PID Control Based On Neural Network Of Backlash Nonlinear Study

Nonlinear Backlash in mechanical transmission servo systems greatly reducing the control system dynamic performance, and even leading to the imbalance between system stability. Therefore, servo systems backlash compensation of nonlinear control research is of great significance.Firstly, different backlash models and compensation approaches are introduced. Secondly, set up a typical gear transmission dynamic model, with continuously differentiable approximation of the model to express the traditional backlash dead-zone model. Finally the application of artificial neural network based on the design of BP neural network PID controller. The simulation results show that the neural network-based BP PID control method can greatly mitigate the effects of backlash, increase tracking accuracy and improve system robustness, compared with traditional PID control law.