Real-Time Path Planning for Humanoid Robot

In order to make humanoid robot walk freely in our environment we discretized humanoid robot´s actions into a set of given actions and state space into grids. The environment map was built by use of stereo vision and the special method of plane segmentation. The shape of humanoid robot was simplified into double-cylinders model for the sake of checking obstacles. In the end a sequence of actions, which waste the least cost, was found out from the environment map as a path. The results of simulation tested the validity of the method.