Ccd And Sonar Sensor Information Fusion-based Mobile Robot Obstacle Avoidance

Multisensor information fusion is a developing practical technology around theworld with a wide range of application areas. It provides a reliable means in the areas ofinformation processing and decision support and has a very high research value as wellas application potential.With the platform of AS-R mobile robot, the application of multisensor informationfusion in the mobile robot obstacle avoidance was studied principally. In this thesis, themain contents are as follows:1.The techniques of information fusion are systematically studied. Then, theapplication of multisensor information fusion in robot is analyzed.2.By experiment, the performance of sensors is tested. In addition, the reasons forerror caused by ultrasonic sensors are analyzed.3.Based on the imaging principle of objects, a bottom-up method for detecting theedges of obstacles based on a trapezoidal region is proposed. The method can effectivelyeliminate the impact of interference information, and then the edges of obstacles aredetected successfully.4.The physics model of the mobile robot is set up. Moreover, the mobile robotobstacle avoidance multisensor system that is made of CCD camera and ultrasonicsensors is researched. In allusion to given environment for mobile robot, the strategy ofavoiding obstacle is proposed.5.The fuzzy neural network information fusion based on the T-S model is adopted inmobile robot obstacle avoidance. Finally, the simulation experiments verify that theapproach proposed above is really feasible and efficient.