AN IMPROVED SLIDING SURFACE AND ARTIFICIAL NEURAL NETWORK WITH AN APPLICATION FOR ROBOT CONTROL

Abstract

Mobile robots are autonomous devices that capable of moving on themselves, with applications ranging from surveillance and warehouse logistics to healthcare and planetary exploration. Precise trajectory tracking control is a key component in robotic applications. This research applies an improved sliding surface and atificial neural network for Mobile robot. The improved sliding surface combined with exponential approach law and hyperbolic tangent function are used to reduce the chattering phenomenon in the sliding mode control. The nonlinear components in the sliding mode control law are estimated using an artificial neural network. The weights of this neural network are updated online using the gradient descent algorithm. Lyapunov theory is used to prove the stability of the system. Simulation results in MATLAB/Simulink show the effectiveness of the proposed method with the rising time achieves 0.071(s), the overshoot is 0.004(%), the steady-state error converges to zero, and the settling time is about 0.0978(s) in x-coordinate and 0.0646(s), 0.0042(%), 0(m) và 0.0902(s) in y-coordinate, respectively; the chattering phenomena has small amplitude and low oscillation frequency.