Research On The Technology Of Autonomous Navigation Of Seeing-eye Robot

Currently because the total number of the visually handicapped in the world is huge and keeps growing, the seeing-eye robot is highly anticipated compared with other guiding device. But most seeing-eye robots are limited to the specific environment or single space and difficult to adapt to the change of environment so the seeing-eye robots are lack of practical value which can't meet the needs of society.In order to expand the environmental range of the adaptation of the seeing-eye robot,this paper proposes an effective method of the seeing-eye robot based on autonomous navigation to break the space limitation and guide the visually impaired in different environments. First of all, design the principle framework of autonomous navigation system of the seeing-eye robot in general. Then create the two-dimensional map of the environment and the seeing-eye robot localize itself at the same time according to the position estimation and observational features of Simultaneous Localization and Mapping method. After that, take the A* heuristic search algorithm to complete the static global path planning in the known map without collision and combine the artificial potential field algorithm which based on relative velocity to avoid the dynamic obstacles that the seeing-eye robot encounter when walking by the planned global path. Afterwards, construct the software control system of the autonomous navigation of the seeing-eye robot in Robot Operating System and the function nodes follow the ROS conventions to communicate with each other, which make the function module more orderly and improve the efficiency of the control system execution. Different kinds of code were used to simulate the A*algorithm and the artificial potential field method, and the function nodes graph are drawn in the ROS framework. Finally, the experiments were carried out in three representative circumstances indoor and outdoor including office, long corridor, rest area, and the results of experiments proved that, compared with other method, with this method the guide robot can navigate by itself, no longer dependent on other devices. So it can break through the constraints of the environment and space working in different environment flexible which wide the application range of the seeing-eye robot and come with high precision comparatively. The error of characteristic estimation and the error of location estimation in the process of map building and the trajectory error in the process of path planning all float within a small range. In contrast, this method can satisfy the needs of blind guiding better.In addition, study and summary the unfavorable factors to autonomous navigation of the seeing-eye robot, put forward the corresponding strategies to provide the basis for the development of the guide robot.The autonomous navigation method of seeing-eye robot designed in this paper can create different environmental maps adapted to varous environments without any extermal facilities which also can escape the dynamic obstacles. The corresponding experimental results demonstrated that the seeing-eye robot can build 2D map more accurate in a variety of environments and the trajectory of the robot is consistent with the planned path by this method that improved the precision of the guide robot. Besides this method can adapt to the change of different environments flexiblely and with which the seeing-eye robot is generally suitable for indoor and outdoor areas where the visually handicapped have activities in daily. Therefore the method proposed by this study is more effective, practical and superior for guiding the blind.