Path Planning For An Indoor Robot With Optimized Floyd Algorithm-based

Robotics is a high technology combining with computer science, control theory, information technology, sensor technology, artificial intelligence, bionics and other disciplines, and is a national high-tech level and an important indicator of industrial automation. Path planning is one of the most fundamental aspects in autonomous navigation of mobile robot, it reflects the interaction ability between the robot and the environment, and is a security requirement for safe driving of mobile robots. To improve the level of intelligence robot, research on path planning has practical significance and bright prospects. The main work of this paper was followed as three points.(1) To achieve the indoor robot path planning map, OpenGL(Open Graphics Library) modeling technique was used to implement the indoor(classroom, laboratory, office, house) path planning map. To generate the path planning map, the passage of a robot based on elevation and headroom of each obstacle was analyzed, Passable area for robot was used as the path planning map. The generated path planning map is simple and reliable, and can be used for path planning.(2) To reduce the computation time of Floyd algorithm, filtering vertices improved method was used to filter the narrow unreachable area from the map, which was able to reduce the number of input vertices for Floyd algorithm. Finally the running time of the algorithm was effectively reduced. The result of test showed that path planning takes large less operation time than the original algorithm before filtering vertices. Vertex count and inverse proportion improved method were used to further optimize the algorithm. Test results showed that the average steps of path planning were significantly decreased. Meanwhile, time-consuming is about 72% of the original algorithm. The effect of improved Floyd path planning algorithm and Dijkstra algorithm were compared, and test results showed that the performance of improved Floyd path planning algorithm is superior to Dijkstra algorithm in all aspects.(3) Through the analysis of software functional requirements, the logical design of interactive interface, path planning module, finally based on Windows 7 system, VC++ programming technology was used to implement the "indoor robot environment map simulation and path planning" software.