Genetic Algorithm-based Control of a Two-Wheeled Self-Balancing Robot

Mobile robots are becoming increasingly popular in a wide array of applications: industrial, item delivery, search and rescue, space, social, and entertainment. A two-wheeled self-balancing mobile robot is a statically unstable non-linear system with strong coupling dynamics. Common practices in the development of control systems for such robots are either to linearise the region of application to be used with linear controllers or to use complex nonlinear controllers such as fuzzy logic, sliding mode, and neural networks. However, self-balancing robots are still restricted by the travelling distance needed to regain an upright stance, the length of settling time, high overshoot and lack of resilience to external disturbance. In this paper, we are proposing a novel genetic algorithm-based switching control to evolve more effective control parameters and increase autonomy. Differently from previous work a genetic algorithm has been used to select the parameters in a sliding mode control and a switching-algorithm-based controller of a two-wheeled self-balancing mobile robot. The performance of the proposed controllers is assessed in simulations using the CoppeliaSim environment. The tests used dynamic criteria (distance travelled, maximum angular deviation), control criteria (settling time, % overshoot). The results showed that the genetic algorithm-based control has better performance in the 55 degree recovery, impulse response and variable inclination tests and that switching algorithm-based control shows better performance in step response tests. The results produced by the evolutionary algorithm are often able to perform better than their analytic counterparts. This shows the potential of meta-heuristic algorithms to obtain solutions for optimization problems encountered by statically unstable non-linear systems in unstructured and fast-changing environments.