Real-time navigation for multiple mobile robots in dynamic environments based on fuzzy logic

Abstract

The work presented in this paper deals with a navigation problem for wheel mobile robots in unknown environments. Dynamic environments usually contain not only static but also moving obstacles. The environments are completely unknown to the robot; thus, in order to guide the robot along a collisions-free path efficiently to reach the goal position quickly, proximity sensors installed on the robot's body must be used to detect information about the surroundings. In this paper, the navigation approach is achieved by a combination of primary behaviors which are used for dealing with different situations, that is, a danger judgment behavior based on the judgment of the direction and dangerous level of upcoming danger is developed for avoiding obstacles; when there is no collision risk, the robot will be guided towards its target according to a target positioning behavior. All the behaviors are achieved by means of fuzzy inference system. The proposed navigation method is demonstrated by simulation experiments implemented using Matlab; and the simulation results verified that the proposed method effectively addresses the mobile robot navigation problem.