Research On Map Building Method Of Indoor Environment Based On Sonar Sensors

At present, the mobile robot has been widely used to the military, transportation, industrial and agricultural,medical and health services. It has been used to extend the human ability and to replace or reduce the heavy human labor. Robot environmental awareness is a prerequisite for the work of robots. Map building accounts for the problem of acquiring spatial models of physical environments through mobile robots. The problem of map building is generally regarded as one of the most important problems in the pursuit of building truly autonomous mobile robots. Despite significant progress in this area, it still reserves great challenges.To explain in hardware system and sonar sensor configuration of the mobile robot. The author has mainly studied on the related method based on grid map building of indoor environment of mobile robot. It takes the mobile robot MORCS-2 as experiment platform, which is developed by Central South University Intelligent Lab. The main contribution and work of this paper are described as follows:1 It takes Uniform Distribution Model as the sonar model. This paper proposes a probabilistic method of grid map building based on MORCS-2 in experiments ,and in the process of the local map transforming to the global map, it has two data integration using Bayesian rules on the global map, and it was achieved in multi-robot experiment.2 Although sonar sensors are fairly accurate in measuring distance from object to robot, the sonar sensors has noise, specular reflection, interference and some errors. In allusion to the indoor environment. Consider the sonar sensors in space-related and time-related degrees, to ensure that data continuity in time and space .The author puts forward a method to determine the noise and specular reflection based on tolerance function and ATM (The Arc-Transversal Median Algorithm) method for improving the accuracy of map building.3 In the feature extraction, The author describes the sonar data conversion to the Hough space for feature extraction. In view of uncertainty of sonar characteristics to feature extraction on line and point from derived sonar arc in the Hough space.4 In the Simultaneous Localization and Map-building, The author puts forward a method to SLAM based on radial collection,it has demonstrated for the method of the simultaneous Localization and Map-building in performance and demonstrate based on the characteristics of line and point. The authenticity of the map has improved in certain degree.