Multi-Robot Cooperative Control Based On Wireless Sensor Network

Multi Wheeled Mobile Robots (WRMs), Wireless Sensor Network (WSN) andNetworked Control System (NCS), are now the hot academic research issues. The controlproblems of the WRMs could split to stabilization and trajectory tracking, but first of all, thelocalization and navigation of the WRMs should be done. WSN is a perfect way for thispurpose, which could not only measured and localized but also implement the tele-operationcontrol of the robot control. On the other side, WSN based systems have been applied to dailylife service, health care, resource exploration etc, especially, the WSN application systemwork well in transportation, logistics, environment monitoring and so on more often.In WSN research area, many of the WMRs are autonomous, which are equipped withseveral kinds of sensor to detect the position and orientation by itself, such as ultrasonic, laser,infrared and visual and just work as a mobile target. WSN is independent from the WMRs.The major research focus on the networking model, topology control, routing, dataforwarding and cross-layer design technique, time synchronization, object tracking, etc.However, WRMs could provide more value by work as a controlled plant via network.Multi-Robots run in the sensor network, the sensor nodes measure and localize for each oneand then send the localization data back, which would be used as the control input of therobots controller design. The followers could now learn the position and orientation of theleader, so that to adjust themselves to move synchronously. And this, the real network controlsystem, could help in the test of the advanced network control system theory.This paper studies the implementation for a distributed network feedback control systemand the trajectory tracking control algorithm for a non-autonomous WMR. In the first chapter,the research history and relative results of WRMs and WSN has been introduced respectively.And then, motion characteristic of a special WRM come out from lots of test with the generalkinematic model. The implemented networked control system consists of a wireless sensornetwork (WSN) and a monitoring base station for measurement and feedback, and threenon-autonomous WMRs as controlled plant. Two followers could cope with the leader. Forthe better performance, a non-linear controller and a LQG controller have been studied andthe simulation results show the system can effectively track a reference trajectory under thecontroller.