The Research And Design Of An Autonomous Following Robot

Autonomous following robot has the function of autonomous navigation and autonomous mobile,which can be used to follow the special populations, such as the elderly, the disabled and so on. And it can help them carry luggage. The robot designed in this paper has three modes:self following mode, user mode and administrator mode. Administrator mode's priority is higher than user mode. The self following mode is based on visual navigation.In user mode, users can control the robot through infrared remote controller. The controller has the function of starting or stoping the autonomous following mode, and it also can directly manipulate the robot's movement. In the administrator mode, the administrator can control the robot with app. Except for the permissions of the user mode,the administrator also can detect the robot's equipment.Firstly, the overall design is made,including robot mechanical structure design and control system design. The robot body has two driving wheel and a camera-rotary platform. Control system consists of upper computer system and lower control system, The upper computer system needs to complete the task of visual navigation. The lower computer system needs to complete the robot's mobile task.Secondly, the upper computer system is designed. The Target tracking algorithm is combines Meanshift and Kalman filtering theory, based on color information and depth information. Adding depth information in color histogram back projection map, which can eliminate the environmental interference. In the back projection map, using Meanshift algorithm to search for the target. Finally, the Kalman filter is added, which can select the optimal tracking window, and predict the target's position of the next frame. The setting of search area improves the reliability and reduces the amount of calculation. When algorithm is determined, the application program with multi thread and human-computer interaction interface is designed.Then, the lower computer system is designed. Firstly, the hardware circuit is designed. Secondly, The robot adopts artificial potential field method with a safe area to planning path. The algorithm is applicable to the situation that the robot has tracking task, as well as obstacle avoidance task in the same time. After the obstacle avoidance algorithm was determined, the software implementation of the algorithm is carried out. At the same time, the user mode and the administrator mode are designed and implemented.Finally, robot body is introduced. The tracking algorithm and obstacle avoidance algorithm is verified. The experimental results show that the tracking algorithm has strong robustness, and the path planning with obstacle avoidance algorithm achieves effective obstacle avoidance. The two algorithms are correct and effective, which enable the robot to complete the task of autonomous following.