Finite Time Motion Control of a Differential Drive Mobile Robot in a Seed Sowing Line Making Application Using Optimal Adaptive Sliding Mode Controller

Abstract

This thesis focuses on the control of a differential drive mobile robot, where a dynamic model is derived using the Lagrangian formulation method. The proposed control method utilizes optimal adaptive sliding mode control theory, combining the strengths of sliding mode control and adaptive control laws. To address the challenges posed by the nonlinear model, parameter uncertainties, and external disturbances, a robust adaptive sliding mode control is designed for the speed control system of the mobile robot. Additionally, the use of finite- time integral sliding mode control improves the asymptotic convergence in conventional sliding mode control. Additionally, a potential field function based on gradient descent is employed for obstacle avoidance.The control system consists of two control laws. The kinematic outer loop controller ensures that the posture error and angular error of the mobile robot converge to zero. The second control law utilizes finite-time integral sliding mode control to generate the desired linear and angular velocity, which is compared with the actual velocity. The controller parameters are optimized using particle swarm optimization (PSO) to achieve desired performance in terms of trajectory tracking accuracy, velocity tracking accuracy, and chattering suppression, as measured by the integral time absolute error (ITAE). The stability of the proposed controller is analyzed using the Lyapunov method. The suggested controller demonstrates efficient trajectory tracking, robustness against random external disturbances, and rapid response time with high tracking performance. The effectiveness of the controller is further validated by subjecting the system to various mismatched parameters and random external disturbances. The findings indicate that along the θ axis, the TSMC (ISE) method with model uncertainty resulted in the smallest error of 2.101, while the TSMC (ITAE) method yielded the largest error of 48.87. Similarly, the PSO-ATSMC (ISE) method produced the smallest errors along the X and Y axes, with errors of 2.139 and 2.247, respectively. In contrast, the TSMC (ITAE) method generated the largest errors of 28.27 along the X axis and 42.95 along the Y axis, as shown in Table 5.5. Thus, the PSO-ATSMC demonstrates superior performance in terms of robustness and time-domain measurement metrics, with shorter rise time, settling time, and less overshoot compared to the TSMC without the adaptive law which shown in a Table 5.6.