Optimizing Continuous Finite-Time Terminal SMC for Two-Link Robot Arms

This paper presents a comprehensive study on the dynamics and control strategy of n-link rigid robot manipulators, with a focus on a two-link manipulator as a representative example. The standard dynamic model is revisited, and a continuous finite-time terminal sliding mode (TSM) controller is considered for trajectory tracking. The controller features a nonlinear sliding surface that ensures finite-time convergence, which significantly enhances tracking performance compared to conventional sliding mode control. The controller parameters are optimized using three global optimization algorithms: Particle Swarm Optimization (PSO), Bayesian Optimization (BO), and Covariance Matrix Adaptation Evolution Strategy (CMA-ES). Simulation results indicate that while all three methods achieve similar objective function values (costs), their optimized parameter configurations vary significantly, providing multiple near-optimal solutions for practical implementation. The study provides practical guidance on selecting both controller parameters and optimizers for low-complexity, high-robustness manipulator control, and lays groundwork for real-time, auto-tuned implementations.