The balance control of two-wheeled robot based on bionic learning algorithm

According to the motion balance for a two-wheeled robot control problems, we put forward a bionic learning algorithm based on growing cell structure (GCS) network and Q-learning. GCS network has in addition to the competitive mechanism of SOM network, and it can also carry out self-organizationally evolution through the continuous growth of new neurons. Q-learning algorithm is a model free reinforcement learning algorithm, and it can improve the learning ability of the control system, but it is only suitable for the discrete state. We made the growth characteristics of GCS network apply to the Q-learning algorithm, and optimized the Q value through the information of the winning neuron which comes from the network. Ultimately, we achieved the model free control of a continuous state system, and made simulation experiments on two-wheeled robot. The results showed that the robot learned to effectively control the movement balance through continuous growth and improvement of neurons, and verified that it was effective and feasible of the bionic learning algorithm based on the growth network for the robot´s motion balance control.