Localization And Real-time Obstacle Avoidance Research For Mobile Robot In Unknown Environment

As one of the key technologies, navigation technology has been paying attention in autonomous mobile robot research field. By means of all kind of sensors to obtain dynamic information of the mobile robot, autonomously moving can be implemented for mobile robot in obstacle existence environment. Obstacle avoidance and localization are two elementary problems for mobile robot navigation. Some of effective navigation control strategies for unknown environment have been presented. However, comparing with the navigation in known environment, to give control scheme in unknown environment is more difficult. In this paper, the studying topics are focused on localization and real-time dynamic obstacle avoidance for mobile robot in unknown environment. Concerned with dynamic obstacle's existence and the boundless cumulated error in odometer localization, an available strategy is presented.Firstly, the environment information acquisition system of mobile robot is introduced in detail. The motion control for mobile robot and the main sensors' model are discussed in chapter 2. And the sensors' error model of laser radar and odometer are established, also, the analysis of the infection and resource of sensor error is given.For solving obstacle avoidance in dynamic and time-variant environment, an angle search strategy based on apperceive-behavior theory is presented in chapter 3. This strategy can be used to implement the real-time obstacle avoidance in unknown environment. According to the actual circumstance of obstacle avoidance motion control, this chapter gave the explanation of the obstacle information abstraction, rotating direction selection of mobile robot and the choice of safe-distance, etc..Based on the presented analysis results of self-localization methods, the general odometer is improved in chapter 4. The direction information obtained by digital compass is used to modify the odometer localization model, so that the direction-lost problem in conventional localization is resolved. And the localization accuracy is improved obviously by checking and compensating the system error and wheel-slippage...