The Research On The Stability Control Of Two-wheeled Self-balancing Robot Based On Genetic Algorithms

Two-wheeled self-balancing robot as a kind of the wheeled robot is anonlinear and strong coupling with multiple variables unstable system which hasbroad development prospects and is applied in service, industry, business, and soon. It makes the particularity of the system itself have not only lowermanufacturing cost but also small volume. Based on the characteristics above,the self-balancing robot not only has strong ability to adapt the terrain and butalso agile motion behavior. At the same time the application of control theory putforward higher request. Two-wheeled self-balancing robot meet the demands. Itis the ideal platform to test various control theory and methods.Two-wheeled self-balancing robot is a system which has two wheeledcoaxial and depends on the monitor angle and angle of a tilt sensor andgyroscope rate of change to maintain its balance and stability. This paper isfocused on the simplified model of the system, and analyzes its controllabilityand observability, then researches for the stability of the system. To achieve itsstability control and motion control, the system is decoupled into two subsystems,and then controlled respectively.With MATLAB to research the LQR control for system model and to applygenetic algorithm to optimize parameters, the paper uses optimized parameters todesign controller, and then applies the controller to the two-wheeled self-balancing robot GBOT1001which is produced by Googol Company. At last, thesimulation and real-time experiment show that the real-time control achieves abetter control effect.