Intelligent Neural Sliding Control for Planetary Gear Type Inverted Pendulum Mechanism

Abstract

An intelligent neural sliding controller is developed for planetary train type inverted pendulum mechanism. The control methodology is based on the sliding mode control. The switching function in the normal control law is replaced with a bipolar sigmoid function. A fuzzy neural network is used to identify the pendulum dynamics. Adaptive tuning law is derived. The bipolar sigmoid function is thus adjusted according to the result of the identification process.