Research on rolling mill hydraulic fault diagnosis based on GA improved ANN algorithm

In view of the difficulties in fault feature extraction, nonlinear signal changes and large data volume in the current fault diagnosis process of rolling mill servo hydraulic cylinder, a hydraulic fault diagnosis study based on GA (Genetic Algorithm) improved ANN (Artificial Neural Network) algorithm was proposed. Taking advantage of the rapidity of artificial neural network fault diagnosis, the signal was normalized and put into ANN for training, and then the network parameters were optimized through back propagation learning to improve the diagnosis accuracy. According to the working principle of the rolling mill system, a rolling mill system simulation model was established to simulate the internal leakage fault condition of the rolling mill. The research results show that the use of GA improved ANN algorithm can effectively optimize the average error and data accuracy, and effectively reduce the number of iterations, indicating that the GA improved ANN method can be used to improve the performance of neural network calculations.