ojme.02.2016.17.19

We propose that using the simulation of human imaginary thinking yields better results than logical thinking in the study of obstacle avoidance behavior of wheeled mobile robots. The simulation of human imaginary thinking can be realized by intelligent algorithm. In experiment, BP Neural Network is used to construct the abstracted vectors of the imaginary thinking, which will train and build the network. The experimental results show that robots using logical thinking have local circling while robots using imaginary think do not, which indicates the realization of human intelligent obstacle avoidance in the latter case.