Approximate estimation of fatigue life of hydraulic support

In view of problem that conventional life approximate analytical method of hydraulic support needs cyclic loading and acquisition of dangerous points which leads to increase of computer load, a life estimation model based on genetic algorithm and BP neural network was proposed. The BP neural network is optimized by using global search performance of genetic algorithm to avoid falling into the local minimum points. The optimized BP neural network is used to establish network mapping model between structural parameters of the dangerous points and fatigue life. The test results of sample size and number of hidden layer nodes show that the estimation accuracy of the model is high when the sample size is 40 and the number of hidden layer nodes is 7; the average value of estimated life of hydraulic support is 36 456 times, and the maximum relative error with the theoretical value is 5.27%.