Research On Indoor Mobile Robot's Path Planning And Localization Technology

Indoor mobile robot has been increasingly widely applied in human's production and life, in mall shopping guide, museum guide, hospital medical guide, exhibition greeting and commentary, security for the important places, restaurant service, etc. Navigation capacity is one of the key indexes to measure the performance of the indoor mobile robot, in order to complete the autonomous navigation of indoor mobile robot, two functions of autonomous location finding and path planning must be realized. RFID technology has the advantages such as fast recognition speed and large capacity of information and can obtain the centimeter-level location finding accuracy within milliseconds when it is applied in the indoor mobile robot, in addition, it adapts to the environment well and is cheap, thus, the autonomous location finding in this thesis adopts the technology based on RFID. This thesis adopts the path planning based on behavior, and the path planning based on behavior has the advantages such as rapid response, high flexibility, good robustness, strong adaptive capacity to environment and good expandability. There are two focuses in this thesis: first, a new improved location finding algorithm on the basis of the classic location finding algorithm is designed; second, the path planning algorithm is designed based on behavior fusion with fuzzy logic control. Main researches are:1. The indoor location finding system of the mobile robot is built by RFID technology, then, the defects of location finding error resulted from that it needs to take ?t time for RFID reader to obtain the wireless signal and the location finding program to calculate the position coordinates of the robot in the maximum likelihood estimation location finding algorithm is analyzed, BP neural network is utilized to conduct fitting for the complex mathematic relations between the three variables influencing the location finding error and the location finding error variable on x-axis and y-axis so as to conduct location finding error compensation, thus, the maximum likelihood estimation location finding algorithm is designed based on BP neural network error compensation, and such two location finding algorithms are analyzed and compared by experiment, the experiment result proves that the improved location finding algorithm is shown to be lower in error and much more accurate in location finding.2. The ultrasonic sensor system of the mobile robot in the experiment of this thesis is built. the target-trending, obstacle-avoiding and walking-along-verge-of-obstacles behaviors are designed with the fuzzy logic control, such three basic behaviors are tested and simulated, and the simulation result proves that the three basic behavioral controllers are shown to be effective.3. The behavioral weighing controller is designed by fuzzy logic control. in addition, the behavior fusion path planning vehicle is designed and simulated, and the simulation results prove that this path planning vehicle is shown to be effective and feasible in three different obstacle environments.4. The “mobile robot path planning algorithm simulation experiment platform” is developed by MATLAB software. this simulation platform can simulate the sensor system of the mobile robot, and this platform has been used to simulate the relevant algorithms involved in this thesis.5. The path planning experiment is conducted by using the designed real mobile robot system in environment of occurrence of complex obstacle and U-type obstacle respectively, and experiment results prove that this mobile robot is shown to be effective and feasible in using the location finding algorithm and behavior fusion path planning algorithm for navigation in such two obstacle environments.