Prediction analysis modeling method for dam settlement based on GA-BP neural network

Abstract

The invention provides a predictive analysis modeling method for dam settlement based on a GA-BP neural network. The predictive analysis modeling method comprises the following steps: 1, establishing a dam filling and settlement amount database; 2, establishing a nonlinear mapping relation among a dam structure, filling parameters and settlement amount by applying the BP neural network; 3, BP neural network parameter setting; 4, optimizing an initial weight and a threshold value of the neural network by using a genetic algorithm, and forming a GA-BP (Genetic Algorithm-Back Propagation) network model; and 5, verifying the accuracy of the model by using test data. According to the method, engineering monitoring data can be utilized more comprehensively and fully, a nonlinear mapping model of dam settlement initial parameters and dynamic settlement volume is established, a large amount of theoretical calculation work is replaced by machine learning, on the other hand, the dam settlement result is predicted according to a GA-BP neural network model method, and guidance is provided for dam construction.

Description

Technical Field

[0001] The invention relates to a prediction analysis modeling method for dam settlement based on a GA-BP neural network, and relates to the calculation of the settlement of a water conservancy project dam. Background Art

[0002] In civil engineering practice, regression fitting methods are often used to model the settlement of structures. However, regression modeling needs to meet certain conditions. For example, linear regression requires data to meet linear, normal, and independent conditions to be applicable, and two-dimensional linear models are difficult to apply to multi-dimensional, complex conditions, and multi-parameter calculations; on the other hand, in the case of nonlinear fitting, the assumptions, model selection, parameter setting, model correction, and model convergence of the regression method are difficult to evaluate, which ultimately makes it difficult to ensure the accuracy of the model. A large number of studies have shown that among the factors affecting the settlement of water conservancy project dams, there is a complex nonlinear relationship between factors such as the location of the dam settlement point and the thickness of the upper fill and the final settlement of the dam. However, there is currently no suitable model to refer to this nonlinear relationship. This makes it difficult to unify multi-dimensional parameters in a single dimension when using traditional nonlinear regression and other methods for modeling, and there are difficulties in many aspects such as model selection, initial value, result correction, and result verification. BP neural network is a branch of deep learning in machine learning. It has good model generalization performance and shows high accuracy when dealing with complex problems. However, in the construction of BP neural network, the selection of initial weights and thresholds is the key to excellent network performance. Too high weights and thresholds will cause gradient explosion, and too low weights and thresholds will cause gradient disappearance. Therefore, how to make BP neural network suitable for engineering practice is an urgent problem to be solved. Summary of the invention

[0003] The purpose of the present invention is to provide a prediction analysis modeling method for dam settlement based on GA-BP neural network, which enables more comprehensive and full utilization of engineering monitoring data, establishes a nonlinear mapping model between initial parameters of dam settlement and dynamic settlement, realizes machine learning to replace a large amount of theoretical calculation work, and on the other hand, predicts dam settlement results according to the GA-BP neural network model method to provide guidance for dam construction.

[0004] In order to realize the above technical features, the purpose of the present invention is achieved as follows: a prediction analysis modeling method for dam settlement based on GA-BP neural network, comprising: Step 1, establish a dam filling and settlement database; Step 2, using BP neural network to establish the nonlinear mapping relationship between dam structure, filling parameters and settlement; Step 3, BP neural network parameter setting; Step 4, using genetic algorithm to optimize the initial weights and thresholds of the neural network to form a GA-BP network model; Step 5: Use the test data to verify the accuracy of the model.

[0005] The specific steps of step 1 are: Step 1.1, establishment of dam framework: By collecting preliminary filling data of the dam, including surrounding terrain environment, dam design shape data and dam structure related information, the dam framework is constructed; Step 1.2, determination of the dam centerline and the upstream and downstream boundaries of the dam: Determine the specific location of the dam's central axis and the upstream and downstream dividing line based on the dam framework; Step 1.3, establishment of electromagnetic sedimentation tube monitoring database: During the dam filling construction process, monitoring points are set up on the central axis of the dam and at the upstream and downstream boundaries of the dam, and electromagnetic settlement tubes are buried at the monitoring points. After the burial is completed, the electromagnetic settlement tubes are used to record the corresponding settlement information of the dam and organize it into an electromagnetic settlement tube monitoring database; Step 1.4, establishment of dam filling and settlement database: Based on the design shape of the dam, the electromagnetic settlement tube monitoring database is transformed into an electromagnetic settlement tube three-dimensional coordinate conversion database, thereby forming a dam filling and settlement database.

[0006] The electromagnetic sedimentation tube monitoring database in step 1.3 is the original data recorded by the electromagnetic sedimentation tubes at the project site, which is provided by the project inspection personnel.

[0007] The electromagnetic settlement pipe monitoring database in step 1.3 is based on the location of the electromagnetic settlement pipe, the filling thickness of the lower part of the electromagnetic settlement pipe, the filling thickness of the upper part of the electromagnetic settlement pipe which is gradually filled up with the passage of time during the construction process, and time factors to form a database based on dam settlement.

[0008] In the specific conversion process of the electromagnetic sedimentation pipe three-dimensional coordinate conversion database in step 1.4, the position of the electromagnetic sedimentation pipe is based on the central axis of the dam as the X-axis, the upstream and downstream dividing line of the dam as the Y-axis, and the dam filling height as the Z-axis to form a three-dimensional Cartesian coordinate system to determine the relative coordinates of the electromagnetic sedimentation pipe; on the central axis of the dam, with the upstream and downstream dividing line of the dam as the center, the upstream is negative and the downstream is positive; on the upstream and downstream dividing line of the dam, with the central axis of the dam as the center, the upstream and downstream directions are the strikes, the left of the central axis of the dam is positive, and the right of the central axis of the dam is negative; with the filling height as the strike, upward is positive and downward is negative; then the data in the electromagnetic sedimentation pipe monitoring database is converted into three-dimensional coordinate data, and the accumulated time and accumulated settlement of each coordinate point corresponding to the monitoring point are arranged in correspondence to form a three-dimensional coordinate conversion database for the electromagnetic sedimentation pipe.

[0009] In the step 1.4, the electromagnetic sedimentation tube three-dimensional coordinate conversion database collects no less than 12,000 sets of data samples, and the data are uniform and dispersed. Before the model is established, the data is cleaned to remove invalid data, and then the data obtained after normalization is used as sample data of the BP neural network.

[0010] The genetic algorithm selection function adopts roulette selection, and iteratively searches for the individual with the highest fitness in the population as the initial threshold and weight of the neural network, thereby optimizing the neural network and forming a GA-BP neural network model.

[0011] The BP neural network adopts a four-layer structure of one input layer, two hidden layers and one output layer, wherein the input layer has five neurons, the first hidden layer has 10 neurons, the second hidden layer has 5 neurons, and the output layer has 1 neuron.

[0012] By inputting the converted sample data into the GA-BP neural network, the neural network model with the best performance is obtained through multiple trainings, and then the nonlinear mapping relationship between dam filling parameters and settlement is established.

[0013] The transfer function of the hidden layer of the BP neural network is the tansig function, the transfer function of the output layer is the purelin linear function, the training algorithm adopts the trainlm algorithm, the learning rate is set to 0.0005, and the network target error is 0~1×10 -5 The network performance evaluation function uses the mean square error MSE and the determination coefficient R 2 ; If the coefficient of determination R 2 If it is greater than 0.8, save the network. If it is less than 0.8, continue training until the determination coefficient R 2 Stop training when it is greater than 0.8; Establish the UI interface of the GA-BP neural network model, call the GA-BP neural network model, and package the program.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. The method of the present invention enables more comprehensive and full utilization of engineering monitoring data, establishes a nonlinear mapping model between the initial parameters of dam settlement and the dynamic settlement, and realizes machine learning to replace a large amount of theoretical calculation work. On the other hand, the GA-BP neural network model method is used to predict the dam settlement results, providing a reference basis for dam foundation treatment.

[0015] 2. The model constructed by the GA-BP network in the present invention can quickly predict the final settlement of the dam in the early construction stage, with a prediction accuracy of 80% to 90%. It has high accuracy and stability, provides effective guidance for the later dam construction, can effectively shorten the analysis cycle of dam settlement data, and reduce the time and labor costs required. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below.

[0017] Figure 1 An example of a dam frame according to the present invention.

[0018] Figure 2 An example of the layout of electromagnetic settlement pipes for a dam project according to the present invention.

[0019] Figure 3 An example of the original database for electromagnetic settlement pipe monitoring of a dam project in the present invention.

[0020] Figure 4 An example of the optimal fitness of the genetic optimization algorithm of the present invention.

[0021] Figure 5 The BP neural network structure of the present invention.

[0022] Figure 6 An example of the BP neural network simulation results of the present invention.

[0023] Figure 7 The GA-BP neural network model of the present invention calls the UI example.

[0024] Figure 8 The GA-BP neural network model logic flow chart of the present invention.

[0025] Table 1 An example of a three-dimensional coordinate transformation database of a single electromagnetic settlement tube of a dam project of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the implementation regulations described 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 are within the scope of protection of the present invention.

[0027] Embodiment 1: See also Figure 1-8 , a prediction analysis modeling method for dam settlement based on GA-BP neural network, including Step 1, establish a dam filling and settlement database; Step 2, using BP neural network to establish the nonlinear mapping relationship between dam structure, filling parameters and settlement; Step 3, BP neural network parameter setting; Step 4, using genetic algorithm to optimize the initial weights and thresholds of the neural network to form a GA-BP network model; Step 5: Use the test data to verify the accuracy of the model.

[0028] Furthermore, the specific steps of step 1 are: Step 1.1, establishment of dam framework: By collecting preliminary filling data of the dam, including surrounding terrain environment, dam design shape data and dam structure related information, the dam framework is constructed; Step 1.2, determination of the dam centerline and the upstream and downstream boundaries of the dam: Determine the specific location of the dam's central axis and the upstream and downstream dividing line based on the dam framework; Step 1.3, establishment of electromagnetic sedimentation tube monitoring database: During the dam filling construction process, monitoring points are set up on the central axis of the dam and at the upstream and downstream boundaries of the dam, and electromagnetic settlement tubes are buried at the monitoring points. After the burial is completed, the electromagnetic settlement tubes are used to record the corresponding settlement information of the dam and organize it into an electromagnetic settlement tube monitoring database; Step 1.4, establishment of dam filling and settlement database: Based on the design shape of the dam, the electromagnetic settlement tube monitoring database is transformed into an electromagnetic settlement tube three-dimensional coordinate conversion database, thereby forming a dam filling and settlement database.

[0029] Furthermore, the electromagnetic sedimentation tube monitoring database in step 1.3 is the original data recorded by the electromagnetic sedimentation tubes at the project site, which is provided by the project inspection personnel.

[0030] Furthermore, the electromagnetic settlement pipe monitoring database in step 1.3 is formed based on the location of the electromagnetic settlement pipe, the filling thickness of the lower part of the electromagnetic settlement pipe, the filling thickness of the upper part of the electromagnetic settlement pipe which is gradually filled up with the passage of time during the construction process, and time factors to form a database based on dam settlement.

[0031] Furthermore, in the specific conversion process of the electromagnetic sedimentation tube three-dimensional coordinate conversion database in step 1.4, the position of the electromagnetic sedimentation tube is based on the central axis of the dam as the X-axis, the upstream and downstream dividing line of the dam as the Y-axis, and the dam filling height as the Z-axis to form a three-dimensional Cartesian coordinate system to determine the relative coordinates of the electromagnetic sedimentation tube; on the central axis of the dam, with the upstream and downstream dividing line of the dam as the center, the upstream is negative and the downstream is positive; on the upstream and downstream dividing line of the dam, with the central axis of the dam as the center, the upstream and downstream directions are the strikes, the left of the central axis of the dam is positive, and the right of the central axis of the dam is negative; with the filling height as the strike, upward is positive, and downward is negative; then the data in the electromagnetic sedimentation tube monitoring database is converted into three-dimensional coordinate data, and the accumulated time and accumulated settlement of each coordinate point corresponding to the monitoring point are arranged in correspondence to form a three-dimensional coordinate conversion database for electromagnetic sedimentation tubes.

[0032] Furthermore, in the step 1.4, the electromagnetic sedimentation tube three-dimensional coordinate transformation database collects no less than 12,000 sets of data samples, the data are uniform and dispersed, and before the model is established, the data is cleaned to remove invalid data, and then the data obtained after normalization is used as sample data of the BP neural network.

[0033] Furthermore, the genetic algorithm selection function adopts roulette wheel selection, and iteratively searches for the individual with the highest fitness in the population as the initial threshold and weight of the neural network, thereby optimizing the neural network and forming a GA-BP neural network model.

[0034] Furthermore, the BP neural network adopts a four-layer structure of one input layer, two hidden layers, and one output layer, wherein the input layer has five neurons, the first hidden layer has 10 neurons, the second hidden layer has 5 neurons, and the output layer has 1 neuron.

[0035] Furthermore, by inputting the sample data of the BP neural network into the GA-BP neural network, the neural network model with the best performance is obtained through multiple trainings, and then the nonlinear mapping relationship between the dam filling parameters and settlement is established.

[0036] Furthermore, the transfer function of the hidden layer of the BP neural network is the tansig function, the transfer function of the output layer is the purelin linear function, the training algorithm adopts the trainlm algorithm, the learning rate is set to 0.0005, and the network target error is 0~1×10 -5 The network performance evaluation function uses the mean square error MSE and the determination coefficient R 2 ; Furthermore, if the coefficient of determination R 2 If it is greater than 0.8, save the network. If it is less than 0.8, continue training until the determination coefficient R 2 Stop training when it is greater than 0.8; Furthermore, a UI interface of the GA-BP neural network model is established, the GA-BP neural network model is called, and the program is packaged.

[0037] Embodiment 2: A dam settlement modeling method based on GA-BP neural network includes a dam frame; an electromagnetic settlement pipe; an electromagnetic settlement pipe monitoring database; an electromagnetic settlement pipe three-dimensional coordinate conversion database; a genetic optimization algorithm; a BP neural network; and a GA-BP neural network model calling UI.

[0038] The following steps are involved: 1. By collecting the preliminary filling data of the dam, including the surrounding terrain environment, the design shape of the dam and other data, the dam framework is established for the construction of the dam model. Based on the dam framework, the specific positions of the dam's central axis and the upstream and downstream dividing line of the dam are determined.

[0039] 2. During the dam filling process, electromagnetic settlement tubes are buried to collect settlement data during the construction process.

[0040] 3. Based on the location of the electromagnetic settlement pipe, the filling thickness of the lower part of the electromagnetic settlement pipe, the filling thickness of the upper part of the electromagnetic settlement pipe which gradually increases with the time during the construction process, and the time factor, a database based on the dam settlement is formed.

[0041] 4. Based on the design shape of the dam, the database is converted into a three-dimensional coordinate conversion database for electromagnetic sedimentation pipes, where the position of the electromagnetic sedimentation pipe is determined by forming a three-dimensional Cartesian coordinate system with the central axis of the dam as the X-axis, the upstream and downstream dividing line of the dam as the Y-axis, and the dam filling height as the Z-axis to determine the relative coordinates of the electromagnetic sedimentation pipe. On the central axis of the dam, with the upstream and downstream dividing line of the dam as the center, the upstream is negative and the downstream is positive; on the upstream and downstream dividing line of the dam, with the central axis of the dam as the center, the upstream and downstream directions are the strike direction, the left of the central axis of the dam is positive, and the right of the central axis of the dam is negative; with the filling height as the strike direction, upward is positive and downward is negative.

[0042] 5. After data cleaning, invalid data is eliminated, and the obtained data is used as sample data of BP neural network after normalization; 6. Design the neural network topology, including the number of neurons in the input layer, the number of hidden layers, the number of neurons in the output layer, the target accuracy of the network model, the learning rate, the activation function and other data.

[0043] 7. Optimize the initial weights and thresholds of the neural network through genetic algorithm to form a GA-BP neural network model; 8. Input the above sample data into the GA-BP neural network, obtain the neural network model with the best performance through multiple training, and establish the nonlinear mapping relationship between dam filling parameters and settlement.

[0044] 9. Use the design program to design a calling UI for the GA-BP neural network model and call the neural network model established in the above steps.

[0045] Embodiment 3: 1. See Figure 1 , from a water conservancy dam project, obtain relevant information on the dam structure, determine and establish the dam framework, and based on the dam framework, determine the specific location of the dam's central axis and the upstream and downstream dividing line of the dam.

[0046] 2. See Figure 2 , Figure 3 , collect electromagnetic settlement pipe data, including the location of the electromagnetic settlement pipe, the filling thickness of the lower part of the electromagnetic settlement pipe, the filling thickness of the upper part of the electromagnetic settlement pipe that gradually fills up with the passage of time during the construction process, and time factors, to form data based on dam settlement.

[0047] 3. Table 1 is an example of a three-dimensional coordinate conversion database of a single electromagnetic settlement tube of a dam project of the present invention. Based on step one, the data in the database of step two is converted into three-dimensional coordinate data, and the accumulated time of each coordinate point corresponding to the monitoring point is arranged in correspondence with the accumulated settlement to form a settlement monitoring database 2, with a total of 12,868 groups of data.

[0048] Table 1 Example of three-dimensional coordinate transformation database of a single electromagnetic sedimentation tube of a dam project of the present invention

[0049] 4. Extract the data of some random monitoring points in the database separately as the test set, and use the other data as the training set to train the network, and normalize the training data to the interval of [-1.1].

[0050] 5. See also Figure 5 , establish a BP neural network model, and determine the initial parameters: In this case, it is determined that the BP neural network adopts a three-layer structure with one input layer, one hidden layer, and one output layer. The input layer has five neurons, the hidden layer has 10 neurons, and the output layer has 1 neuron. The transfer function of the hidden layer of the BP neural network is the tansig function, and the transfer function of the output layer is the purelin linear function. The training algorithm adopts the trainlm algorithm, the learning rate is set to 0.0005, and the network target error is 0~1×10-5.

[0051] 6. See Figure 4 , a genetic algorithm optimization model was established, the roulette wheel selection was used as the selection function, the population size was set to 50, and the iteration was performed 100 times. The individual with the highest fitness in the population was found through iteration as the initial threshold and weight of the neural network.

[0052] 7. See Figure 8 , establish the GA-BP neural network model.

[0053] 8. See Figure 6 After the network training is completed, the test set is input into the trained network for testing. The network performance evaluation function uses the mean square error MSE and the determination coefficient R 2 .

[0054] 9. If the coefficient of determination R 2 If it is greater than 0.8, save the network. If it is less than 0.8, repeat steps 5 to 7 until the determination coefficient R 2 If it is greater than 0.8, training is stopped.

[0055] 10. See Figure 7 , establish the GA-BP neural network calling UI, and package the program to make the model easier to use.

[0056] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A prediction analysis modeling method for dam settlement based on GA-BP neural network, characterized by: include Step 1, establish a dam filling and settlement database; Step 2, using BP neural network to establish the nonlinear mapping relationship between dam structure, filling parameters and settlement; Step 3, BP neural network parameter setting; Step 4, using genetic algorithm to optimize the initial weights and thresholds of the neural network to form a GA-BP network model; Step 5: Use the test data to verify the accuracy of the model.

2. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 1 is characterized in that: The specific steps of step 1 are: Step 1.1, establishment of dam framework: By collecting preliminary filling data of the dam, including surrounding terrain environment, dam design shape data and dam structure related information, the dam framework is constructed; Step 1.2, determination of the dam centerline and the upstream and downstream boundaries of the dam: Determine the specific location of the dam's central axis and the upstream and downstream dividing line based on the dam framework; Step 1.3, establishment of electromagnetic sedimentation tube monitoring database: During the dam filling construction process, monitoring points are set up on the central axis of the dam and at the upstream and downstream boundaries of the dam, and electromagnetic settlement tubes are buried at the monitoring points. After the burial is completed, the electromagnetic settlement tubes are used to record the corresponding settlement information of the dam and organize it into an electromagnetic settlement tube monitoring database; Step 1.4, establishment of dam filling and settlement database: Based on the design shape of the dam, the electromagnetic settlement tube monitoring database is transformed into an electromagnetic settlement tube three-dimensional coordinate conversion database, thereby forming a dam filling and settlement database.

3. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 2 is characterized in that: The electromagnetic sedimentation tube monitoring database in step 1.3 is the original data recorded by the electromagnetic sedimentation tubes at the project site, which is provided by the project inspection personnel.

4. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 2 is characterized in that: The electromagnetic settlement pipe monitoring database in step 1.3 is based on the location of the electromagnetic settlement pipe, the filling thickness of the lower part of the electromagnetic settlement pipe, the filling thickness of the upper part of the electromagnetic settlement pipe which is gradually filled up with the passage of time during the construction process, and time factors to form a database based on dam settlement.

5. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 2 is characterized in that: In the specific conversion process of the electromagnetic sedimentation pipe three-dimensional coordinate conversion database in step 1.4, the position of the electromagnetic sedimentation pipe is based on the central axis of the dam as the X-axis, the upstream and downstream dividing line of the dam as the Y-axis, and the dam filling height as the Z-axis to form a three-dimensional Cartesian coordinate system to determine the relative coordinates of the electromagnetic sedimentation pipe; on the central axis of the dam, with the upstream and downstream dividing line of the dam as the center, the upstream is negative and the downstream is positive; on the upstream and downstream dividing line of the dam, with the central axis of the dam as the center, the upstream and downstream directions are the strikes, the left of the central axis of the dam is positive, and the right of the central axis of the dam is negative; with the filling height as the strike, upward is positive and downward is negative; then the data in the electromagnetic sedimentation pipe monitoring database is converted into three-dimensional coordinate data, and the accumulated time and accumulated settlement of each coordinate point corresponding to the monitoring point are arranged in correspondence to form a three-dimensional coordinate conversion database for the electromagnetic sedimentation pipe.

6. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 2 is characterized in that: In the step 1.4, the electromagnetic sedimentation tube three-dimensional coordinate conversion database collects no less than 12,000 sets of data samples, and the data are uniform and dispersed. Before the model is established, the data is cleaned to remove invalid data, and then the data obtained after normalization is used as sample data of the BP neural network.

7. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 6 is characterized in that: The genetic algorithm selection function adopts roulette selection, and iteratively searches for the individual with the highest fitness in the population as the initial threshold and weight of the neural network, thereby optimizing the neural network and forming a GA-BP neural network model.

8. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 7 is characterized in that: The BP neural network adopts a four-layer structure of one input layer, two hidden layers and one output layer, wherein the input layer has five neurons, the first hidden layer has 10 neurons, the second hidden layer has 5 neurons, and the output layer has 1 neuron.

9. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 7 is characterized in that: By inputting the converted sample data into the GA-BP neural network, the neural network model with the best performance is obtained through multiple trainings, and then the nonlinear mapping relationship between dam filling parameters and settlement is established.

10. The prediction analysis modeling method for dam settlement based on GA-BP neural network according to claim 8 is characterized in that: The transfer function of the hidden layer of the BP neural network is the tansig function, the transfer function of the output layer is the purelin linear function, the training algorithm adopts the trainlm algorithm, the learning rate is set to 0.0005, and the network target error is 0~1×10 -5 The network performance evaluation function uses the mean square error MSE and the determination coefficient R 2 ; If the coefficient of determination R 2 If it is greater than 0.8, save the network. If it is less than 0.8, continue training until the determination coefficient R 2 Stop training when it is greater than 0.8; Establish the UI interface of the GA-BP neural network model, call the GA-BP neural network model, and package the program.

Images