Collision Obstacle in Dynamic Environment Based Heuristic Using Sonar and Optical Flow Sensors

A method of real-time collision obstacle for autonomous mobile robots in local dynamic environment using multi-sensor data fusion is proposed. Due to the effectiveness of the optical flow sensors, the robot can acquire more accurate self-motion information. This paper also proposes an approach to complete enhancement based on the earlier Vector Field Histogram algorithm (VFH), which is called VFH#. This paper improved a new heuristic function to the VFH# algorithm in particular. By using this way, robots can select a more appropriate steeling angle in local dynamic environments. The results of experiments with a Pioneer robot demonstrated that robot collision obstacle can be obtained using the proposed method.