Research On Risk Prediction System Of Booster Production Line Based On GA-Elman Neural Network

As a kind of blasting equipment to detonate insensitive explosives,Booster is widely used in infrastructure construction,which plays an important role in civil blasting,and makes a great contribution to economic construction.At the same time,we should pay more attention to the safety of the production line of the Booster.If the safety problems of the production line of the Booster are not strictly controlled and the risks are not predicted,the explosion will bring serious consequences.The production processes of the Booster are complex,involving many factors,so the traditional prediction methods can not solve the problems of risk management and control of detonator production line.Therefore,in this paper,the improved genetic algorithm(GA)optimized Elman neural network is used to establish the risk prediction model,and based on the model,the risk prediction of the Booster production line is carried out.Finally,the analysis and verification are carried out in the laboratory environment.The main contents of this paper are stated as follows:(1)Firstly,it introduces the research background and significance of this paper and the development trend of risk management at home and abroad,summarizes the purpose of risk management and the shortcomings of current risk management methods,and leads to the solution of risk prediction based on neural network,At the same time,the realization scheme of risk prediction for detonator production line is given.(2)In the original structure of Elman neural network,the input layer and output layer are added to retain the previous information of input layer and output layer.Considering the concentration and dispersion of population in genetic algorithm,the crossover probability and mutation probability of genetic algorithm are improved adaptively.At the same time,in order to speed up the optimization effect of the model,the sorting selection strategy and the optimal saving strategy are combined to optimize the original selection operator operation.Then the improved genetic algorithm is used to optimize the weight and threshold of Elman neural network structure,which can solve the problem of Elman neural network weight threshold falling into the minimum value,and effectively improve the training efficiency of the model.(3)Then,based on the detailed analysis of the process and equipment of the Booster production line,the risk factors in the production line are analyzed by the risk identification theory,and the PCA method is used to reduce the dimension of the data samples of controllable risk factors,and several items with larger influence factors are obtained as the input of the prediction model.And the concrete realization scheme of the risk prediction system of Booster is given.(4)Based on the improved GA-Elman neural network,the risk prediction model of Booster production line is established,and the parameters of genetic algorithm and Elman neural network structure are selected.The improved GA-Elman neural network model and the traditional GA-Elman neural network model are used for simulation and comparison tests,and the training performance and prediction results of the two models are analyzed and compared,reflecting the improved GA-Elman spirit The accuracy of network prediction is higher,and the training speed of network model is faster.Then,the comparative experiments of the improved GA-Elman neural network,the improved GA-BP neural network and the single Elman neural network are carried out,and the simulation results are analyzed to show the superiority of the improved GA-Elman neural network model.(5)Finally,the risk prediction system of the Booster production line is designed in detail from two aspects of software and hardware.In the aspect of hardware,it mainly selects sensors,controllers and industrial computers.In the aspect of software,the interface of the system is designed,and different situations are simulated to verify the effectiveness of the prediction system.