Research on Process Parameter Prediction Method of Medium Carbon Steel Heat Treatment Based on Improved Neural Network Algorithm
In the past practice of heat treatment engineering,the traditional BP neural network algorithm was applied to the practice of heat treatment engineering,and the data convergence speed was often low,and it was easy to fall into the local optimal solution.By establishing a neural network optimized by genetic algorithm,namely GA-BP neural network,the powerful global optimization of genetic algorithm was utilized.The traditional BP neural network is easy to fall into the local optimal solution,easy to overfit and other drawbacks,and then by collecting heat treatment experiment data,training and testing the separate BP neural network and GA-BP neural network.With the help of MATLAB,which can run some programs,the optimization results of low carbon steel heat treatment by BP neural network before optimization and GA-BP neural network after optimization are compared to confirm the role and advantages of GA-BP neural network algorithm in metal heat treatment optimization.The GA-BP neural network algorithm proposed in this paper can be applied to other kinds of metal heat treatment process optimization to guide engineering practice.
BP neural networkgenetic algorithmmedium carbon steelheat treatmentprocess optimization
NETL
NSTL
万方数据
