Offline intelligent optimization control method for chemical and volume control system of nuclear power plant
By adopting offline intelligent optimization control method in the chemical and volume control system of nuclear power plants, using mathematical models and multi-objective optimization algorithms, the problems of overshoot and long response time of RCV are solved, and more efficient and stable control effects are achieved.
Patent Information
- Application Number
- CN202411746012.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-29
- Publication Date
- 2025-04-15
AI Technical Summary
The existing chemical and volume control system (RCV) of nuclear power plants has problems such as severe overshoot and long response time, which cannot meet the safety and economic requirements of modern nuclear power plants.
Offline intelligent optimization control method is adopted to detect the operating status in real time, obtain historical data, establish a variety of mathematical models (such as polynomial regression, support vector regression, etc.), and find the best operation method through multi-objective optimization algorithm to achieve optimization control of RCV.
It effectively solves the problems of overshoot and long response time of RCV, improves control accuracy and stability, reduces energy consumption and maintenance costs, and ensures the safe and efficient operation of nuclear power plants.
Smart Images
Figure CN119833185A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence and automated control technology, and in particular to an offline intelligent optimization control method for a nuclear power plant chemical and volume control system, a computer device, and a computer-readable storage medium. Background Art
[0002] With the development of nuclear power technology, the requirements for the safety and economy of nuclear power plants are becoming higher and higher. As one of the auxiliary systems of nuclear power plants, the chemical and volume control system (RCV) of nuclear power plants is mainly responsible for maintaining the primary side pressure and water level of the reactor coolant, and ensuring the water quality by adjusting the pH value of the primary coolant. At the same time, RCV also bears the important responsibility of preventing the spillage of radioactive materials. Therefore, the safe and stable operation of RCV is crucial to the safety of the entire nuclear power plant.
[0003] At present, the existing RCV usually adopts PID controller for closed-loop control, but this method has disadvantages such as severe overshoot and long response time, which cannot meet the requirements of modern nuclear power plants. Summary of the invention
[0004] In order to improve the above technical problems, the technical solution of the first aspect of the present invention provides an offline intelligent optimization control method for a nuclear power plant chemical and volume control system, which is used to solve the problems of severe overshoot and long response time of RCV in the prior art.
[0005] An offline intelligent optimization control method for a nuclear power plant chemical and volume control system provided by a technical solution of the present invention, referred to as a control method, comprises real-time detection of the operating state of the nuclear power plant chemical and volume control system, and acquisition of first historical data, including a primary circuit temperature, a primary circuit flow, a pressurizer liquid level, and a top-up flow of the nuclear power plant chemical and volume control system;
[0006] Setting a first mathematical model according to the first historical data; wherein the first mathematical model is trained using the first historical data as a sample;
[0007] Real-time detection of the operating status of the nuclear power plant chemical and volume control system, and acquisition of the second historical data, including the charging flow, charging valve opening, discharge flow, and charging pump speed signal of the nuclear power plant chemical and volume control system;
[0008] Setting a second mathematical model according to the second historical data; wherein the second mathematical model is trained using the second historical data as a sample;
[0009] At least one objective function is set; wherein each objective function corresponds to a certain performance index of the chemical and volume control system of the nuclear power plant;
[0010] Training the first mathematical model and the second mathematical model with the optimal solution of the objective function as the goal;
[0011] The trained first mathematical model and the second mathematical model are used to predict the state change trend of the nuclear power plant chemical and volume control system in the future, and the best operation mode is sought through a multi-objective optimization algorithm to achieve optimized control.
[0012] In the above technical solution, the control method further includes the following steps: saving the first historical data and the second historical data.
[0013] In the above technical scheme, the first mathematical model is one or any combination of a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosting tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, and a LightGBM regression model; and / or the second mathematical model is one or any combination of a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosting tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, and a LightGBM regression model.
[0014] In the above technical solution, the objective function includes the stabilizer liquid level deviation rate, the charging flow fluctuation rate and the discharging flow fluctuation rate.
[0015] In the above technical solution, the optimal solution of the objective function is solved by a global optimization algorithm, wherein the global optimization algorithm is a genetic algorithm, a particle swarm algorithm or a simulated annealing algorithm.
[0016] In the above technical solution, the hyperparameters of the first mathematical model are the same as the hyperparameters of the second mathematical model; or the hyperparameters of the first mathematical model are partially the same as the hyperparameters of the second mathematical model and partially different; or the hyperparameters of the first mathematical model are different from the hyperparameters of the second mathematical model.
[0017] In the above technical solution, the control method further includes the following steps: solidifying the trained first mathematical model and the second mathematical model.
[0018] In the above technical solution, the operation method for realizing the optimized control includes adjusting the rotation speed of the upstroke pump and / or the opening of the upstroke valve.
[0019] In summary, the offline intelligent optimization control method and guarantee method of the nuclear power plant chemical and volume control system provided by the present invention can at least have the following beneficial effects:
[0020] (1) The present invention obtains first historical data and establishes a first mathematical model based on the first historical data, then obtains second historical data and establishes a second mathematical model based on the second historical data, and finally determines an objective function based on the first mathematical model and the second mathematical model and obtains an optimal solution based on the objective function, thereby realizing offline intelligent optimization control of a nuclear power plant chemical and volume control system, and solving the problems of severe overshoot and long response time of RCV in the prior art.
[0021] (2) The present invention takes into account the changes in multiple parameters in the RCV and performs machine learning based on the mutual influence between the multiple parameters, thereby improving the control effect.
[0022] (3) The combination of the two models can improve the prediction accuracy and make it easier to find the optimal configuration, which helps to improve the offline intelligent optimization control effect of RCV.
[0023] The technical solution of the second aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant in any one of the technical solutions of the first aspect of the present invention can be implemented. Therefore, the computer device provided by the technical solution of the second aspect of the present invention has all the beneficial effects of any of the technical solutions of the first aspect of the present invention, which will not be repeated here.
[0024] The technical solution of the third aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed, the steps of the offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant in any one of the technical solutions of the first aspect of the present invention can be implemented. Therefore, the computer-readable storage medium provided by the technical solution of the third aspect of the present invention has all the beneficial effects of any of the technical solutions of the first aspect of the present invention, which will not be repeated here.
[0025] It should be noted that those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, a magnetic tape storage, or any other medium of a computer scale that can be used to carry or store data. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 The figure is a flow chart of an offline intelligent optimization control method for a nuclear power plant chemical and volume control system in one embodiment of the present invention. DETAILED DESCRIPTION
[0027] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the protection scope of the present invention.
[0028] Some embodiments of the present invention provide an offline intelligent optimization control method for a chemical and volume control system of a nuclear power plant.
[0029] like Figure 1 As shown, the offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant provided in this embodiment includes the following steps:
[0030] Obtaining first historical data;
[0031] Establishing a first mathematical model based on the first historical data;
[0032] Obtaining second historical data;
[0033] establishing a second mathematical model based on the second historical data;
[0034] determining an objective function according to the first mathematical model and the second mathematical model;
[0035] Obtain the optimal solution based on the objective function.
[0036] Furthermore, before acquiring the first historical data, the method further includes: collecting first measurement data including 7 related variables; and obtaining the first historical data according to the first measurement data.
[0037] Furthermore, the method further includes: obtaining second historical data according to the first measurement data:
[0038] Furthermore, the first historical data includes: historical data of primary circuit temperature, primary circuit flow, stabilizer liquid level, and charging flow.
[0039] Furthermore, the second historical data includes: historical data of the charging flow, the charging valve opening, the discharge flow, and the charging pump speed signal.
[0040] Furthermore, the first mathematical model and / or the second mathematical model is: one or any combination of a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosting tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, and a LightGBM regression model.
[0041] Furthermore, the optimal solution is: the upstroke pump speed signal and / or the upstroke valve opening signal.
[0042] The present invention is further described below by means of a specific embodiment.
[0043] Example 1
[0044] See also Figure 1 As shown, this embodiment provides a method for chemical and volume control of a nuclear power plant, comprising the following steps:
[0045] S101. Obtain first historical data.
[0046] In this embodiment, the status of each RCV device can be monitored in real time through field instruments, for example, a temperature sensor can be used to detect the primary circuit temperature, a flow meter can be used to detect the primary circuit flow, a liquid level meter can be used to detect the stabilizer liquid level, etc. The operation data of each device can also be collected through DCS or other control systems and saved in a database for subsequent analysis.
[0047] S102: Establish a first mathematical model according to the first historical data.
[0048] In this embodiment, a suitable modeling method can be selected according to different application scenarios. For example, a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosting tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, a LightGBM regression model, etc. can be selected.
[0049] S103: Obtain second historical data.
[0050] In this embodiment, the status of each RCV device can be monitored in real time through field instruments, for example, a temperature sensor can be used to detect the primary circuit temperature, a flow meter can be used to detect the primary circuit flow, a liquid level meter can be used to detect the stabilizer liquid level, etc. The operation data of each device can also be collected through DCS or other control systems and saved in a database for subsequent analysis.
[0051] S104: Establish a second mathematical model according to the second historical data.
[0052] In this embodiment, a suitable modeling method can be selected according to different application scenarios. For example, a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosting tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, a LightGBM regression model, etc. can be selected.
[0053] S105. Determine an objective function according to the first mathematical model and the second mathematical model.
[0054] In this embodiment, multiple objective functions may be set, corresponding to different performance indicators, for example, multiple objective functions such as the stabilizer liquid level deviation rate, the charging flow fluctuation rate, and the discharge flow fluctuation rate may be set.
[0055] S106. Obtain an optimal solution according to the objective function.
[0056] In this embodiment, the optimal solution of the objective function can be solved by a global optimization algorithm such as a genetic algorithm, a particle swarm algorithm, a simulated annealing algorithm, etc. The optimal solution is an operation mode that can make all objective functions reach the optimal state, for example, it can be a speed signal of the upstroke pump and / or an opening signal of the upstroke valve.
[0057] In this embodiment, multiple mathematical models are established by statistical analysis and machine learning of historical data, and these models are used to predict the state change trend of RCV in the future, and then the best operation mode is sought through a multi-objective optimization algorithm. This method can not only improve the control accuracy and stability of RCV, but also reduce energy consumption and maintenance costs, which is conducive to the safe and efficient operation of nuclear power plants.
[0058] In the description of this specification, the description of the terms "one embodiment", "some embodiments", "specific embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0059] In the description of this specification, specific features, structures, materials or characteristics may be combined in an appropriate manner in any one or more embodiments or examples.
[0060] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, technicians familiar with the field may make various corresponding changes and modifications based on the present invention, but these changes and modifications should all fall within the scope of protection of the claims of the present invention.
Claims
1. An offline intelligent optimization control method for a nuclear power plant chemical and volume control system, characterized in that: The steps include: Real-time detection of the operating status of the nuclear power plant chemical and volume control system, and acquisition of the first historical data, including the primary circuit temperature, primary circuit flow, pressurizer liquid level and top-up flow of the nuclear power plant chemical and volume control system: Setting a first mathematical model according to the first historical data; wherein the first mathematical model is trained using the first historical data as a sample; Real-time detection of the operating status of the nuclear power plant chemical and volume control system, and acquisition of the second historical data, including the charging flow, charging valve opening, discharge flow, and charging pump speed signal of the nuclear power plant chemical and volume control system; Setting a second mathematical model according to the second historical data; wherein the second mathematical model is trained using the second historical data as a sample; At least one objective function is set; wherein each objective function corresponds to a certain performance index of the chemical and volume control system of the nuclear power plant; Training the first mathematical model and the second mathematical model with the optimal solution of the objective function as the goal; The trained first mathematical model and the second mathematical model are used to predict the state change trend of the nuclear power plant chemical and volume control system in the future, and the best operation mode is sought through a multi-objective optimization algorithm to achieve optimized control.
2. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The following steps are also included: The first historical data and the second historical data are saved.
3. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The first mathematical model is one or any combination of a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosted tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, and a LightGBM regression model; and / or The second mathematical model is one or any combination of a polynomial regression model, a support vector regression model, a random forest regression model, a gradient boosting tree regression model, a Gaussian process regression model, a deep neural network regression model, an XGBoost regression model, and a LightGBM regression model.
4. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The objective function includes the stabilizer liquid level deviation rate, the charging flow fluctuation rate and the discharge flow fluctuation rate.
5. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The optimal solution of the objective function is solved by a global optimization algorithm, wherein the global optimization algorithm is a genetic algorithm, a particle swarm algorithm or a simulated annealing algorithm.
6. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The hyperparameters of the first mathematical model are the same as the hyperparameters of the second mathematical model; or The hyperparameters of the first mathematical model are partially the same as and partially different from the hyperparameters of the second mathematical model; or the hyperparameters of the first mathematical model are all different from the hyperparameters of the second mathematical model.
7. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The following steps are also included: The trained first mathematical model and the second mathematical model are solidified.
8. The offline intelligent optimization control method for the chemical and volume control system of a nuclear power plant according to claim 1, characterized in that: The operation mode of realizing the optimized control includes adjusting the speed of the upstroke pump and / or the opening of the upstroke valve.
9. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the off-line intelligent optimization control method for the chemical and volume control system of a nuclear power plant as described in any one of claims 1 to 8 can be implemented.
10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed, the steps of the off-line intelligent optimization control method for the chemical and volume control system of a nuclear power plant as described in any one of claims 1 to 8 can be implemented.