Comparative Study on Control Methods for Two-Degree-of-Freedom Robotic Manipulators

Abstract

Robotic manipulator systems are inherently nonlinear systems. They usually show model uncertainty and external disturbances. This paper does a systematic comparative analysis of three main control methods for two-degree-of-freedom (2-DOF) robotic manipulators: classical Proportional-Integral-Derivative (PID) control, Model-Based Adaptive Control (MBAC), and intelligent adaptive control based on Radial Basis Function (RBF) neural networks. First, a Lagrangian dynamic model of the 2-DOF robotic manipulator is established. Subsequently, the control laws of the three controllers are elaborated in detail, and system stability is proven through the Lyapunov stability principle. Comparative verification of trajectory tracking is conducted on a simulation platform. The performance differences among the three methods are analyzed in terms of tracking accuracy, convergence speed, disturbance rejection capability, and computational cost. Experimental results indicate that control precision progressively improves from PID to MBAC, and further to RBF neural network control, albeit with a corresponding increase in computational complexity. This study provides a comprehensive performance comparison of classical, model-based, and intelligent control strategies, offering guidance for selecting optimal controllers in nonlinear robotic systems.