Research On Autonomous Navigation Of Mobile Robot Under Indoor Circumstance

In Recent years, with the further development of mobile robot towards direction of minimization and intelligence, the problems including path planning and autonomous navigation have been pay more and more attentions and become a hotspot research sector of mobile robot. In this thesis, the problem of path navigation is studied intensively under the condition of corridor and unknown indoor environment on the platform of Pioneer3-AT mobile robot. Moreover, the simulation verifications and basic experimental exploration are conducted. The main work and achievements of this thesis are shown as follows:1. Mobile robot communication platform is designed by VC++6.0 based on the C/S structure. The platform overcomes the bottleneck of mobile robot logging, which has promising prospect of engineering application.2. In order to solve the problem of navigation for mobile robot under corridor environments, one kind of visual navigation algorithm based on parallels is proposed. During the process of vision navigation, the two corridor baselines are viewed as a group of parallels. The feature points of the projection straight line of corridor baselines in the visual plane are used to control the angle velocity and the linear velocity of mobile robot. It makes the autonomous move of mobile robot come true. The validity and correctness of controlling algorithm are verified by the results of Matlab 7.0 simulation platform.3. In order to solve the problem of navigation for mobile robot under unknown environments, an autonomous navigation method based on Taboo searching is presented. The analyzed data collected by laser radar is used to identify the sub-goal. Current optimize sub-goal is generated by Taboo searching algorithm that is evaluated by heuristic function, and furthermore it makes use of fuzzy logic controlling method to trace the sub-goal. So the free-obstacle and optimization path is achieved by gradual way. The algorithm has the advantages of smoother path and shorter path-length. At the same time it raises the foreseability of navigation movement and overcomes the shortcoming of repetition visiting and local minimum. Simulation results verify the correctness and effectiveness of the proposed method. Finally, the obtained results are summarized and future is addressed.