The Research Of The Methods In Autonomous Mobile Robot Mapping

In the research of the autonomous mobile robot, the purpose is to realize that the robot can efficiently achieve its task by itself, especially in the unknown environment. The precondition of the moving and working autonomously is that the robot is able to know the work environment in detail, and make path planning and task planning in order to achieve the task successfully.In the paper, we research the apperceiving and comprehending of the robot to the work environment, in order to establish the image of the environment in the robot's brain and describe it in an easily understanding way. The main outcome is as follows:一,Aiming at the indoor cleaning robot, we present a method of robot autonomously mapping. It is based on grid map, and subareas it and gains the free areas order in the width-first search and depth-first coverage (WSDC) way, navigate the robot in the exploration of the environment. The experiment result indicates the feasibility that the subareaing and coverage in using grid map, using the WSDC way can reduce the moving distance and improve the work efficiency of the robot.二,Based on the self-organizing map (SOM), we construct a new growing self-organizing map(GSOM). Its growing characteristic is that its neuron number is not fixed, it can change according to the distribution of the samples in order to carry out its structure self-organization and self-study. The experiment indicates the GSOM model is able to recognize the structure of the input space in the self-growing of the map structure, and classify and cluster the input sample exactly.三,We represent a new method of the autonomous mobile robot mapping. It is based on the growing characteristic of the GSOM model to make true the image shaping gradually in the robot brain. It can find the optimal SOM neuron number for describing the environment, and use less SOM neurons distributing to show the environment structure, establish the topologic map describing more exactly the environment. The experiment indicates the topologic map basing on the GSOM can effectively gain the number of the topologic nodes and carry out the data compressed and distribution identified to make it convenient in the robot path planning.This work is supported by the National Nature Science Foundation of China(60375017),Funding Project for Academic Human Resources Development in Institutions of Higher Learning Under the Jurisdiction of Beijing Municipality (PHR (IHLB))and the Special Founding for Doctoral Program (20050005002),and the related work has been approved for the Journal of Robot Technique and Application and Journal of System Simulation(EI).