Balance control of two-wheeled robot based on reinforcement learning
Sun Liang, Feimei Gan · 2011
Two-wheeled robot is a high-order, non-stable, non-linear, typical control system. This paper present a novel reinforcement learning algorithm to balance control of two-wheeled robot, when its model is not available and the agent has no a priori control knowledge. And it constructs performance evaluation function by using neural networks, uses their own neural network learn online, it can achieve balance control of self-learning two-wheeled robot. The simulation results demonstrate that it can successfully achieve self-learning balance control of two-wheeled robot System in a short time.