Multi-model process sensitivity analysis method, storage medium and equipment
By defining a multi-model process sensitivity index and using a novel Monte Carlo algorithm to eliminate nested structures, the problem of high computational cost of the multi-model process sensitivity index is solved, and fast and accurate multi-model process sensitivity analysis is achieved.
Patent Information
- Application Number
- CN202410978897.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing multi-model process sensitivity index calculations are too costly, and traditional Monte Carlo algorithms are time-consuming and difficult to apply to complex groundwater models.
A multi-model process sensitivity analysis method is proposed. By defining a multi-model process sensitivity index, the Monte Carlo algorithm is used to calculate and eliminate the nested structure of parameter sampling, and a novel Monte Carlo algorithm is used to calculate the process sensitivity index.
It significantly improves the computational efficiency of the multi-model process sensitivity index, saves costs, and takes less time, enabling rapid and accurate calculation of the multi-model process sensitivity index.
Smart Images

Figure CN118940489A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of sensitivity analysis, and in particular to a multi-model process sensitivity analysis method, storage medium, and equipment. Background Art
[0002] Numerical simulation methods represented by reactive solute transport models are powerful tools for simulating groundwater flow, solute transport, and biogeochemical reactions. With the deepening of groundwater-related research, groundwater models are becoming increasingly large and complex, and the running time cost of the models is also increasing. Without the help of statistical analysis methods, it is often very difficult to understand and improve complex groundwater models. At the same time, due to various reasons such as the lack of relevant data and researchers' insufficient understanding of groundwater processes, there are uncertainties in parameters and model structures in groundwater models. These uncertainties will lead to uncertainty in model output and have a negative impact on the prediction accuracy of the model. Sensitivity analysis is an effective tool for identifying important sources of uncertainty and improving models. Sensitivity analysis of complex groundwater models can identify important uncertainty factors in the model, guide researchers to focus on important uncertainty factors in model improvement, data collection, and parameter calibration, and provide technical support for reasonably simplifying model structures, guiding data collection for field work, and improving model performance.
[0003] In recent years, model uncertainty has received increasing attention in the field of sensitivity analysis. It is considered to be the largest source of uncertainty in the model and will seriously affect the accuracy of the model output. Among the many sensitivity analysis methods, the global sensitivity analysis method, especially the variance-based method, is considered to be one of the most suitable and comprehensive sensitivity analysis methods for complex groundwater models. However, this method relies on Monte Carlo (MC) simulation and requires multiple calls to the model for a large number of calculations. The operation of complex groundwater models is usually very time-consuming, which makes the variance-based global sensitivity analysis method usually difficult to apply. Based on the variance-based global sensitivity analysis method, some scholars have proposed a new process sensitivity index for identifying important system processes under process model and parameter uncertainty. The process sensitivity index is based on the model average and global sensitivity analysis methods, and explicitly decomposes the system model into a process model. It is used to analyze the uncertainty of the process model and model parameters, and can identify important processes while considering the parameter uncertainty and model uncertainty in the groundwater model. However, this method also has the problem of high computational cost, which limits its application. Summary of the invention
[0004] The purpose of the present invention is to solve the problem that the calculation cost of the existing multi-model process sensitivity index is too high, and to propose a multi-model process sensitivity analysis method, comprising the following steps:
[0005] S1. Obtain the hydrological system driven by multiple processes and identify the relevant parameters of the processes;
[0006] S2, define multi-model process sensitivity index based on relevant parameters of hydrological systems and processes driven by multiple processes;
[0007] S3. According to the definition of multi-model process sensitivity index, the Monte Carlo algorithm is used to calculate and obtain the calculation formula containing parameter sampling nested structure;
[0008] S4. According to the calculation formula containing the nested structure of parameter sampling, the nested structure of parameter sampling is eliminated through three groups of parameter samples, and the multi-model process sensitivity analysis results of the hydrological system are obtained.
[0009] Furthermore, each process in the hydrological system driven by multiple processes includes multiple process models.
[0010] Further, the process sensitivity index of process A is expressed as:
[0011]
[0012] Among them, PS A represents the process sensitivity index of process A, Δ represents the system output, V(Δ) represents the total variance of the system output, M Ai Represents the model set M A The i-th process model in M A Represents the model set of process A, M ~A represents the model set of all processes except process A, Indicates that the system output Δ is in M Ai Under the condition of ~A Expected value; Represents the model set M A All models in Then find the variance of the value.
[0013] Furthermore, the process sensitivity analysis is carried out according to the process sensitivity index as follows:
[0014] Will Rewritten as:
[0015]
[0016] Among them, M B Replace M ~A ,θ A For the model set MA The corresponding parameter set is the model set M B The corresponding parameter set, θ Ai and θ Bi The parameter set θ is A and θ B The i-th group of parameters in [Δ|θ Ai ,M Ai ,θ Bi ,M Bi ] indicates that the process model of the system in process A and process B takes M Ai and M Bi , the corresponding parameter values are θ Ai and θ Bi The output Δ under the condition of M Bi Represents the model set M B The i-th process model in represents [Δ|θ Ai ,M Ai ,θ Bi ,M Bi About process model M Bi The corresponding parameter set θ B The expected value of express About Model Set M B The expected value of express About process model M Ai The corresponding parameter set θ A The expected value of express About the process model set M A The expected value of express About process model M Ai The corresponding parameter set θ A The expected value of express About the process model set M A expected value.
[0017] Furthermore, the nested structure of parameter sampling includes four loop structures. The first loop structure is used for the model set M A ,right and Calculate; the second loop structure is used for the model set M A The corresponding parameter set θ A ,right and Calculate; the third loop structure is used for the model set M B ,right and Calculate; the fourth loop structure is used for the model set M B The corresponding parameter set θ B ,right Perform calculations.
[0018] Further, Rewritten as:
[0019]
[0020] in, Represents process models The parameter set, and It is the model set M B The jth and lth process models in The process models representing the system in process A and process B are M Ai and M Bj , the corresponding parameter values are respectively Ai and θ Bj The system output Δ under the condition of [Δ|θ A ,M A ,θ B ,M B ] indicates that the model set of the system in process A and process B is M A and M B , the corresponding parameter sets are respectively A and θ B The system output Δ under the condition of represents [Δ|θ A ,M A ,θ B ,M B About Model Set M B The corresponding parameter set θ B The expected value of express About Model Set M B The expected value of express About Model Set M A The corresponding parameter set θ A The expected value of express About Model Set M A The expected value of Indicates that the model set and parameter set of the system in process A are M A and θ A , the process model and parameter values of process B are M Bj and θ Bj The system output Δ under the condition of express About process model MNj The corresponding parameter θ Bj The expected value of express About the process model Corresponding parameters The expected value of The process models representing the system in process A and process B are M Ai and M Bl , the corresponding parameter values are respectively Ai and θ Bl The system output Δ under the condition of express About the process model Corresponding parameters The expected value of express About the process model Corresponding parameters The expected value of The random variables in the brackets are related to the process model M Ai The corresponding parameter set The expected value in The random variables in the brackets are related to the model set M A The corresponding parameter set θ A The expected value of Represents the model set M B Process Model The associated probability, Represents the model set M B Process Model The associated probability, Represents the model set M A Process Model The probability of correlation, n A Represents the number of process models for process A, n B represents the number of process models for process B;
[0021] Expands to:
[0022]
[0023] in, Process Model Conditional parameter The probability density function of
[0024] It is expressed as:
[0025]
[0026] Among them, the parameter set It is a process model Two different parameter sets obtained by repeated sampling, and Represents parameters and by is the probability density function of the condition;
[0027] Substituting formula (5) into formula (4), we can get:
[0028]
[0029] in, Represents the parameter set Integral in the range of It means that under the condition that the system model consists of the i-th model of process A and the j-th model of process B, the corresponding parameter value is and The model output Δ when The process model representing the system in process A and process B is M Ai and M Bj , the corresponding parameter values are respectively Ai and The system output Δ under the condition of
[0030] Formula (6) is expressed as:
[0031]
[0032] The integral in formula (7) corresponds to the expected value Formula (7) can be rewritten as:
[0033]
[0034] In formula (8), the nested expectation Transformed into a single expectation Through three sets of parameter samples Successfully eliminated The nested structure of
[0035] It is expressed as:
[0036]
[0037] in, Nested Expectations Transformed into a single expectation
[0038] Rewrite formula (7) as:
[0039]
[0040] Rewrite formula (8) as:
[0041]
[0042] Where N is the number of parameter samples, represents the rth θ sampled Ai , Indicates the rth sample Indicates the rth sample
[0043] Will Rewritten as:
[0044]
[0045] Rewrite the following formula as:
[0046]
[0047]
[0048] The sensitivity index Calculated by a new Monte Carlo algorithm:
[0049]
[0050] The present invention also proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-model process sensitivity analysis method is implemented.
[0051] The present invention also proposes an electronic device, comprising a processor and a memory, wherein the processor and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises computer-readable instructions, and the processor is configured to call the computer-readable instructions to execute the above-mentioned multi-model process sensitivity analysis method.
[0052] The beneficial effects brought by the technical solution provided by the present invention are:
[0053] The multi-model process sensitivity analysis method of the present invention is based on the traditional Monte Carlo algorithm and has undergone rigorous theoretical derivation. It eliminates the nested structure of parameter sampling, greatly reduces the number of simulations required by the traditional Monte Carlo algorithm, and significantly improves the calculation efficiency of the multi-model process sensitivity index. It can quickly and accurately calculate the multi-model process sensitivity index, save costs, and save time. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Figure 1is a flow chart of a multi-model process sensitivity analysis method according to an embodiment of the present invention;
[0055] Figure 2 is a graph showing the sensitivity of the model output of an embodiment of the present invention to process 1;
[0056] Figure 3 It is a block diagram of an electronic device in an exemplary embodiment of the present invention. DETAILED DESCRIPTION
[0057] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.
[0058] The flowchart of the multi-model process sensitivity analysis method of the embodiment of the present invention is as follows: Figure 1 , specifically including the following steps:
[0059] S1. Obtain a hydrological system driven by multiple processes and identify relevant parameters of the processes.
[0060] A system driven by multiple processes is represented as: M = {M A (θ A ),M B (θ B ),...}, each process in the system includes multiple optional process models, among which: M represents a system driven by multiple processes. A (θ A ),M B (θ B ),... is the model set of processes A, B,... in a system driven by multiple processes, Optional process model for process A, Process Model The corresponding parameter set, Process Model The corresponding parameter set, θ A For the model set M A The corresponding parameter set, θ B For the model set M B The corresponding parameter set.
[0061] Specifically, the present invention selects a hydrological or groundwater model (system) to be subjected to multi-model process sensitivity analysis. The model is a function file based on MATLAB software, and the model is defined as a function of output=model(process_models, param_values), so that the computing software can call the model for simulation according to the specific process model number and parameter value, and return the model output data. Among them, process_models is a row vector with a length equal to the number of processes (i.e., equal to 3), and the values in the vector represent the process model numbers of process 1, process 2, and process 3 from left to right. param_values is a row vector with a length equal to the number of parameters, and the values in the vector represent the values of parameter 1, parameter 2, and parameter 3 from left to right. Output is the model output of interest to the user, which is determined by the user and adjusted to a one-dimensional row vector. The model file name should be "model".
[0062] S2. Calculate the process sensitivity index of the process based on multiple process-driven systems and process-related parameters.
[0063] The process sensitivity index of process A is expressed as:
[0064]
[0065] Among them, PS A represents the process sensitivity index of process A, Δ represents the system output, V(Δ) represents the total variance of the system output, M Ai Represents the model set M A The i-th process model in M A Represents the model set of process A, M ~A represents the model set of all processes except process A. The formula involves calculating the conditional expectation, which is the expected value under given conditions. The conditional expectation is similar to the ordinary expectation, except that the conditional expectation requires averaging the random variables under given conditions. Indicates that the system output Δ is in M Ai Under the condition of ~A Expected value; Represents the model set M A All available models in Then find the variance of the value.
[0066] Determine the key processes, process models, content and corresponding quantities of uncertain parameters in the hydrological or groundwater model, and write the parameter attribute tables required by the program.
[0067] The function of the parameter attribute table is to input the hierarchical relationship and value range of the parameters into the software. The parameter attribute table is established using a text file. The content of the parameter attribute table is shown in Table 1. The first line is the title line, and each line beyond that represents a parameter. Each parameter needs to give its serial number, distribution, value 1, value 2, as well as information such as the process and process model to which it belongs. Among them, the serial number, process, and process model are all numbered from 1 according to the rule of 1, 2, 3, 4... The parameters follow the normal distribution or uniform distribution, respectively, and are represented by N and U. The values 1 and 2 represent the mean and variance of the parameter value range under the normal distribution, or the lower and upper limits of the parameter value range under the uniform distribution. The specific settings of the parameter attribute table can be referred to Table 1. The file of the parameter attribute table established using a text file is named "param.txt".
[0068] Table 1
[0069] Serial number distributed Value 1 Value 2 process Process Model 1 N 16.88 1 2 1 2 U 0.1 0.2 2 2 3 N 15 1 1 1 4 N 20 1 1 2 5 N 10 1 1 2 6 N 3.5 0.75 3 1 7 N 2.5 0.3 3 2 8 N 0.3 0.05 3 2
[0070] Determine the key process of the sensitivity index to be output as input. In the embodiment of the present invention, process 1 is selected as input and the process is coded as 1 in "param.txt" and "model.m".
[0071] S3. Perform process sensitivity analysis based on the process sensitivity index and use the Monte Carlo algorithm to calculate and obtain the nested structure of parameter sampling.
[0072] The specific process sensitivity analysis based on the process sensitivity index is as follows:
[0073] Will Rewritten as:
[0074]
[0075] Among them, M B Replace M ~A ,θ A For the model set M A The corresponding parameter set, θ B For the model set M B The corresponding parameter set, θ Ai and θ Bi The parameter set θ is A and θ B The i-th group of parameters in [Δ|θ Ai ,M Ai ,θ Bi ,M Bi ] indicates that the process model of the system in process A and process B takes M Ai and M Bi , the corresponding parameter values are θ Ai and θ BiThe output Δ under the condition of M Bi Represents the model set M B The i-th process model in Indicates that the parameters in brackets are related to the process model M Bi The corresponding parameter set θ B The expected value of represents [Δ|θ Ai ,M Ai ,θ Bi ,M Bi About process model M Bi The corresponding parameter set θ B The expected value of Indicates that the parameters in brackets are about the model set M B The expected value of express About Model Set M B The expected value of Indicates that the parameters in brackets are related to the process model M Ai The corresponding parameter set θ A The expected value of express About process model M Ai The corresponding parameter set θ A The expected value, E MA () indicates that the parameters in brackets are related to the process model set M A The expected value of express About the process model set M A The expected value of express About process model M Ai The corresponding parameter set θ A The expected value of express About the process model set M A expected value.
[0076] The nested structure for obtaining parameter sampling includes four loop structures. The first loop structure is used for the model set M A ,right and Calculate; the second loop structure is used for the model set M A The corresponding parameter set θ A ,right and Calculate; the third loop structure is used for the model set M B ,right and Calculate; the fourth loop structure is used for the model set M B The corresponding parameter set θ B,right Perform calculations.
[0077] Will Rewritten as:
[0078]
[0079] in, Represents process models The parameter set, and It is the model set M B The jth and lth process models in The process model representing the system in process A and process B is M Ai and M Bj , the corresponding parameter values are respectively Ai and θ Bj The system output Δ under the condition of [Δ|θ A ,M A ,θ B ,M B ] indicates that the model set of the system in process A and process B is M A and M B , the corresponding parameter sets are respectively A and θ B The system output Δ under the condition of Indicates that the model set and parameter set of the system in process A are M A and θ A , the process model and parameter values of process B are M Bj and θ Bj The system output Δ under the condition of The process model representing the system in process A and process B is M Ai and M Bl , the corresponding parameter values are respectively Ai and θ Bl The system output Δ under the condition of Bj Represents the model set M B The j-th process model in The variables in brackets represent the process model M. Bj The corresponding parameter set θ Bj The expected value of Indicates that the variables in brackets are related to the process model M Ai The corresponding parameter set θ Ai The expected value in The random variables in the brackets are related to the model set M A The corresponding parameter set θ A The expected value of represents [Δ|θ A ,MA ,θ B ,M B About Model Set M B The corresponding parameter set θ B The expected value of Indicates that the parameters in brackets are about the model set M B The expected value of Indicates that the parameters in brackets are about the model set M A The expected value of Indicates that the parameters in brackets are related to the process model Corresponding parameters The expected value of Represents the model set M B Process Model The associated probability, Represents the model set M B Process Model The associated probability, Represents the model set M A Process Model The probability of correlation, n A Represents the number of process models for process A, n B Represents the number of process models for process B.
[0080] Expands to:
[0081]
[0082] in, Process Model Conditional parameter The probability density function of
[0083] It is expressed as:
[0084]
[0085] Among them, the parameter set It is a process model Two different parameter sets obtained by repeated sampling, and Represents parameters and by is the probability density function of the condition;
[0086] Substituting formula (5) into formula (4), we can get:
[0087]
[0088] in, Represents the parameter set Integral in the range of It means that under the condition that the system model consists of the i-th model of process A and the j-th model of process B, the corresponding parameter value is and The model output Δ when The process models representing the system in process A and process B are M Ai and M Bj , the corresponding parameter values are respectively Ai and The system output Δ under the condition of
[0089] Formula (6) is expressed as:
[0090]
[0091] The integral in formula (7) corresponds to the expected value Formula (7) can be rewritten as:
[0092]
[0093] In formula (8), the nested expectation Transformed into a single expectation Through three sets of parameter samples Successfully eliminated The nested structure of
[0094] It is expressed as:
[0095]
[0096] in, Nested Expectations Transformed into a single expectation
[0097] Rewrite formula (7) as:
[0098]
[0099] Rewrite formula (8) as:
[0100]
[0101] Where N is the number of parameter samples, represents the rth θ sampled Ai , Indicates the rth sample Indicates the rth sample
[0102] Will Rewritten as:
[0103]
[0104] Rewrite the following formula as:
[0105]
[0106] The sensitivity index Calculated by a new Monte Carlo algorithm:
[0107]
[0108] S4. According to the nested structure of parameter sampling, relevant parameters are sampled, and the nested structure of parameter sampling is eliminated through parameter samples to obtain the multi-model process sensitivity analysis results of the hydrological system.
[0109] A program based on the novel Monte Carlo algorithm is written based on MATLAB software, and the program based on the novel Monte Carlo algorithm (the method of the present invention) is used to perform analysis and calculation of multi-model process sensitivity.
[0110] Create a new M script file, copy and paste the source code into the file, and customize the file name (for example, "MultimodelPSA"). There are three files in the current MATLAB folder: "param.txt", "model.m", and "MultimodelPSA.m". Run the new multi-model process sensitivity index calculation software to obtain the sensitivity index of the model output to process 1.
[0111] Before sensitivity analysis, the user needs to assign values to N and Noutput. The former represents the number of parameter sampling times. The larger the value, the easier it is for the calculated sensitivity index to converge to an accurate result, but the simulation time will also be longer. In the embodiment of the present invention, N=1000; the latter represents the length of the model output vector, which is consistent with the output output of "model.m". After the parameters are assigned, you can run the software directly by clicking the "Run" button under the editor in the menu bar of the MATLAB operation interface.
[0112] Output the sensitivity index result of the selected process. When the new multi-model process sensitivity index calculation software is finished running, the number of model runs and the total time will be displayed in the command line window of the running interface. At the same time, a line graph will pop up on the software interface, such as Figure 2 As shown, it reflects the change in the sensitivity of the model output to process 1.
[0113] After the software is finished running, a data file "SI_A.mat" will be obtained in the folder where the source script file is located, which stores the sensitivity results of the model output to process 1.
[0114] In an exemplary embodiment, a computer-readable storage medium is included, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-model process sensitivity analysis method described above is implemented.
[0115] See also Figure 3 In an exemplary embodiment, an electronic device is also included, including at least one processor, at least one memory, and at least one communication bus.
[0116] Wherein, a computer program is stored in the memory, and the computer program includes computer-readable instructions. The processor calls the computer-readable instructions stored in the memory through the communication bus to execute the above-mentioned multi-model process sensitivity analysis method.
[0117] The method of this application can be used in the sensitivity analysis of complex groundwater models. By screening out the parameters that have the greatest impact on the model operation results through multi-model process sensitivity analysis, the important uncertainty factors in the model can be effectively identified. Compared with the traditional Monte Carlo calculation method, the method of this application has lower calculation cost and shorter time consumption, is more convenient for the application of multi-model process sensitivity analysis, improves the accuracy of groundwater flow numerical simulation, reduces the error caused by the uncertainty of hydrogeological parameters, and can be applied to complex problems with a large number of processes, process models and parameters.
[0118] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A multi-model process sensitivity analysis method, characterized in that: The following steps are involved: S1. Obtain the hydrological system driven by multiple processes and identify the relevant parameters of the processes; S2, define the multi-model process sensitivity index based on the relevant parameters of the hydrological systems and processes driven by multiple processes; S3. According to the definition of multi-model process sensitivity index, the Monte Carlo algorithm is used to calculate and obtain the calculation formula containing parameter sampling nested structure; S4. According to the calculation formula containing the nested structure of parameter sampling, the nested structure of parameter sampling is eliminated through three groups of parameter samples, and the multi-model process sensitivity analysis results of the hydrological system are obtained.
2. A multi-model process sensitivity analysis method according to claim 1, characterized in that: Each process in a hydrological system driven by multiple processes includes multiple process models.
3. A multi-model process sensitivity analysis method according to claim 1, characterized in that: The process sensitivity index of process A is expressed as: Among them, PS A represents the process sensitivity index of process A, Δ represents the system output, V(Δ) represents the total variance of the system output, M Ai Represents the model set M A The i-th process model in M A Represents the model set of process A, M ~A represents the model set of all processes except process A, Indicates that the system output Δ is in M Ai Under the condition of ~A Expected value; Represents the model set M A All models in Then find the variance of the value.
4. A multi-model process sensitivity analysis method according to claim 3, characterized in that: The specific process sensitivity analysis based on the process sensitivity index is as follows: Will Rewritten as: Among them, M B Replace M ~A ,θ A For the model set M A The corresponding parameter set is the model set M B The corresponding parameter set, θ Ai and θ Bi The parameter set θ is A and θ B The i-th group of parameters in [Δ|θ Ai ,M Ai ,θ Bi ,M Bi ] indicates that the process model of the system in process A and process B takes M Ai and M Bi , the corresponding parameter values are θ Ai and θ Bi The output Δ under the condition of M Bi Represents the model set M B The i-th process model in represents [Δ|θ Ai ,M Ai ,θ Bi ,M Bi About process model M Bi The corresponding parameter set θ B The expected value of express About Model Set M B The expected value of express About process model M Ai The corresponding parameter set θ A The expected value of express About the process model set M A The expected value of express About process model M Ai The corresponding parameter set θ A The expected value of express About the process model set M A expected value.
5. A multi-model process sensitivity analysis method according to claim 4, characterized in that: The nested structure for obtaining parameter sampling includes four loop structures. The first loop structure is used for the model set M A ,right and Calculate; the second loop structure is used for the model set M A The corresponding parameter set θ A ,right and Calculate; the third loop structure is used for the model set M B ,right and Calculate; the fourth loop structure is used for the model set M B The corresponding parameter set θ B ,right Perform calculations.
6. A multi-model process sensitivity analysis method according to claim 5, characterized in that: Will Rewritten as: in, Represents process models The parameter set, and It is the model set M B The jth and lth process models in The process models representing the system in process A and process B are M Ai and M Bj , the corresponding parameter values are respectively Ai and θ Bj The system output Δ under the condition of [Δ|θ A ,M A ,θ B ,M B ] indicates that the model set of the system in process A and process B is M A and M B , the corresponding parameter sets are respectively A and θ B The system output Δ under the condition of represents [Δ|θ A ,M A ,θ B ,M B About Model Set M B The corresponding parameter set θ B The expected value of express About Model Set M B The expected value of express About Model Set M A The corresponding parameter set θ A The expected value of express About Model Set M A The expected value of Indicates that the model set and parameter set of the system in process A are M A and θ A , the process model and parameter values of process B are M Bj and θ Bj The system output Δ under the condition of express About process model M Bj The corresponding parameter θ Bj The expected value of express About the process model Corresponding parameters The expected value of The process models representing the system in process A and process B are M Ai and M Bl , the corresponding parameter values are respectively Ai and θ Bl The system output Δ under the condition of express About the process model Corresponding parameters The expected value of express About the process model Corresponding parameters The expected value of The random variables in the brackets are related to the process model M Ai The corresponding parameter set The expected value in The random variables in the brackets are related to the model set M A The corresponding parameter set θ A The expected value of Represents the model set M B Process Model The probability of the correlation, Represents the model set M B Process Model The associated probability, Represents the model set M A Process Model The probability of correlation, n A Represents the number of process models for process A, n B represents the number of process models for process B; Expands to: in, Process Model Conditional parameters The probability density function of It is expressed as: Among them, the parameter set It is a process model Two different parameter sets obtained by repeated sampling, and Represents parameters and by is the probability density function of the condition; Substituting formula (5) into formula (4), we can get: in, Represents the parameter set Integral within the range of It means that under the condition that the system model consists of the i-th model of process A and the j-th model of process B, the corresponding parameter value is and The model output Δ when The process models representing the system in process A and process B are M Ai and M Bj , the corresponding parameter values are respectively Ai and The system output Δ under the condition of Formula (6) is expressed as: in, The integral in formula (7) corresponds to the expected value Formula (7) can be rewritten as: In formula (8), the nested expectation Transformed into a single expectation Through three sets of parameter samples Successfully eliminated The nested structure of It is expressed as: in, Nested Expectations Transformed into a single expectation Rewrite formula (7) as: Rewrite formula (8) as: Where N is the number of parameter samples, represents the rth θ sampled Ai , Indicates the rth sample Indicates the rth sample Will Rewritten as: Rewrite the following formula as: The sensitivity index Calculated by a new Monte Carlo algorithm:
7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.
8. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the processor and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises computer-readable instructions, and the processor is configured to call the computer-readable instructions to execute the method according to any one of claims 1 to 6.
Citation Information
Cited By
Control method for achieving sewage treatment plant effluent standard reaching and greenhouse gas N2O cooperative emission reduction through model prediction
CN121680093A
