Research On Indoor Navigation Algorithm And Autonomous Charging System Of Service Robot In Hospital

The hospital service robot belongs to the category of service robot. Infectious diseases, the aged tendency of population and industrial demand mainly contribute to the study of its related technologies gradually. Based on the company’s demand, this paper mainly research on the indoor navigation, positioning and self-charging methods of the hospital se rvice robot. According to summarize the shortcomings of the previous studies, this paper decides to adopt a more easy transplant operating system ’ROS’ to program the robot’s control system to avoid repetitive work.The ROS navigation algorithm is improved first. Repair items of observation model in the existing self- localization method is proposed to increase the probability of the robot self- positioning in dynamic environment. Derivation and related experiments can prove the improved method ha s more robust self- localization. Also linear navigation and BUG obstacle avoidance scheme is adopted to instead of the original COSTMAP and local speed control method, the simulation results demonstrate new method has a shorter distance and more real- time control. Then, based on the existing robot platform, this paper designs the whole robot self-charging system. A combination of mechanical structure and software control is used to ensure that the robot can achieve self- charging in a high success rate. Since the implementation of the charging process using the mileage meter coordinates as its global coordinate system, the accuracy of the odometer data has a special significance for the successful completion of the self-charging. Kalman filter is used to fuse the information of the current popular optical flow odometer and the information of the robot’s encoder odometer to improve the accuracy of the robot’s odometer. Through the formula derivation and the actual experiment, it can be concluded that this method is feasib le. But due to its easily disturbed, this method has some difficulty to realize in engineering implementation. Finally, using the real robot platform, this paper gradually verifies the feasibility of the local navigation modification method and self charging system which has proposed in the front chatter, and experiments the robot prototype during the simulation hospital scene. The reliability experiment vertifie s the method proposed by this paper has some real practicability.