Design of Sliding Mode Controller for a Two Wheeled Balancing Robot

Ng, Shin Huan (2020) Design of Sliding Mode Controller for a Two Wheeled Balancing Robot. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text NgShinHuan_Full Text.pdf Restricted to Registered users only Download (3MB)

Abstract

A two-wheeled balancing robot (TWBR) is a robot which can balance itself through its pendulum. One of the most common of the TWBRs is Segway. The TWBR can balance itself no matter in what condition, a flat surface, a slope or even in an uneven terrain. To balance the TWBR, there are many ways to develop. One of the methods is implement the control algorithm to control the TWBR. The controllers consist of the most famous techniques, namely proportional-integral-derivative (PID) controller, cascade controller which is the combination of two or more PID controller, sliding mode controller (SMC), non-linear PID (N-PID) controller, H-infinity controller, higher order SMC (HOSMC) super twisting SMC (STSMC) and the hybrid controller like SMC-PID controller. These controllers are able to help to balance the TWBR but the performances are different for each type of the controller. Most of these methods are done through the simulation especially in MATLAB/Simulink. In this project, a new method is to be developed namely the STSMC-PID control. This controller has the benefits from both STSMC and PID controller. The STSMC has a better disturbance rejection than the PID controller while the PID controller can help to increase the transient responses in the STSMC like hybrid controller (SMC-PID control). In this project, 3 controllers are used to simulate the TWBR through SIMLUINK of MATLAB which are the PID controller, the STSMC and the STSMC-PID controller. The PID controller is used as the benchmark. The mathematical modelled of the TWBR is derived from the state space. These 3 controllers are simulated by 2 types of disturbances namely impulse (force exerted in fixed interval) and step (force exerted in a period of time then released). After that, the control performances include settling time, maximum overshoot, steady state error and disturbance rejection of each type of the controller are observed and analysed. The reference input is 0. After observing and analysing those control performances through 2 disturbances, the final control performance is the robustness of the controller which means that particular controller can handle both type of disturbance without changing its parameters. Under impulse disturbance, the STSMC-PID controller showed the best result as compared to the STSMC and the PID controller. The STSMC-PID controller improves the settling time by 55.38% as compared to the STSMC which just improves by 52.31%. Under step disturbance, the STSMC has the lowest settling time among 3 types of controllers but only the STSMC-PID controller can balance itself when the force is exerted on the TWBR. Undeniably, the STSMC-PID controller is still better than individual the STSMC. As a result, the STSMC-PID controller will greatly improve the stability performance of the TWBR as well as having a better disturbance rejection. Lastly, the STSMC-PID controller is the highest robustness controller throughout the simulation because the tuning parameters are maintained under 2 disturbances through simulation.