A predictive collision-free fuzzy algorithm for mobile robots to avoid obstacles moving at variable speeds

In this paper, a new algorithm for mobile robot navigation is presented. Some previous works have been proposed to solve this problem which only assumes the obstacles moving at constant speeds. However, differently from the unrealistic assumption, the obstacles can move at variable speed in the real world. The goal of the proposed algorithm is assisting mobile robots smoothly avoid moving obstacles which continuously change their speed by predicting collision. The mechanism is based on fuzzy logic system of two inputs which are the angle of obstacle relative to the direction of robot and the distance between obstacle and robot. The output will be adjustment angle of the heading of robot to avoid possible collisions with obstacles.