Researches On Path Programming Of Maze Robot Using Combined Depth-Width Algorithm

With the progress of worldwide science and technology, the robot is toincrease geometrically way into all areas of human life, from the unmannedhelicopter to the AGV (Automatic Guided Vehicle) on the highway, from theindustrial production line robot arm to a wide range of entertainment robots andteaching robots, from the low-level intelligence to senior intelligence, from thecomputer simulation to the actual experimental system, the development ofrobotics is gaining more and more momentum. However, how the autonomousmobile robot in unknown and complex environment, self-planning from thebeginning to the end of the path, and avoid the obstacle has always been acomplex technical problems. Therefore, studies for mobile robot path planningalgorithm have certain theoretical and engineering significance. And the mazerobot is a typical application of mobile robots, path planning algorithm is agood test platform.This paper studies the history and current situation of mobile robots basedon the focus of the mobile robot obstacle avoidance and route selection optionscommonly used in grid planning method, potential field, genetic algorithms andother methods which were analyzed and compared. Example for a miniaturerobot by myself, this paper discusses the infrared-based extraction method ofthe unknown environment characteristics and the principles about selecting safedistance of obstacle avoidance, and we give a strategy based on broad and deepintegration algorithm for mobile robot obstacle avoidance, describe the optimalpath planning application based on the broad and deep integration algorithm inthe robotic maze map.In addition, this paper has also built a communications platform for mobilerobots under the Visual C++, and the simulations have proved the effectivenessand accuracy of the principle and algorithm used. Text is divided into sixchapters:The first chapter introduces the research background, significance andstatus of domestic and international robotics development, describes thecontents and research ideas of this study.The second chapter describes several mobile robot path planning methods. The third chapter outlines the hardware system architecture and controlprinciple of the maze robot, and emphasis the important relationship betweenthe robot’s initial location and the path programming by which the mobile robotachieve to avoid obstacles.Chapter IV introduces the path programming algorithm of the mobile robotand its realization when the map is known. And the classical maze algorithms,depth-first search, breadth-first search and genetic algorithm, are integrated.An improved maze search method with memory capacity is provided which iscombined with depth-first algorithm and breadth-first algorithm. In addition,the principle of infrared sensors also is introduced with the basic obstacleavoidance action.In Chapter V, the maze robot software has been designed in accordancewith the modular idea. The robot’s each functional module is designed inaccordance with the requirements of part programming, and the eightconditions the robot may encounter are detailly analyzed. With the used roadmemory, dead-end memory and steering memory, the appropriate design isprovided in accordance with the norms of behavior-based responses.In Chapter VI, the actual movement path of the robot on the maze map isprovided, based on the combined depth-width algorithm path programming.Severaltypical cases of robot motion characteristics of Infrared sensor-based obstacleavoidance are analyzed, and the used algorithm is proved its correctness and validity.The last chapter sums up path programming algorithms and puts forwardresearch tasks for further study.