Optimizated RBF Neural Network With Co-evolutionary PSO In The Application Of Gearbox Faults Diagnosis

Fault diagnosis technology is the product of modern manufacturing's development, gearbox is the most important component of engineering machinery, gear and rolling bearing is the vulnerability components in gearbox. According to statistics, 80 percent of transmission machinery fault is caused by gearbox fault, and therefore, the operation of gearbox condition monitoring and fault diagnosis has been the focus of the mechanical fault diagnosis technology.On the basis of concentrating study on the related theory of co-evolutionary PSO algorithm, combining with the specific problems of RBF neural network optimization, the article proposes a kind of RBF neural network optimization model based on co-evolutionary PSO algorithm, and then optimized RBF neural network will be applied in gearbox fault diagnosis technology in order to make timely, accurate and effective judgment for gearbox fault condition or abnormal state, guiding the operation of gearbox system, improve the reliability, safety and effectiveness of gearbox system, eventually brought the economic loss by gearbox fault to a minimum level.Experimental results show that the optimization model is feasible, and then comparing the optimized RBF neural network test results with the traditional RBF neural network test results, this paper proposed optimization model has higher training accuracy and higher convergence speed. Therefore, through this paper research, not only provides a new optimization approach for the RBF neural network, but also greatly improves RBF neural network in gearbox fault diagnosis technology, then enriches and develops the application of PSO and neural network in gearbox fault diagnosis.