Study And Desigh For Claeaning Robot Navigation Based On Recognition And Tracking Of Boundary

With the development of road cleaning industry, the intelligent cleaning robot, as one of main application fields of service robots, has become a hot researched topic. The paper, based on the visual sensor, laser sensor and other equipments, studied the road boundary identification, path planning and tracking control in the structured road environment.Firstly, the boundary identification of the structured road using vision sensor is studied, and the visual model of the robot is established. The process of the boundary identification of the road image is: grayscale, filtering, threshold segmentation, edge extraction. Fitted boundary line using Hough transform after image processing, Obtain the angle and distance information of the cleaning robot relative to the road boundary, building the foundation of path planning and tracking control of the robot.Secondly, studied the path planning of the cleaning robot in a structured road environment, established the kinematic model of the robot and designed the method of navigation point path planning. As well, this paper researched the aspect of obstacle avoidance in path planning by laser sensor, achieved simulation experiment.Thirdly, analyzed the model of two common styles of mobile robot, its representations in space and simplified motion model, with the characteristics of the cleaning robot, did further analysis and research of the robot with four-wheels. The paper designed anthropomorphic fuzzy control method to achieve the tracking control of road boundary, achieved simulation experiment using MATLAB and analyzed the results.Finally, the paper showed the design of cleaning device for the robot and the transformation of the robot moving body, designed the software and hardware of its control system, proved the feasibility of the design in this paper by experiment.