Stability Model of Robot Manipulators by using Adaptive Fast Terminal WNN-based Sliding Mode Control

Stability Model of Robot Manipulators

Abstract

This paper investigates a control strategy for robot manipulators to address the stability of the system, in which the introduced controller is based on adaptive fast terminal sliding mode control (FTSMC) with wavelet neural networking. To handle uncertainties and unknown dynamics in the robotic system, a wavelet neural network (WNN) is introduced to compensate and approximate for the nonlinearity of the dynamic model through specifically designed the adaptive update laws. The adaptive tuning rules for the WNN parameters are determined through a Lyapunov-based stability analysis, which ensures convergence and stability of the closed-loop system. To enhance control performance, a FTSMC framework is designed to determine the system's tracking error to the sliding surface within finite time and subsequently improves convergence characteristics. The proposed control scheme demonstrates that the WNN can effectively learn and represent the robot dynamics while maintaining finite-time convergence and adaptive FTSMC performance.