systems can be matched with the parameters of thermal power unit thermal systems. By storing excess heat energy in high-temperature molten salt, the stored heat can be released during peak electricity demand for secondary power generation or to provide heating services. This achieves thermal-electric decoupling to a certain extent, improves the flexibility of heating units, and releases the deep peak-shaving potential of heating units. Therefore, the use of molten salt thermal energy storage systems is becoming increasingly widespread.
[0060] Currently, the operation control methods for molten salt thermal energy storage systems in coal-fired power units typically fall into two categories: 1) Rule-based fixed control methods, which rely on pre-set rules and scheduling strategies. For example, the timing of heat storage or release in the molten salt thermal energy storage system is determined based on key factors such as grid load, electricity price, coal-fired power unit operating conditions, and ambient temperature. Its advantages are low implementation cost and simple logic, making it suitable for situations where the operating conditions are relatively stable or have a low frequency of change; 2) Intelligent control methods based on artificial intelligence and machine learning, which rely on historical operating data of the molten salt thermal energy storage system and use artificial intelligence algorithms such as neural networks and support vector machines to establish a high-dimensional, multi-factor coupled system model. The established model is used to guide the operation of the molten salt thermal energy storage system. This approach not only improves the flexibility of scheduling the system's operation but also has strong adaptability in the face of sudden or complex operating conditions.
[0061] However, in complex and variable load scenarios with strong randomness, the former (heat storage and release) often struggles to timely and accurately schedule the heat storage and release processes using fixed rules, potentially leading to decreased system efficiency and poor economic returns. The latter (heat storage and release) model performance largely depends on the scale and diversity of the training data, requiring comprehensive data acquisition, feature selection, and online learning strategies during actual deployment to ensure the accuracy and stability of the control strategy. Therefore, in summary, the complex factors influencing load fluctuations in molten salt thermal storage systems make existing operation control methods inadequate for handling these fluctuations, resulting in low system heat storage and release efficiency and slow dynamic response. Thus, there is an urgent need for an operation control method for molten salt thermal storage systems that can improve the control effectiveness of these systems.
[0062] In one embodiment, such as Figure 1 As shown, a method for operating and controlling a molten salt thermal storage system is provided. This embodiment uses the method applied to a control terminal of the molten salt thermal storage system as an example. The control terminal includes, but is not limited to, a local control cabinet, an industrial-grade computer, an embedded controller, and a host computer with remote monitoring and control functions. The control terminal includes an acquisition module, a training module, a prediction module, a generation module, and a control module. The acquisition module is used to acquire historical operating data of the molten salt thermal storage system, which includes historical operating training data and historical operating test data. The training module is used to iteratively train the initial load prediction model of the molten salt thermal storage system based on the historical operating training data to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system. The prediction module is used to, after testing the target load prediction model based on the historical operating test data, convert the target load prediction model into a real-time operating data. Data is input into the target load prediction model, which predicts the operating load for each discrete prediction time interval within the load prediction time domain. The generation module generates corresponding operating control rules for each discrete prediction time interval based on all operating load prediction values. The control module controls the operation of the molten salt thermal storage system within the load prediction time domain based on all operating control rules. Through information interaction between the acquisition module, training module, prediction module, generation module, and control module, the target load prediction model can accurately identify the load demand change trend and numerical range of the molten salt thermal storage system in each discrete time interval in the future. Ultimately, it can generate operating control rules that are fully adapted to the load changes in each discrete prediction time interval, thus effectively controlling the operation of the molten salt thermal storage system and improving the control effect of the system's operation. In this embodiment, the method includes the following steps 202 to 210:
[0063] Step 202: Obtain historical operational data of the molten salt thermal storage system, including historical operational training data and historical operational test data.
[0064] It should be noted that historical operational baseline data refers to complete data from the past operation of the molten salt thermal storage system, specifically including historical load data, historical meteorological data, and historical time variable data. This can be understood as the system entering its initial operational state after startup. During the data acquisition phase, historical heat load data, historical electricity load data, meteorological data, time variables, and electricity spot market price data can be collected as initial data sources to establish a complete data acquisition system and implement unified data management to ensure data accuracy and real-time performance. After the initial data source is collected, the initial data can be cleaned and outlier processed. Specifically, methods such as Exponentially Weighted Moving Average (EWMA) and Kalman filtering can be used for data imputation and noise reduction to improve data quality and provide high-quality data for subsequent model development. Simultaneously, feature engineering techniques such as one-hot encoding and partial autocorrelation functions can be used to optimize the model's input variables. After obtaining the historical operational baseline data, a combination of hierarchical clustering and stratified sampling can be used to divide the data into historical operational baseline training data and historical operational baseline test data, ensuring consistency in the distribution of various samples in both datasets.
[0065] As an example, step 202 includes: after the molten salt thermal storage system is started, collecting historical operational data of the molten salt thermal storage system, and dividing the historical operational data into historical operational training data and historical operational test data.
[0066] Step 204: Based on historical operational training data, iteratively train the initial load prediction model of the molten salt thermal storage system to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system.
[0067] It should be noted that, based on the load prediction time domain length of the molten salt thermal storage system, an initial load prediction model can be determined. Then, by repeatedly adjusting the model parameters of the initial load prediction model using historical training parameters, a target load prediction model with satisfactory prediction accuracy can be obtained. For example, in one feasible approach, the load prediction time domain length H of the molten salt thermal storage system is directly detected, and the category of the initial load prediction model is determined by H. If the load prediction time domain length H is greater than a preset load prediction time domain length threshold, the direct prediction model is used as the initial load prediction model; if the load prediction time domain length H is less than or equal to the preset load prediction time domain length threshold, the recursive prediction model is used as the initial load prediction model. It is understandable that the introduction of the model selection mechanism ensures the accuracy and stability of the prediction. Afterwards, based on historical operational training data and the selected prediction model, the initial load prediction model is established.
[0068] As an example, step 204 includes: if the load prediction time domain length of the molten salt thermal storage system is detected to be greater than the preset load time domain length threshold, a first preset prediction model is used as the initial load prediction model; if the load prediction time domain length of the molten salt thermal storage system is detected to be less than or equal to the preset load time domain length threshold, a second preset prediction model is used as the initial load prediction model; after determining the initial load prediction model, the initial load prediction model of the molten salt thermal storage system is iteratively trained using historical operational training data to obtain the target load prediction model.
[0069] Step 206: After the target load prediction model meets the standard based on historical operation test data, the real-time operation data is input into the target load prediction model, and the operation load prediction value for each discrete prediction time interval within the load prediction time domain is predicted by the target load prediction model.
[0070] It should be noted that after the load forecasting model is established, it will be validated to ensure the accuracy of the forecasts. Specifically, mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination can be used respectively. Evaluation indicators such as accuracy, robustness and generalization ability are used to comprehensively measure the performance of the target load prediction model in terms of accuracy, robustness and generalization ability. If the prediction accuracy meets the standard, the next step of multi-step load prediction can be carried out directly. If the prediction accuracy is insufficient, the model parameters are adjusted, the model is fine-tuned or reconstructed to optimize the target load prediction model until the target load prediction model meets the standard.
[0071] As an example, step 206 includes: determining the model accuracy evaluation index obtained by fusing the mean absolute error, mean absolute percentage error, and coefficient of determination; determining that the target load prediction model meets the standard when the model accuracy evaluation index is detected to be greater than the preset model accuracy evaluation index threshold; and inputting real-time operating basic data into the target load prediction model; and predicting the operating load prediction value for each discrete prediction time interval within the load prediction time domain length through the target load prediction model; wherein, the load prediction time domain length refers to the preset prediction time range of the model, which can be the next 24 hours or the next 48 hours; the discrete prediction time interval is a number of consecutive and equal-length short time intervals divided within the load prediction time domain length, such as 24 1-hour intervals or 48 0.5-hour intervals; and the operating load prediction value is the output result of the target load prediction model, which can be the load value that the molten salt thermal storage system may face within the discrete prediction time interval t, such as a heating load of 70 MJ / h.
[0072] Step 208: Generate corresponding operation control rules for each discrete prediction time interval based on all predicted operating load values.
[0073] It should be noted that after the predicted electrical and thermal loads are output based on the target load prediction model, specific heat storage-heat release strategies can be formulated for the molten salt thermal storage system to simultaneously meet the peak-shaving and heating needs of the molten salt thermal storage system. It can be understood that the mapping relationship between the predicted operating load value and the operating control rules can be set in advance, so that after obtaining the predicted operating load value, the corresponding operating control rules can be directly queried based on the predicted operating load value.
[0074] As an example, step 208 includes: querying the operation control rules for each discrete forecast time interval using all operational load forecast values as an index.
[0075] Step 210: Perform operation control on the molten salt thermal storage system within the load prediction time domain according to all operation control rules.
[0076] It should be noted that within the load forecast time domain, the molten salt thermal storage system can be operated and controlled according to the selected operation control rules, thereby automatically adjusting the system operation and responding to the forecast load demand in real time.
[0077] As an example, step 210 includes: performing operational control on the molten salt thermal storage system within the load prediction time domain based on operational control rules.
[0078] The aforementioned operation and control method for the molten salt thermal energy storage system first acquires historical operational data, including training data and test data. This divides the historical operational data into training and test data. Then, using the training data, the initial load prediction model is iteratively trained to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time series length of the molten salt thermal energy storage system. After the target load model is tested and found to meet the standards based on the historical test data, real-time operational data is input into the target load prediction model. The target load prediction model then predicts the operational load for each discrete prediction time interval within the load prediction time series, enabling accurate identification of the load demand of the molten salt thermal energy storage system in future discrete time intervals using the compliant target load prediction model. The purpose of this study is to determine the changing trends and numerical ranges, and then generate corresponding operation control rules for each discrete prediction time interval by running composite predicted values. Finally, based on all operation control rules, the molten salt thermal storage system is operated and controlled within the load prediction time domain. Since the target load prediction model is trained and verified based on historical operation data, the accuracy of load demand identification can be ensured. At the same time, by running load prediction values, operation control rules that are fully adapted to the load changes in each discrete prediction time interval can be generated, thus enabling effective control of the molten salt thermal storage system. Therefore, this study overcomes the technical shortcomings of the molten salt thermal storage system, which is prone to low heat storage and release efficiency and slow dynamic response due to the complexity of load fluctuation influencing factors during the operation of the molten salt thermal storage system, even when using neural network models. This improves the control effect of the molten salt thermal storage system.
[0079] In one embodiment, such as Figure 2 As shown, based on all predicted operating load values, corresponding operating control rules are generated for each discrete prediction time interval, including steps 302 to 306:
[0080] Step 302: Based on all predicted operating load values, construct operating constraints of different dimensions for each discrete prediction time interval.
[0081] It should be noted that, in the process of formulating the system's heat storage and release control strategy, to ensure accurate operational control for each discrete prediction time interval, the load prediction time domain length can first be divided into multiple discrete prediction time intervals, which can be represented as follows: This allows for individual decision-making within each discrete prediction time interval, thus enabling the construction of different dimensional operational constraints for each discrete prediction time interval.
[0082] As an example, step 302 includes: taking the candidate operating constraints for each discrete forecast time interval obtained from the query as operating constraints based on all operating load forecast values.
[0083] Step 304: Based on all operational constraints, generate a multi-stage operation control decision model for the molten salt thermal storage system within the load prediction time domain.
[0084] As an example, step 304 includes: fitting a multi-stage operation control decision model for the molten salt thermal storage system within the load prediction time domain, according to all operating constraints.
[0085] Step 306: Output the corresponding operation control rules for each discrete prediction time interval through the multi-stage operation control decision model.
[0086] It should be noted that by using a multi-stage operation control decision model, multi-step predictions can be completed in stages.
[0087] As an example, step 306 includes: outputting the corresponding operation control rules for each discrete prediction time interval through a multi-stage operation control decision model.
[0088] In this embodiment, relying on all predicted operating load values, the power dimension constraints are fully considered during the construction of operating constraints to ensure the power supply and demand balance of the molten salt thermal storage system in each discrete prediction time interval. Then, based on the operating constraints, a multi-stage operation control decision model is generated for the molten salt thermal storage system within the load prediction time domain. Finally, the multi-stage operation control decision model outputs the corresponding operation control rules for each discrete prediction time interval, thereby achieving the goal of matching the corresponding operation control rules for each discrete prediction time interval. Therefore, this lays the foundation for further improving the control effect of the operation control of the molten salt thermal storage system.
[0089] In one embodiment, the operating constraints include the molten salt capacity constraint of the molten salt thermal storage system, the heat storage and release rate constraint of the molten salt thermal storage system, the heat load constraint of the molten salt thermal storage system, and the electrical load constraint of the molten salt thermal storage system.
[0090] It should be noted that, for each discrete prediction time interval, the following parameters and key variables can be defined, specifically including: 1) the heat (MJ) stored in the molten salt during time interval t, which can be expressed as... ;2) The heat (MJ) released by the molten salt in time period t can be expressed as 3) The heat of molten salt storage in the high-temperature tank during time period t (MJ) can be expressed as: ;4) The total usable internal energy (MJ) of the molten salt in the system can be expressed as 5) The steam heat extracted by the unit in time period t (MJ) can be expressed as: 6) The heat generated by the steam during the exothermic process (MJ) can be expressed as: 7) The maximum extraction heat of the unit in time period t (MJ) can be expressed as: ;8) The predicted heat load demand (MJ) for time period t can be expressed as ;9) The predicted electricity load demand (MJ) for time period t can be expressed as ;10) The maximum heat storage (MJ) of the molten salt system during time period t can be expressed as ;11) The maximum heat release (MJ) of the molten salt system during time period t can be expressed as ;12) The maximum thermal storage capacity (MJ) of the high / low temperature molten salt tank can be expressed as follows: / ;13) The minimum allowable thermal storage capacity (MJ) of high / low temperature molten salt tanks can be expressed as follows: / ;14) Molten salt heat exchanger efficiency (%), which can be expressed as .
[0091] As an example, in a molten salt thermal storage system, there are constraints including molten salt capacity (high-temperature tank storage limit), maximum heat storage / release rate constraints, heat load demand satisfaction constraints, and electrical load demand satisfaction constraints (steam extraction capacity constraints).
[0092] In one embodiment, the molten salt capacity constraint of the molten salt thermal storage system includes:
[0093] Minimum energy storage value of the first molten salt tank Real-time energy storage value The maximum energy storage value of the first molten salt tank and the minimum energy storage value of the second molten salt tank. Allowable energy dispatch value for molten salt thermal storage systems The maximum energy storage value of the second molten salt tank.
[0094] Among them, the minimum energy storage value of the first molten salt tank can be specifically the minimum allowable heat storage capacity of the high-temperature molten salt tank; the real-time energy storage value can be specifically the molten salt stored heat of the high-temperature tank in time period t; the maximum energy storage value of the first molten salt tank can be specifically the maximum heat storage capacity of the high-temperature molten salt tank; the minimum energy storage value of the first molten salt tank can be specifically the minimum allowable heat storage capacity of the low-temperature molten salt tank; the energy allowable dispatch value of the molten salt thermal energy storage system can be specifically the total available internal energy of the molten salt in the molten salt thermal energy storage system and the molten salt stored heat of the high-temperature tank in time period t; and the maximum energy storage value of the second molten salt tank can be specifically the maximum heat storage capacity of the low-temperature molten salt tank.
[0095] In one feasible approach, the molten salt capacity constraint of the molten salt thermal storage system can be specifically expressed by the following formula:
[0096] .
[0097] In one embodiment, the heat storage and release rate constraint of the molten salt thermal energy storage system includes:
[0098] 0 Real-time stored heat value of molten salt thermal storage system The maximum heat storage value of the molten salt thermal energy storage system, and 0 The heat value of steam generated by the molten salt thermal storage system The upper limit of real-time heat release of the molten salt thermal storage system.
[0099] The real-time stored heat value of the molten salt thermal storage system can be specifically defined as the heat stored in the molten salt during time period t, and can be expressed as: The maximum heat storage value of a molten salt thermal energy storage system can be specifically defined as the maximum heat storage capacity of the molten salt during time period t, which can be expressed as: The heat value of the steam generated during the heat release process in a molten salt thermal storage system can be specifically expressed as: The real-time heat release upper limit of the molten salt thermal storage system can be specifically defined as the maximum heat release of the molten salt system in time period t, which can be expressed as: .
[0100] As an example, the molten salt heat storage and release rate is constrained at any given time period as follows: , .
[0101] In one embodiment, the heat load constraints of the molten salt thermal storage system include:
[0102] The combined steam heat value is obtained by combining the heat output steam value of the molten salt thermal storage system and the heat output steam generated by the molten salt thermal storage system. Heat load requirements of molten salt thermal storage system.
[0103] The steam heat fusion value is generated from the steam extraction heat value of the molten salt thermal storage system and the steam heat value generated by the heat release of the molten salt thermal storage system. Specifically, the steam extraction heat value of the molten salt thermal storage system can be expressed as the steam heat extracted by the unit in time period t (MJ), which can be represented as... The heat generated by the steam released during the molten salt thermal storage system can be specifically expressed as the heat generated during the heat release process. The heat load demand value of the molten salt thermal storage system can be specifically the predicted heat load demand for time period t, which can be expressed as: .
[0104] As an example, the heat load demand constraint is as follows: the extracted steam is used entirely for molten salt heat exchange, and the actual external steam supply is the sum of the heat generated by the unit's external steam extraction and heat release, i.e. .
[0105] In one embodiment, the electrical load constraints of the molten salt thermal storage system include:
[0106] Real-time steam extraction heat value of molten salt thermal storage system Electrical load requirements for molten salt thermal storage systems.
[0107] The real-time extraction heat value of the molten salt thermal storage system, specifically the steam heat extracted by the unit in time period t (MJ), can be expressed as: The electrical load demand of the molten salt thermal energy storage system, specifically the predicted electrical load demand for time period t, can be expressed as: .
[0108] As an example, the electrical load constraints of a molten salt thermal storage system are as follows: The unit's extraction of steam heat is limited by the electrical load. , where the function Indicates electrical load At that time, the maximum amount of steam heat that the unit can extract can be determined by the unit characteristic curve.
[0109] In one embodiment, a multi-stage operation control decision model for the molten salt thermal storage system over the time domain of load prediction is generated based on all operational constraints, including:
[0110] Based on all operational constraints, a multi-stage optimization objective function is constructed; based on the energy balance relationship of the molten salt thermal storage system, state transition equations for adjacent stages within the load prediction time domain are established, and an initial multi-stage operation control decision model is generated based on the objective function, state transition equations, and all operational constraints; the initial multi-stage operation control decision model is iteratively optimized to obtain a multi-stage operation control decision model.
[0111] It should be noted that the extracted steam heat is diverted during the same period, with part entering molten salt thermal storage and the other part being directly used for external steam supply, which can be expressed as follows: ,in, f t ∈ [ 0 , 1 ] This refers to the proportion of extracted steam allocated to the thermal storage system. It is understandable that, with the goal of improving the operational flexibility and economy of the molten salt thermal storage system, a multi-stage objective function can be constructed, as shown below:
[0112]
[0113] The first objective is to maximize the total amount of steam heat supplied; the second objective is to minimize changes in the operating state of the molten salt thermal storage system, penalizing excessive fluctuations in the thermal storage state to improve the stability of the molten salt thermal storage system. A penalty coefficient for changes in the energy storage state of molten salt (configurable); furthermore, the heat in the high-temperature molten salt tank at each stage can be specified. As a state variable, and For decision variables; the state transition equation can be defined as The boundary conditions are shown in the constraint section. The objective function for stage t can be expressed as:
[0114]
[0115] Among them, record Let be the maximum cumulative return at time t. Then the dynamic programming recurrence relation is:
[0116] V t ( E h , t ) = max Q store , t , Q release , t [ R t ( E h , t , Q store , t , Q release , t ) + V t − 1 ( E h , t − 1 ) ]
[0117] Initial conditions are defined as Furthermore, starting from t=1, the calculation proceeds stage by stage: 1) Given all possible states for each stage, enumerate all legal decisions, and using the decisions obtained in the previous step, determine the optimal heat storage and release strategy for each stage; 2) Calculate and save the optimal decision and maximum benefit for each stage; 3) Recursively calculate to the end of the prediction period t=T, and save the state and cumulative benefit at this time as the initial value for the next stage; Finally, based on the above steps, calculate the heat storage and release strategies for each time period. and This leads to the formulation of a heat storage and release control strategy for the molten salt system.
[0118] Understandably, after setting up multi-stage operation control decision models, human-machine intelligent hybrid systems can also be used as a basis for determining whether the molten salt thermal storage system can meet the actual load demand. The human-machine intelligent hybrid system can be divided into four core sub-modules: experience acquisition and rule-making, strategy consistency verification, human-machine collaborative decision-making and execution, and experience base learning and updating. The functions and roles of each sub-module will be described below.
[0119] (1) Experience acquisition and rule-making: Based on the historical operating conditions of the unit and the operation rules of experts, IF-THEN type control logic rules are extracted. Each rule has an applicable domain (such as load range, equipment operating status, etc.) and is accompanied by priority and confidence, and finally forms a searchable and maintainable rule base. This module can realize the structured storage and online call of expert knowledge, and provide a standardized basis for subsequent strategy verification and manual decision-making. (2) Strategy consistency verification: First, the optimal storage-heat release strategy calculated based on the dynamic programming method is obtained and a candidate strategy set is formed; the upper and lower limits of tank capacity, the upper limit of storage / heat release power, and the heat load satisfaction are constrained to form a constraint set, and the time period action in the verification is checked. If it is satisfied, it is retained; if it is not satisfied, the action of the time period is adjusted according to the principle of minimum modification, and if it is still not feasible, it is backed to the conservative feasible solution. Finally, the candidate sequence after consistency verification is obtained. (3) Human-machine collaborative decision-making and execution: For each time period t, the applicable rule set is retrieved according to the system operating conditions. ;in, Let t be the working condition vector. This is the rule trigger function, triggered when rule r is in the current working condition. When applicable, the function takes the value 1; otherwise, it takes the value 0. Then, the deviation between the candidate action and the applicable rule is calculated. ,like (Set tolerance), then set ;like In the manual adjudication stage, operators select, integrate, or fine-tune candidate solutions and applicable rules to arrive at a final decision. ;for Perform a consistency check again, and write the data after it passes the check. (4) Experience base learning and updating: After the strategy is executed, the system will collect several key performance indicators (such as external heating satisfaction rate, improvement in operating benefits, system response time and load tracking accuracy, etc.) and record the performance corresponding to the strategy execution. Subsequently, the applicable scope, priority and confidence of the rules in the experience base will be updated according to the collected results. Valid strategies formed in manual adjudication will be added as new rules, while rules that have been invalid for a long time or have contradictions will be downgraded or removed. Through this process, a closed-loop iterative mechanism of 'optimization solution - consistency check - manual adjudication - execution - evaluation - feedback' will be formed, thereby realizing the continuous correction and improvement of the control strategy and gradually improving the intelligence level and overall efficiency of the system operation.
[0120] In one feasible approach, refer to Figure 3 , Figure 3 This is a flowchart illustrating the operation and control process of a molten salt thermal energy storage system. First, historical operational data of the system is acquired, including training data and test data. Then, based on the training data, the initial load prediction model is iteratively trained to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal energy storage system. After verifying the target load prediction model's compliance with historical test data, real-time operational data is input into the target load prediction model, and the target load prediction model predicts the load prediction time domain length. The system generates operating load forecasts for each discrete forecast time interval within a given time period. Based on these forecasts, corresponding operating control rules are generated for each interval. These rules are then used to control the molten salt thermal storage system within the load forecast time domain. During actual operation, the load forecast is updated periodically by adding new data, and the system re-plans the heat storage and release strategies for subsequent periods according to the aforementioned strategy. This enhances the system's adaptability to uncertainties or load fluctuations. Load demand satisfaction analysis is then performed: the system's operating status is monitored in real-time to determine if the current load demand is met. If the demand is met, the current operating control strategy is maintained; if not, the heat storage-release strategy is adjusted promptly through rolling correction, and the main steam supply can be increased by adjusting the boiler load. Finally, the process terminates, and the control system enters standby or periodic monitoring mode until the next iteration. As the number of iterations increases, the model's prediction accuracy gradually increases, improving the performance of the molten salt thermal storage operation control system. Based on accurate predictions, the control system can formulate more precise heat storage control strategies, ultimately forming an adaptive control closed loop.
[0121] Understandably, this scheme, through a closed-loop system design of multi-step load prediction, thermal storage and release operation control strategy formulation, and real-time feedback adjustment of operating status, effectively improves the operational stability and economy of molten salt thermal storage systems in coal-fired power units. It is applicable to various operation scheduling and optimization control scenarios of cogeneration coal-fired power units. Simultaneously, because the target load prediction model is trained and validated based on historical operational training data, it ensures the accuracy of load demand identification. Furthermore, through the predicted operating load values, it can generate operation control rules that are fully adapted to load changes in each discrete prediction time interval, thus enabling effective control of the molten salt thermal storage system's operation at all times. Therefore, it overcomes the technical shortcomings of complex load fluctuation factors affecting molten salt thermal storage systems during operation, which, even with neural network models, struggle to accurately capture the dynamic load fluctuation patterns under multi-factor coupling, leading to low thermal storage and release efficiency and slow dynamic response in molten salt thermal storage systems. Thus, it improves the control effect of operating molten salt thermal storage systems.
[0122] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.
[0123] Based on the same inventive concept, this application also provides an operation control device for a molten salt thermal storage system to implement the operation control method of the molten salt thermal storage system described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the operation control device for a molten salt thermal storage system provided below can be found in the limitations of the operation control method for the molten salt thermal storage system described above, and will not be repeated here.
[0124] In one exemplary embodiment, such as Figure 4 As shown, an operation control device for a molten salt thermal energy storage system is provided, comprising: an acquisition module 401, a training module 402, a prediction module 403, a generation module 404, and a control module 405, wherein:
[0125] The acquisition module 401 is used to acquire historical operational basic data of the molten salt thermal storage system, wherein the historical operational basic data includes historical operational basic training data and historical operational basic test data.
[0126] Training module 402 is used to iteratively train the initial load prediction model of the molten salt thermal storage system based on historical operational training data to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system.
[0127] The prediction module 403 is used to input real-time operational data into the target load prediction model after the target load prediction model has met the test results based on historical operational data, and to predict the operational load prediction value for each discrete prediction time interval within the load prediction time domain through the target load prediction model.
[0128] The generation module 404 is used to generate corresponding operation control rules for each discrete prediction time interval based on all predicted operating load values.
[0129] The control module 405 is used to control the operation of the molten salt thermal storage system within the load prediction time domain according to all operating control rules.
[0130] In one embodiment, the generation module 404 is further configured to:
[0131] Based on all predicted operating load values, different dimensions of operating constraints are constructed for each discrete prediction time interval. Based on all operating constraints, a multi-stage operation control decision model is generated for the molten salt thermal storage system within the load prediction time domain. The operation control rules corresponding to each discrete prediction time interval are output through the multi-stage operation control decision model.
[0132] In one embodiment, the operating constraints include the molten salt capacity constraint of the molten salt thermal storage system, the heat storage and release rate constraint of the molten salt thermal storage system, the heat load constraint of the molten salt thermal storage system, and the electrical load constraint of the molten salt thermal storage system.
[0133] In one embodiment, the molten salt capacity constraint of the molten salt thermal storage system includes:
[0134] Minimum energy storage value of the first molten salt tank Real-time energy storage value The maximum energy storage value of the first molten salt tank and the minimum energy storage value of the second molten salt tank. Allowable energy dispatch value for molten salt thermal storage systems The maximum energy storage value of the second molten salt tank.
[0135] In one embodiment, the heat storage and release rate constraint of the molten salt thermal storage system includes:
[0136] 0 Real-time stored heat value of molten salt thermal storage system The maximum heat storage value of the molten salt thermal energy storage system, and 0 The heat value of steam generated by the molten salt thermal storage system The upper limit of real-time heat release of the molten salt thermal storage system.
[0137] In one embodiment, the heat load constraints of the molten salt thermal storage system include:
[0138] The combined steam heat value is obtained by combining the heat output steam value of the molten salt thermal storage system and the heat output steam generated by the molten salt thermal storage system. Heat load requirements of molten salt thermal storage system.
[0139] In one embodiment, the electrical load constraints of the molten salt thermal storage system include:
[0140] Real-time steam extraction heat value of molten salt thermal storage system Electrical load requirements for molten salt thermal storage systems.
[0141] In one embodiment, the generation module 404 is further configured to:
[0142] Based on all operational constraints, a multi-stage optimization objective function is constructed; based on the energy balance relationship of the molten salt thermal storage system, state transition equations for adjacent stages within the load prediction time domain are established, and an initial multi-stage operation control decision model is generated based on the objective function, state transition equations, and all operational constraints; the initial multi-stage operation control decision model is iteratively optimized to obtain a multi-stage operation control decision model.
[0143] Each module in the operation control device of the aforementioned molten salt thermal storage system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.
[0144] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5As shown. The computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements an operation control method for a molten salt thermal storage system. Those skilled in the art will understand that... Figure 5 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.
[0145] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.
[0146] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.
[0147] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.
[0148] Those skilled in the art will understand that all or part of the processes in the methods of 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 of the above methods. 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). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.
[0149] 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.
[0150] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.
Claims
1. A method for operating and controlling a molten salt thermal energy storage system, characterized in that, The method includes: Obtain historical operational data of the molten salt thermal storage system, wherein the historical operational data includes historical operational training data and historical operational test data; Based on the historical operational training data, the initial load prediction model of the molten salt thermal storage system is iteratively trained to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system. After the target load prediction model is tested and found to meet the standards based on the historical operation data, the real-time operation data is input into the target load prediction model, and the operation load prediction value for each discrete prediction time interval within the load prediction time domain is predicted by the target load prediction model. Based on all predicted operating load values, generate corresponding operating control rules for each discrete prediction time interval; Based on all operational control rules, the molten salt thermal storage system is operated and controlled within the load prediction time domain.
2. The method according to claim 1, characterized in that, The step of generating corresponding operation control rules for each discrete prediction time interval based on all predicted operating load values includes: Based on all predicted operating load values, different dimensions of operating constraints are constructed for each discrete prediction time interval; Based on all operational constraints, a multi-stage operation control decision model for the molten salt thermal storage system is generated within the time domain of the load prediction. The multi-stage operation control decision model outputs the operation control rules corresponding to each discrete prediction time interval.
3. The method according to claim 2, characterized in that, The operational constraints include the molten salt capacity constraint of the molten salt thermal storage system, the heat storage and release rate constraint of the molten salt thermal storage system, the heat load constraint of the molten salt thermal storage system, and the electrical load constraint of the molten salt thermal storage system.
4. The method according to claim 3, characterized in that, The molten salt capacity constraint of the molten salt thermal storage system includes: Minimum energy storage value of the first molten salt tank Real-time energy storage value The maximum energy storage value of the first molten salt tank and the minimum energy storage value of the second molten salt tank. The allowable energy dispatch value of the molten salt thermal storage system The maximum energy storage value of the second molten salt tank.
5. The method according to claim 3, characterized in that, The heat storage and release rate constraints of the molten salt thermal storage system include: 0 The real-time stored heat value of the molten salt thermal storage system The maximum heat storage value of the molten salt thermal storage system, and 0 The heat value of the steam generated by the molten salt thermal storage system The upper limit of real-time heat release of the molten salt thermal storage system.
6. The method according to claim 3, characterized in that, The heat load constraints of the molten salt thermal storage system include: The steam heat fusion value is obtained by combining the heat output steam value of the molten salt thermal storage system and the heat output steam generated by the molten salt thermal storage system. The heat load requirement value of the molten salt thermal storage system.
7. The method according to claim 3, characterized in that, The electrical load constraints of the molten salt thermal storage system include: Real-time steam extraction heat value of the molten salt thermal storage system The electrical load requirement of the molten salt thermal storage system.
8. The method according to claim 2, characterized in that, The step of generating a multi-stage operation control decision model for the molten salt thermal storage system within the load prediction time domain based on all operational constraints includes: Based on all the aforementioned operational constraints, construct a multi-stage optimization objective function; Based on the energy balance relationship of the molten salt thermal storage system, state transition equations for adjacent stages within the load prediction time domain are established, and an initial multi-stage operation control decision model is generated based on the objective function, the state transition equations, and all operating constraints. The initial multi-stage operation control decision model is iteratively optimized to obtain a multi-stage operation control decision model.
9. An operation control device for a molten salt thermal storage system, characterized in that, The device includes: The acquisition module is used to acquire the historical basic operating data of the molten salt thermal storage system, wherein the historical basic operating data includes historical basic operating training data and historical basic operating test data; The training module is used to iteratively train the initial load prediction model of the molten salt thermal storage system based on the historical operational training data to obtain the target load prediction model. The initial load prediction model is selected based on the load prediction time domain length of the molten salt thermal storage system. The prediction module is used to input real-time operational data into the target load prediction model after testing the target load prediction model based on the historical operational data, and to predict the operational load prediction value for each discrete prediction time interval within the load prediction time domain through the target load prediction model. The generation module is used to generate corresponding operation control rules for each discrete prediction time interval based on all predicted operating load values. The control module is used to control the operation of the molten salt thermal storage system within the load prediction time domain according to all operation control rules.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.
