The Research Of Campus Patrol Robot Localization And Obstacles Avoidance Technology

With the rapid development of service robotics, its application has gradually extended to all kinds of fields. The patrol robot is a new type of service robots, it’s produced under the environment that the supermarkets、logistics warehouses and other places with large flow of people and logistics are expanding at size and number with the social-economic development, the demand for security automation is increasing urgently. Using the patrol robot to complete fixed-time, fixed-point monitoring patrols or uninterruptible patrols has become technically possible, and the robot has the ability to accurately localization and to avoid the obstacles that may exist in the environment, is the premise and basis of the robot to complete patrol tasks.The paper aims to the localization and obstacle avoidance technology of the mobile robot in the campus environment and has achieved certain results, the specific contents are as follows:(1) In view of the wide-range and simple-structure characteristics of campus environment, and combined with the information transmission function of RFID sensor, the paper proposed a new large range of topology environment learning and modeling method. According to the context information needs of campus environment, designed the RFID tag format to guide the robot for environmental learning, and on behalf of the environmental whole structure with the distribution of RFID tags, the navigation tasks can be passed by the tag content, thus built the environment model.(2) Aiming to the problem that the laser data exists noises, the paper designed a dynamic adaptive median filtering method, using the relationship on the time and space of the laser data, to remove the noises, make the laser data smooth.(3)According to the characteristics of the campus road, this paper use the laser sensor to distinguish road area by the height of road curb, further to complete the robot localization, and dew to the problem that one single sensor cannot be reliably localization, here proposed a angle information fusion method connected the laser sensor with the AHRS based on the Extended Kalman Filter, for which greatly improved the accuracy and reliability of the robot localization on the straightway. In the intersection area, using RFID、laser、AHRS and odometer cooperate with each other to make up for deficiencies, so that ensured the robot localization accuracy of the intersection area.(4) To ensure the security of robot patrols, a new angle potential field method was proposed in this paper, based on the summary of commonly used avoidance strategies, the method transformed the potential field force around the robot environment to the angle information, after a comprehensive analysis about access of all angles, lead to the best selection of robot direction. Furthermore, aiming to the defect of traditional detection methods for the obstacle, install the laser sensor in tilt downward orientation to detect an inclined plane, taking into account the front and the lower area of laser sensor on the robot to get better detection effect.