Inverse Compensation Method for the Executor with Deadzone Based on Adaptive Inverse Control
Ping Yu, Zhongbin Zhao, Guo-Ju Bao · 2007
A deadzone inverse compensation of second-order neural network method based on adaptive inverse control is proposed for the nonlinear compensation of the system input. The neural network here is applied to approximating the inverse of deadzone to eliminate the effect of system deadzone, which achieved the dynamic compensation for the unknown nonlinearity of deadzone. This method doesn't require any hypothesis of nonlinear and constraint for deadzone, and realized the compensation strategy composed of estimator and compensator using neural network, which provides a new solution for nonlinear compensation.