Position Control of a Intelligent Module based on Neural Network Sliding Mode Control

Abstract

Based on the standard intelligent module of the modular robot, a kind of neural network sliding mode control method is put forward. The neural network is used to approach to the function between the state hyperplane of the system and the reaching law. A hyperbolic tangent function is applied to replace the saturated function in order to realize the boundary method design of the sliding mode control. Simulation results show that system owns quick response and strong antijamming capability. Moreover, the chattering of neural network sliding mode control is weakened effectively, and problems that cannot be solved with the traditional PID control method under complicated environment and conditions such as variable load etc can be solved.