Intelligent sequencing method and system for reactor thermal hydraulic experiment working conditions

Abstract

The invention discloses an intelligent sequencing method and system for reactor thermal-hydraulic experiment working conditions. The intelligent sequencing method comprises the steps that firstly, a pre-constructed machine learning model is adopted to achieve automatic prediction of switching time between the working conditions of a thermal-hydraulic experiment; then, each working condition point is used as a node, the switching time between working conditions is used as an edge between the nodes, that is, the length of the edge represents the time distance between the nodes, the longer the switching time is, the longer the edge is, the smaller the length of an anti-positive edge is, and a directed graph of the working condition switching time is generated; and finally, carrying out shortest path search on the digraph of the working condition switching time by adopting an ant colony algorithm to obtain a path with the shortest total switching time, namely, a working condition sequence with the shortest completion time, and realizing working condition control according to the working condition sequence with the shortest completion time, so that the completion time of the thermal hydraulic experiment can be shortest, the experiment efficiency is effectively improved, and the working condition control is realized. And the whole process is fully automatically realized, so that the working intensity of experimenters is greatly reduced.

Description

Technical Field

[0001] The present application relates to the technical field of nuclear reactor thermal hydraulic system experimental research, and in particular to an intelligent sorting method and system for reactor thermal hydraulic experimental conditions. Background Art

[0002] Reactor thermal hydraulic experiment is a key link in the nuclear reactor R&D process. Thermal hydraulic experiment needs to complete many working conditions, and the working condition conversion often takes a long time. Traditional working condition conversion is mainly achieved manually by relevant experimenters. In addition, the working condition sequence is usually determined manually based on experience or convenient operation. Calculating the working condition conversion sequence will take a lot of time, which will lead to long experiment time, low efficiency, high experiment cost, and high work intensity of experimenters.

[0003] Therefore, it is urgent to study an intelligent sequencing technology for experimental conditions that can reasonably arrange the sequence of thermal hydraulic experimental conditions, reduce experimental time, reduce the workload of experimenters, and improve experimental efficiency, so as to provide technical support for the intelligent operation of thermal hydraulic experiments. Summary of the invention

[0004] In order to solve the problems that the conversion of existing thermal-hydraulic experimental conditions mainly relies on manpower, resulting in long experimental time, high experimental cost, low experimental efficiency, and high work intensity of experimental personnel, this application proposes an intelligent sorting method and system for reactor thermal-hydraulic experimental conditions. This application adopts an intelligent optimization sorting algorithm, which can reasonably sort the conditions through an intelligent algorithm without human intervention, effectively improve the experimental efficiency, and provide a prerequisite for the intelligent operation of thermal-hydraulic experiments.

[0005] In the first aspect, the present application is implemented through the following technical solutions:

[0006] An intelligent sorting method for reactor thermal hydraulic test conditions, the intelligent sorting method comprising:

[0007] Based on the thermal hydraulic test condition data, the pre-built machine learning model and various safety boundary constraints, the switching time between various conditions of the thermal hydraulic test is predicted;

[0008] According to the working conditions of the thermal hydraulic experiment and the switching time between the working conditions, a directed graph of the working condition point switching time is generated; the directed graph of the working condition point switching time uses each working condition point as a node and the switching time as an edge between nodes to represent the time distance between nodes;

[0009] An ant colony algorithm is used to perform a shortest path traversal search on the directed graph of the switching time of the operating point, and a ranking of the experimental operating points with the shortest total switching time is obtained;

[0010] The experimental operating points with the shortest total switching time are sorted and corresponding operating point switching control is performed.

[0011] In some embodiments, the machine learning model building process specifically includes:

[0012] Obtain historical working condition data and historical working condition switching time of thermal hydraulic experiments, and process to obtain each historical working condition switching time and its corresponding change amount of each historical working condition data;

[0013] By using the processed historical working condition switching time and the corresponding historical working condition data changes, fitting regression is performed to obtain the relationship function between the working condition switching time and the data changes of each working condition, and various safety boundaries are used as constraint conditions;

[0014] The data variation of the working condition point to be tested is input into the relationship function to predict the switching time between the working condition points.

[0015] In some embodiments, the relationship function between the operating condition switching time and the change amount of each operating condition data obtained by processing and fitting regression is obtained by utilizing the historical operating condition switching time obtained by processing and the change amount of each historical operating condition data, specifically including:

[0016] Construct multivariate polynomial fitting regression function;

[0017] Substituting the historical operating condition data changes corresponding to the processed several historical operating condition switching times into the multivariate polynomial fitting regression function to predict the operating condition switching time; taking the processed several historical operating condition switching times as measured values;

[0018] Based on the predicted value and the measured value, the coefficients of the multivariate polynomial fitting regression function are fitted, and the obtained coefficient values ​​are substituted into the multivariate polynomial fitting regression function to obtain the relationship function between the operating condition switching time and the change amount of each operating condition data.

[0019] In some embodiments, the machine learning model building process specifically includes:

[0020] By constructing a thermal hydraulic model of the research object, the system characteristics are studied, the transfer function model of the switching time and each working condition is obtained, and various safety boundary constraints are introduced into the model;

[0021] The switching time between any two operating points can be quickly obtained through the transfer function model.

[0022] In some embodiments,

[0023] The constraints of the safety boundary specifically include:

[0024] Constraints on the rate of temperature increase, rate of temperature decrease, rate of flow change, and rate of pressure change.

[0025] In some implementations, the ant colony algorithm is used to perform a shortest path traversal search on the directed graph of the operating point switching time to obtain a ranking of the experimental operating points with the shortest total switching time, specifically including:

[0026] Initialize parameters and initialize the same amount of pheromone for all paths;

[0027] The solution space is constructed, and the ants choose the next path to take based on the pheromone concentration and heuristic information on the path;

[0028] Update pheromones; determine whether the number of iterations has been reached. If so, output the optimal solution and end the process; otherwise, record the number of iterations + 1, clear the path record table, and return to the step of constructing the solution space;

[0029] The pheromone is a probability of an ant choosing a path constructed based on the pheromone concentration and a heuristic factor; the heuristic factor is the inverse of the path length, and the pheromone concentration increases accordingly as the ants pass by.

[0030] In some implementations, the probability of an ant selecting a path is expressed as:

[0031]

[0032] In the formula, α and β are regulatory factors used to adjust the pheromone concentration τ ij (t) and the heuristic factor η ij The role between; allowed k Indicates the path that ant k has not yet walked.

[0033] In some embodiments, the updating formula of the pheromone concentration is:

[0034] τ ij (t+1)=(1-ρ)·τ ij (t)+Δτ ij

[0035] Where ρ is a number between 0 and 1, (1-ρ) is the volatility factor, and Δτ ij represents the total amount of pheromone left by ants on all paths from i to j after traversing all working points once, τ ij (t) represents the pheromone concentration on path i to j at time t, τ ij (t+1) represents the pheromone concentration on the path from i to j at time t+1, and m represents the number of ants in the algorithm.

[0036] In a second aspect, the present application proposes an intelligent sorting system for reactor thermal hydraulic test conditions, the intelligent sorting system comprising:

[0037] A calculation module, wherein the calculation module predicts the switching time between various working conditions of the thermal hydraulic experiment based on the working condition data of the thermal hydraulic experiment in combination with a pre-built machine learning model and various safety boundary constraints;

[0038] A directed graph construction module, wherein the directed graph construction module generates a directed graph of the switching time of the working condition points according to the working conditions of the thermal hydraulic experiment and the switching time between the working conditions; the directed graph of the switching time of the working condition points uses each working condition point as a node and the switching time as an edge between the nodes to represent the time distance between the nodes;

[0039] An optimization module, wherein the optimization module uses an ant colony algorithm to perform a shortest path traversal search on the directed graph of the operating point switching time to obtain a ranking of experimental operating points with the shortest total switching time;

[0040] And, a control module, the control module sorts the experimental operating points with the shortest total switching time and performs corresponding operating point switching control.

[0041] In some embodiments, the calculation module further includes:

[0042] A data processing unit, the data processing unit is used to obtain historical working condition data and historical working condition switching time of thermal hydraulic experiments, and process to obtain each historical working condition switching time and its corresponding change amount of each historical working condition data;

[0043] A fitting regression unit, wherein the fitting regression unit uses the switching time of each historical working condition and the corresponding change amount of each historical working condition data to obtain a relationship function between the switching time and the change amount of each historical working condition switching data by fitting regression, and uses various safety boundaries as constraint conditions;

[0044] And, a prediction unit, wherein the prediction unit inputs the data change of the experimental operating point into the relationship function to calculate the switching time between each experimental operating point.

[0045] In a fourth aspect, the present application proposes a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0046] In a fourth aspect, the present application proposes a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0047] The present application proposes an intelligent sorting method and system for reactor thermal hydraulic experimental conditions. First, a pre-built machine learning model is used to realize the automatic prediction of the switching time between various operating conditions of the thermal hydraulic experiment; then each operating condition point is used as a node, and the switching time between various operating conditions is used as the edge between the nodes, that is, the length of the edge represents the time distance between the nodes, and the longer the switching time, the longer the length of the edge, and anyway, the shorter the length of the edge is, thereby generating a directed graph of the operating condition switching time; finally, an ant colony algorithm is used to perform a shortest path search on the directed graph of the operating condition switching time to obtain the path with the shortest total switching time, that is, the operating condition sorting with the shortest completion time, and the operating condition control is realized according to the operating condition sorting with the shortest completion time, which can realize the shortest completion time of the thermal hydraulic experiment, effectively improve the experimental efficiency, reduce the experimental cost, and the whole process is fully automatic, which greatly reduces the workload of the experimenters, while avoiding the adverse effects caused by human intervention, and ensuring the reliability of the experiment.

[0048] The present application proposes an intelligent sorting method and system for reactor thermal hydraulic test conditions, which uses an intelligent algorithm to automatically sort the conditions, so that the experiment can be completed in the shortest time, thereby achieving the purpose of improving the experiment efficiency and providing a prerequisite for the intelligent operation of thermal hydraulic experiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings described herein are used to provide a further understanding of the embodiments of the present application, constitute a part of the present application, and do not constitute a limitation on the embodiments of the present application. In the drawings:

[0050] Figure 1 A flow chart of the intelligent sorting method proposed in the embodiment of the present application;

[0051] Figure 2 An example of a directed graph of operating mode switching time generated for an embodiment of the present application;

[0052] Figure 3 This is a flow chart of the ant colony algorithm used in the embodiment of the present application;

[0053] Figure 4 This is a functional block diagram of the intelligent sorting system proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with examples and drawings. The illustrative implementation scheme of the present application and its description are only used to explain the present application and are not intended to limit the present application.

[0055] Example:

[0056] The nuclear reactor thermal hydraulic system experiment needs to carry out multiple groups of working conditions. For example, in order to determine the friction resistance coefficient and heat transfer coefficient of the core channel, it is necessary to carry out experiments of many working conditions. The friction resistance coefficient and heat transfer coefficient correlation of the core channel are determined by the data of many working conditions. The working condition conversion takes a lot of time. At present, the working condition conversion is usually mainly realized manually, and the working condition sequence is usually determined based on experience or convenient operation, which leads to problems such as long experiment time, low efficiency, and high work intensity of experimenters. In view of this, the present embodiment provides an intelligent sorting method for reactor thermal hydraulic experimental conditions. The intelligent sorting method provided in the present embodiment adopts an intelligent optimization sorting algorithm. The work is optimally sorted by an intelligent algorithm without human intervention, which effectively improves the experimental efficiency, reduces the experimental cost, and reduces the work intensity of the experimenters, providing a prerequisite for the intelligent operation of thermal hydraulic experiments.

[0057] like Figure 1 As shown, the intelligent sorting method provided in this embodiment specifically includes:

[0058] Step 100, based on the thermal hydraulic test condition data, combined with a pre-built machine learning model, predict the switching time between various conditions of the thermal hydraulic test.

[0059] The process of building a machine learning model specifically includes:

[0060] The historical operating condition data of the thermal hydraulic experiment and the thermal circuit simulation data are obtained, including but not limited to: the combination of the circuit pressure, circuit flow, the inlet temperature of the experimental body and the outlet temperature of the experimental body, and the historical operating condition switching time, and the historical operating condition switching time and the corresponding historical operating condition data changes are obtained.

[0061] Table 1 shows a typical thermal hydraulic test condition.

[0062] Table 1 Typical thermal hydraulic test conditions

[0063]

[0064] As shown in Table 1, this typical thermal hydraulic experiment includes a total of n working conditions. The working condition data corresponding to the working condition include loop pressure, loop flow, experimental body inlet temperature, and experimental body outlet temperature. There are differences in the working condition data between each working condition. For example, the working condition data corresponding to working condition 1 is {P1, Q1, T in1 ,T out1}, the corresponding working condition data of working condition 2 is {P1,Q1,T in1 ,T out2}, the switch from working condition 1 to working condition 2 changes the outlet temperature of the experimental body. It can be seen that the switching time is related to the changes in these working condition data. Therefore, this embodiment uses the historical working condition switching time and its corresponding historical working condition data changes to obtain the relationship function between the switching time and the changes in each working condition data through fitting regression. According to this relationship function, the switching time between each working condition to be tested can be directly predicted. Optionally, this embodiment can use the least squares fitting regression method, and the curve fitting regression function is:

[0065] y=f(x1)+f(x2)+…+f(x p ) (1)

[0066] Among them, the change of switching time caused by the change of each working condition data is shown in the following formula:

[0067]

[0068] Where y represents the switching time, x1, x2, ..., x p Respectively represent the changes in p working condition data, a0, a1, ..., a q Represents the coefficient.

[0069] Substitute the historical operating condition data changes corresponding to several historical operating condition switching times into the above fitting regression function to predict the operating condition switching time, and use the processed several historical operating condition switching times as the measured values. Then use the least squares fitting method (that is, the sum of the squares of the difference between the predicted value and the measured value is the smallest) to obtain the above coefficients, thereby obtaining the above functional relationship. And introduce the safety boundary constraint condition into the model.

[0070] Based on the above functional relationship and constraints, the switching time between various experimental operating points is calculated according to the data of the experimental operating point.

[0071] In another optional implementation, a thermal hydraulic model of the object under study may be constructed to conduct a system characteristic study, obtain a transfer function model of the switching time and each working condition, and introduce various safety boundary constraints into the model;

[0072] The switching time between any two working conditions can be quickly obtained through the transfer function model.

[0073] Optionally, the above mentioned safety boundary constraints specifically include:

[0074] Constraints on the temperature increase rate, temperature decrease rate, flow rate change rate and pressure change rate. For example, for a typical reactor thermal hydraulic device, the temperature increase rate is set to ≤1℃ / min, while the temperature decrease rate is set to ≤-0.8℃ / min. The flow rate change rate is set to ≤±10% / min. The pressure change rate is set to ≤0.1MPa / min.

[0075] Step 200, generating a directed graph of the switching time of the operating point according to the operating conditions of the thermal hydraulic experiment and the switching time between the operating conditions.

[0076] In step 200, after the switching time between all the working points of the thermal hydraulic experiment is obtained, each working point is used as a node, and the working point switching time is used as the edge between nodes to ensure the time distance between nodes, thereby generating a directed graph of the working point switching time of the thermal hydraulic experiment, such as Figure 2 As shown, Figure 2 The switching time directed graph of three operating points (i.e., operating point 1, operating point 2, and operating point 3) is given. It can be seen from the figure that the switching from operating point 1 to operating point 2, the switching from operating point 2 to operating point 3, the switching from operating point 3 to operating point 1, the switching from operating point 1 to operating point 3, the switching from operating point 3 to operating point 2, and the switching from operating point 2 to operating point 1. It should be noted that Figure 2 This is only an exemplary description and does not limit the number of operating points.

[0077] Step 300, using an ant colony algorithm to perform a shortest path traversal search on a directed graph of operating point switching time, and obtain a ranking of experimental operating points with the shortest total switching time.

[0078] The principle of the shortest path traversal by the ant colony algorithm is:

[0079] Assume there are n operating points, i and j are any two points, d ij represents the distance between two points (i.e. the switching time between the two points), b i (t)(i=1,…,n) represents the number of ants at point i at time t, then Represents the total number of ants, using τ ij (t) is used to simulate the pheromone concentration on path i to j at time t. The pheromone concentration on each path will evaporate and increase accordingly as the ants pass by, so the update formula for the pheromone concentration is:

[0080] τ ij (t+1)=(1-ρ)·τ ij (t)+Δτ ij (3)

[0081] In the formula, τ ij(t+1) represents the pheromone concentration on the path from i to j at time t+1, τ ij (t) represents the pheromone concentration on path i to j at time t, ρ is a number between 0 and 1, and (1-ρ) is the volatility factor. In addition, Δτ ij It represents the total amount of pheromone left by ants on all paths from i to j after traversing all working points once, that is: In the formula, represents the amount of pheromone left by the kth ant on the path from i to j, and m represents the number of ants. If the kth ant passes through the path from i to j, then: L k is the total length of the path that the ants have walked, and Q is the pheromone constant, which represents the total amount of pheromone released by the ants in one cycle.

[0082] Introduced a heuristic factor: η ij =1 / d ij , represents the expected degree of ants moving from city i to city j. By combining the pheromone concentration and the heuristic factor, we can get the probability that the ant chooses path i to j at time t. for:

[0083]

[0084] In the formula, α and β are adjustment factors used to adjust τ ij (t) and η ij In addition, allowed k is a set that represents the paths that ant k (k = 1, 2, 3, ... m) has not yet walked (using the taboo table to store the paths that have been walked). This storage can ensure that all solutions are logically feasible. If the information density on the path from i to j is greater, τ ij The larger the value of (t), the greater the probability that the path is selected; similarly, if the path length (operating state switching time) is shorter, then η ij =1 / d ij The larger it is, the greater the probability that the path will be selected.

[0085] Preferably, in this embodiment, ρ is set to 0.16-0.26 for better calculation effect; α=2, β=3-5 for better calculation effect in experiments; m is equal to the number of working conditions. If the number of working conditions is less than 9, m=9; experiments show that the above values ​​have the best convergence calculation effect.

[0086] like Figure 3As shown in FIG. 1 , the ant colony algorithm process used in this embodiment mainly includes: initializing parameters; constructing a solution space; updating pheromones; judging whether the number of iterations has been reached, if so, outputting the optimal solution and ending the process; otherwise, recording the number of iterations + 1, clearing the path record table and returning to the step of constructing the solution space. The initialization parameter is to initialize the same amount of pheromones for all paths. The ants select the next path to take based on the pheromone concentration and heuristic information on the path. After the ants complete the path, the solution space is constructed, and then the pheromones are updated according to the path length. The iterative process is to repeat the path selection and pheromone update steps until the termination condition is met. This embodiment can adopt three optimization strategies: positive feedback information mechanism, pheromone update, and ant screening of accessible paths, so that its computing power and operation efficiency are high, and the heuristic probability search is used, which is not easy to fall into the local optimum. It should be noted that the pheromone concentration calculation method here is shown in formula (3), which is not a direct quantity, but an indirect quantity in the ant colony algorithm process. The higher the pheromone concentration, the higher the probability that the ant will choose the path. When selecting a path (operating state switching time), as more and more ants pass through a certain path, the number of pheromone tracks they leave behind also increases, indicating that the probability of ants choosing this path later is higher, thereby increasing the pheromone intensity of the path.

[0087] Step 400, sorting the experimental operating points with the shortest total switching time, and performing corresponding operating point switching control.

[0088] Through the shortest path search in step 300 above, a path with the shortest switching time is determined from the directed graph of the operating point switching time, that is, the optimal experimental operating point ranking. According to the optimal experimental operating point ranking, the corresponding operating point intelligent switching control is performed.

[0089] Compared with the existing working point sorting determined by manual sorting, experience sorting, etc., the intelligent sorting method proposed in this embodiment first uses a pre-built machine learning model to realize the automatic prediction of the switching time between each working condition of the thermal hydraulic experiment; then each working point is used as a node, and the switching time between each working condition is used as the edge between the nodes, that is, the length of the edge represents the time distance between the nodes, the longer the switching time, the longer the length of the edge, anyway, the smaller the length of the edge, and a directed graph of the working condition switching time is generated; finally, the ant colony algorithm is used to search the shortest path of the directed graph of the working condition switching time to obtain the path with the shortest total switching time, that is, the working condition sorting with the shortest completion time, and the working condition control is realized according to the working condition sorting with the shortest completion time, which can realize the shortest completion time of the thermal hydraulic experiment, effectively improve the experimental efficiency, and the whole process is fully automatic, which greatly reduces the work intensity of the experimenters, while avoiding the adverse effects caused by human intervention, and ensuring the reliability of the experiment. In addition, the intelligent sorting method proposed in this embodiment uses an intelligent algorithm to automatically complete the working condition sorting, providing a prerequisite for the intelligent operation of the thermal hydraulic experiment.

[0090] Based on the same technical concept as above, this embodiment also proposes an intelligent sorting system for reactor thermal hydraulic test conditions, such as Figure 4 As shown, the intelligent sorting system proposed in this embodiment specifically includes:

[0091] The calculation module predicts the switching time between various working conditions of thermal hydraulics based on the thermal hydraulic experimental working condition data combined with the pre-built machine learning model.

[0092] A directed graph construction module generates a directed graph of the switching time of the operating point according to the operating conditions and switching time of the thermal hydraulic experiment. The directed graph uses each operating point as a node and the switching time as an edge between nodes to represent the time distance between nodes.

[0093] The optimization module uses the ant colony algorithm to traverse the directed graph information of the switching time of the operating point in the shortest path to search and obtain the ranking of the experimental operating points with the shortest total switching time.

[0094] And, a control module, which sorts the experimental working conditions with the shortest total switching time and performs corresponding working condition point switching control.

[0095] Furthermore, the calculation module also includes:

[0096] A data processing unit is used to obtain historical working condition data and historical working condition switching time of thermal hydraulic experiments and obtain each historical working condition switching time and its corresponding change amount of each historical working condition data.

[0097] The fitting regression unit uses the switching time of each historical working condition and the corresponding change amount of each historical working condition data to obtain the relationship function between the switching time and the change amount of each working condition data by fitting regression, and uses various safety boundaries as constraint conditions.

[0098] And, a prediction unit, after obtaining the above relationship function, inputs the data change of the experimental operating point into the above relationship function, and can calculate the switching time between each experimental operating point.

[0099] Furthermore, the fitting regression unit is configured to perform the following process:

[0100] Construct multivariate polynomial fitting regression function;

[0101] Substituting the historical operating condition data changes corresponding to the processed several historical operating condition switching times into the multivariate polynomial fitting regression function to predict the operating condition switching time; taking the processed several historical operating condition switching times as measured values;

[0102] Based on the predicted values ​​and the measured values, the coefficients of the above-mentioned multivariate polynomial fitting regression function are fitted by the least squares method, that is, a set of coefficients that minimize the sum of the squares of the differences between the predicted values ​​and the measured values ​​are obtained. Substituting the obtained coefficient values ​​into the above-mentioned multivariate polynomial fitting regression function can obtain the relationship function between the operating condition switching time and the change in the data of each operating condition.

[0103] Furthermore, the optimization module is configured to perform the following process:

[0104] Initialization parameters;

[0105] Construct solution space;

[0106] Update pheromones;

[0107] Determine whether the number of iterations has been reached. If so, output the optimal solution and end the process; otherwise, record the number of iterations + 1, clear the path record table, and return to the step of constructing the solution space.

[0108] The initialization parameter is to initialize the same amount of pheromone for all paths. The ant selects the next path to take based on the pheromone concentration and heuristic information on the path. After the ant completes the path, the solution space is constructed, and the pheromone is updated based on the path length. The iterative process repeats the path selection and pheromone update steps until the termination condition is met.

[0109] It should be noted that the pheromone described here does not directly represent the pheromone concentration, but is the probability of ant choosing a path constructed based on the pheromone concentration and the heuristic factor, as described in step 300 of the above method, which will not be described in detail here.

[0110] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0111] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0112] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0113] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0114] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only the specific implementation method of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. An intelligent sorting method for reactor thermal hydraulic test conditions, characterized in that: The intelligent sorting method comprises: Based on the thermal hydraulic test condition data, combined with the pre-built machine learning model and various safety boundary constraints, the switching time between various conditions of the thermal hydraulic test is predicted; According to the working conditions of the thermal hydraulic experiment and the switching time between the working conditions, a directed graph of the working condition point switching time is generated; the directed graph of the working condition point switching time uses each working condition point as a node and the switching time as an edge between nodes to represent the time distance between nodes; An ant colony algorithm is used to perform a shortest path traversal search on the directed graph of the switching time of the operating point, and a ranking of the experimental operating points with the shortest total switching time is obtained; The experimental operating points with the shortest total switching time are sorted and corresponding operating point switching control is performed.

2. The intelligent sorting method for reactor thermal hydraulic test conditions according to claim 1, characterized in that: The machine learning model construction process specifically includes: Obtain historical working condition data and historical working condition switching time of thermal hydraulic experiments, and process to obtain each historical working condition switching time and its corresponding change amount of each historical working condition data; By using the processed historical working condition switching time and the corresponding historical working condition data changes, fitting regression is performed to obtain the relationship function between the working condition switching time and the data changes of each working condition, and various safety boundaries are used as constraint conditions; The data variation of the working condition point to be tested is input into the relationship function to predict the switching time between the working condition points.

3. The intelligent sorting method for reactor thermal hydraulic test conditions according to claim 2, characterized in that: The relationship function between the switching time of each historical working condition obtained by processing and the change amount of each historical working condition data corresponding to the switching time of each historical working condition obtained by fitting regression specifically includes: Construct multivariate polynomial fitting regression function; Substituting the historical operating condition data changes corresponding to the processed several historical operating condition switching times into the multivariate polynomial fitting regression function to predict the operating condition switching time; taking the processed several historical operating condition switching times as measured values; Based on the predicted value and the measured value, the coefficients of the multivariate polynomial regression function are obtained by least squares fitting. The obtained coefficient values ​​are substituted into the multivariate polynomial fitting regression function to obtain the relationship function between the operating condition switching time and the change amount of each operating condition data.

4. The intelligent sorting method for reactor thermal hydraulic test conditions according to claim 1, characterized in that: The machine learning model construction process specifically includes: By constructing a thermal hydraulic model of the research object, the system characteristics are studied, the transfer function model of the switching time and each working condition is obtained, and various safety boundary constraints are introduced into the model; The switching time between any two operating points can be quickly obtained through the transfer function model.

5. According to the intelligent sorting method for reactor thermal hydraulic test conditions according to any one of claims 1 to 4, the constraint conditions of the safety boundary specifically include: Constraints on the rate of temperature increase, rate of temperature decrease, rate of flow change, and rate of pressure change.

6. An intelligent sorting method for reactor thermal hydraulic test conditions according to any one of claims 1 to 4, characterized in that: The ant colony algorithm is used to perform the shortest path traversal search on the directed graph of the operating point switching time to obtain the experimental operating point ranking with the shortest total switching time, which specifically includes: Initialize parameters and initialize the same amount of pheromone for all paths; The solution space is constructed, and the ants choose the next path to take based on the pheromone concentration and heuristic information on the path; Update pheromones; determine whether the number of iterations has been reached. If so, output the optimal solution and end the process; otherwise, record the number of iterations + 1, clear the path record table, and return to the step of constructing the solution space; The pheromone is a probability of an ant choosing a path constructed based on the pheromone concentration and a heuristic factor; the heuristic factor is the inverse of the path length, and the pheromone concentration increases accordingly as the ants pass by.

7. The intelligent sorting method for reactor thermal hydraulic test conditions according to claim 6, characterized in that: The probability of the ant choosing a path is expressed as: In the formula, α and β are regulatory factors used to adjust the pheromone concentration τ ij (t) and the heuristic factor η ij The role between; allowed k Indicates the path that ant k has not yet walked.

8. The intelligent sorting method for reactor thermal hydraulic test conditions according to claim 7, characterized in that: The updating formula of the pheromone concentration is: t ij (t+1)=(1-ρ)·τ ij (t)+Δτ ij Where ρ is a number between 0 and 1, (1-ρ) is the volatility factor, and Δτ ij represents the total amount of pheromone left by ants on all paths from i to j after traversing all working points once, τ ij (t) represents the pheromone concentration on path i to j at time t, τ ij (t+1) represents the pheromone concentration on the path from i to j at time t+1, and m represents the number of ants in the algorithm.

9. An intelligent sequencing system for reactor thermal hydraulic test conditions, characterized in that: The intelligent sorting system comprises: A calculation module, wherein the calculation module predicts the switching time between various working conditions of the thermal hydraulic experiment based on the working condition data of the thermal hydraulic experiment in combination with a pre-built machine learning model and various safety boundary constraints; A directed graph construction module, wherein the directed graph construction module generates a directed graph of the switching time of the working condition points according to the working conditions of the thermal hydraulic experiment and the switching time between the working conditions; the directed graph of the switching time of the working condition points uses each working condition point as a node and the switching time as an edge between the nodes to represent the time distance between the nodes; An optimization module, wherein the optimization module uses an ant colony algorithm to perform a shortest path traversal search on the directed graph of the operating point switching time to obtain a ranking of experimental operating points with the shortest total switching time; And, a control module, the control module sorts the experimental operating points with the shortest total switching time and performs corresponding operating point switching control.

10. The intelligent sequencing system for reactor thermal hydraulic test conditions according to claim 9, characterized in that: The calculation module further includes: A data processing unit, the data processing unit is used to obtain historical working condition data and historical working condition switching time of thermal hydraulic experiments, and process to obtain each historical working condition switching time and its corresponding change amount of each historical working condition data; A fitting regression unit, wherein the fitting regression unit uses the switching time of each historical working condition and the corresponding change amount of each historical working condition data to obtain a relationship function between the switching time and the change amount of each historical working condition switching data by fitting regression, and uses various safety boundaries as constraint conditions; And, a prediction unit, wherein the prediction unit inputs the data change of the experimental operating point into the relationship function to calculate the switching time between each experimental operating point.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images