The Reseaching On Active Probe Mobile Robot Based On Embedded Linux

The remote control and active probe of mobile robots based on network have increasingly got more attension than ever, with the rapid development of network and embedded technologies. The remote control of robots can make the operator keep away from the danderous scene to avoid life injury, which can be applied in out space explore, underwater operation and nuclear industry and many other fields. Active probe of the mobile robot is one important part of the technique of mobile robots. The researching on it has significant meaning in theories and engineering applications. After researching on main methods about path planning, this paper provides the different method of planning.Embeded Linux, with many advantages such as open source, easy to customize, multitask and multithreading. In this paper, a mobile robot with a Web server and a network control server integrated into S3C2410, and a wireless linkage to network is afforded, which meet the needs of mobile robots for minitupe, parallel, moving and group cooperation. Remote operators are able to program themselves by network control server to send orders to the robot, and also look over the states and the images of the robot so as to monitor its moving status.The main study work and results are just as followed:(1) Construct the system platform based on embedded Linux, which is composed of bootloader, Linux kernel porting, USB WLAN card driver and USB camera driver.(2) Realize the movement control and sensor information collection of the mobile robot including motor control inner sensors information collection, image information collection and data acquisition of the wireless sensor network.(3) The connection between the robot bottom control and upper computer is actualize by using USB and RS232 interface in order to improve the open and extendable ability of the system.(4) Introduce risk-degree of collision which is defined using distance and relative orientation between the mobile robot and obstacles, and then plans the path using the method based on risk-degree of collision and fuzzy logic control. The obstacle-avoidance simulation is carried out; the results indicate that the method is feasible.