With the expansion of robotic applications, sometimes the robotic system with local sensing can not meet the increasing complexity demand of tasks and environments. Network robot system may expand the sensing ability of robotic system by interacting with environments, which has become one of important research directions and received many attentions. This thesis focuses on target tracking and pursuit with network robot system. The main contents are as follows: Firstly, the research development of network robot system is presented. The multiple robots system, robot localization and wireless sensor network are introduced briefly. The background and structure of this thesis are also given. Secondly, a framework for network robot system is given. The framework is composed of three parts: console, server units and multi-robot system, and the server units are the provider of environment intelligence. These three parts can interact by WLAN, and the server units and console may also transfer information by LAN. Thirdly, the research on target localization of moving object based on fixed CCD camera in indoor environments is conducted. The BP neural network is used to map the information of target in image into the target’s position, which may achieve the rough localization. Also, target localization based on multi-camera is discussed. Fourthly, the target pursuit with network robot system is conducted. The console with electronic map and server unit with visual localization supply the task-related information to robots. On this basis, the robot makes decision with encoders, electronic compass and sonar information to achieve target pursuit. Fifthly, we design a target pursuit approach for multi-robot system in a multi-region workspace with the help of hybrid wireless sensor network that comprises static sensor nodes and mobile sensor nodes. The layout of hybrid wireless sensor network is presented. Considering static nodes, hollow zones and multi-region, the dynamic cluster and cooperative target tracking strategy are adopted to track the targets. On this basis, the user layer allocates the task based on targets information from the regions, and the robots execute the task with multi-region target pursuit strategy. Finally, the conclusions are given and future work is addressed.
修改评论