Nonsingular Fixed-Time Terminal Sliding Mode Control Based Emotional Neural Network for Nonlinear Systems

A fixed-time nonsingular terminal sliding mode control (TSMC) method based on emotional neural network (ENN) is proposed for nonlinear systems. First of all, a new terminal sliding surface (TSS) is proposed in this article, which can adaptively adjust when the system states are far and near the origin and achieve fixed time fast convergence. Second, the equivalent control term in the controller is approximated by the ENN, which solves the singularity problem of traditional TSMC and the problem of unknown model information simultaneously. Finally, by using Lyapunov stability theorem, it is proved that the system states are stable in a fixed time during both sliding and reaching stages, which is independent of the initial values. The proposed method is verified and compared on an inverted pendulum system, which shows a better control performance.