Research Of Relative Locating Algorithm In Swarm Robotics System For Coal Mine Rescuing

Coal production has always been a high-risky work, and because of poorconditions of the coal mining system, accidents such as gas explosion, water inflowhave happened frequently in recent years.While mine disaster occur, the scene of thecoal mine after accident is a tricky and dangerous environment. The difficulty ofrescue work is increased by the potential security risks. To meet that challenge,people try to make the robot used in the mine accident rescue. But in the actualapplication people found that the individual rescue robots, because of some adversefactors, such as large equipment it carriage, size, high weight, are not do very goodon its viability.To provide kinds of information and complete the various tasks, determine itsposition is a foundation, this article is study on the robot relative positioningalgorithm based on the study of the early stage.（1）Based on the extended kalman filtering algorithm, each individual robotsensor data fusion itself, estimate their position. Then obtain the coordinates of theadjacent robot by the mutual communication and calculate the distance betweenadjacent nodes.（2）With the aid of wireless sensor networks, based on the individual robotmapping relationship and network, and by using joint distance measurement methodbased on RSSI/LQI, measured by the distance to the robot location data obtained forcorrection and fusion.（3） The trilateral localization algorithm is used to determine the final locationestimation of each robot.The simulation and experimental results shows that:(1) Compared withsingle-sensor methods, this method exploits the differences in uncertainty betweenultrasonic-uensor and odometer systems, and consequently improves theself-localization precision of mobile robot.(2) The ranging method based onRSSI/LQI improve the ranging accuracy. To sum up, the simulation and experimentverify the effectiveness of the relative positioning algorithm.