Research On The Complete Area Coverage Algorithm Of Mobile Robot

This dissertation was supported in part by the National Natural Science Foundation of China under Grant 60234030 and in part by the National Basic Scientific Research Founds under Grant A1420060159. As a part of the above-mentioned projects, this dissertation presents complete coverage algorithms for single robot and multi-robots under unknown static areas respectively. The algorithms enable mobile robots to traverse their surrounding reliably and efficiently, and enable the robots to accomplish their relevant missions.According to the targets, the following researches are carried out:(1) We propose a new complete coverage algorithm-improved Acar algorithm, we obtain the walking direction of the mobile robot according to the changes in its position and we can detect the critical points of the obstacles on the basis of the changes of its direction when it walks along the border of the obstacle. So, our method outperforms the Acar algorithm, which needs omni-directional sensors and can only deal with obstacles with smooth boundary and the slice can't parallel to the border of the obstacle.(2) The experiments on MORCS-1, which is made by the Intelligent Control Research Institute affiliated with Central South University, validate the feasibility and correctness of the algorithm. The experimental results are also analyzed in detail.(3) To overcome the limits led by the complete coverage carded out by a single robot, we propose a complete algorithm for multi-robots. The algorithm is based on the algorithm of Rekleitis I and was integrated with the new complete coverage algorithm-improved Acar algorithm. Both the walking direction changes of the two exploratory robots and the discontinuity of the interruption of the line-of-sight between them are used to detect the critical points of the obstacles. We design the walking rule for two exploratory robots to keep line-of-sight when walking along the border. We also design the rule to recover line-of-sight of the two exploratory robots when the visibility is broken and the rule to ensure the circular coverage of the neighboring areas of the obstacles, the algorithm can apply to both fast environmental exploration and efficient complete coverage.(4) Our complete coverage algorithms for a single robot and multi-robots don't take reeb graph to denote the connection relationship between free cell and the critical points to ensure complete coverage and lead robot go to uncovered cell. Instead, we present a new method for complete coverage, in which the neighboring cells of all the obstacles are covered in turn according to some rules. When the robot finds concave critical points and there exists no other critical points in both the left convex critical point setand the right convex critical point set, the complete coverage is accomplished. So, our method can overcome the common defect of both Acar algorithm and Rekleitis I algorithm, which a slice can't include multiple critical points.