Robust Optimal Controller for Two-wheel Self-Balancing Vehicles Using Particle Swarm Optimization

Abstract

—Control of a self-balancing vehicle is a challenging but exciting research topic. The challenge of researching self-balancing bicycles is maintaining balance when the bike is stationary and when the bike is moving. This paper, through analysis and comparison of two-wheeled vehicle balancing methods, shows that the method that best meets the requirements of the two-wheeled vehicle balance control problem is the balancing method using a flywheel stabilizer. Compared with the gyroscopic flywheel stabilizer, the inverted pendulum flywheel stabilizer has the advantages of fast response speed and energy saving, so we choose the pendulum flywheel stabilizer to reverse to control the balance of the two-wheeler. By modeling and analyzing the two-wheel vehicle model, it shows that the vehicle model is subjected to uncertainties. Hence, the robust controller is an appropriate controller for balancing two-wheel vehicles. However, the controller designed according to the robust control algorithm RH ∞ is often high-order, affecting the actual control quality. We proposed using the particle swarm optimization (PSO) algorithm to find a low-order robust controller from the high-order robust controller. By comparing the efficiency of the low-order robust controller according to PSO with the high-order robust controller and other low-order robust controllers, we have proven the correctness of the low-order robust controller according to PSO. Simulation results show that a two-wheel vehicle using a low-order robust controller according to PSO can stabilize the vehicle and give good control quality.