Pattern recognition and following of wall by behavior-based mobile robot

The movement of mobile robot along wall is classified into wall type recognition, straight wall following and turning around corners according to the behavior based idea. Mobile robot measures distance between its body and the wall it follows by three ultrasonic sensors, it recognizes the wall type online with neural network, follows straight wall with fuzzy controller and turns around 3 kinds of corners at their optimal turning radius by calculating models for the corners. The results of simulation and experiments according to TIT-I mobile robot indicate that the control strategy for wall recognition and following proposed in this paper is practicable and effective.