Research On Tunnel Inspection Robot Localization Method Based On WSN

The application of robots for tunnel inspection has the advantage of low cost, high efficiency and high factor of safety, and in the future this kind of inspection method will be a trend. The funcion of inspection Robot is to monitor the status of cable in the tunnel in the way of using a variety of sensors. If there is something wrong with the cable in the tunnel, the robot needs to position the faulted cable, according to this positioning result the technical staff can find the location of faulted cable, and then repair it. Therefore, accurate positioning system has great significance on the patrol robot.How to use the technology based on WSN to position patrol robot is the point of this research. Through the analysis of the researching background and the existing positioning technology and algorithms, this paper proposed a more precise localization method.This dissertation mainly consists of four sections for design and improvement of tunnel positioning method, which includes appropriate selection of distance-measuring technology, adopting reasonable calculation method, programming for integrated layout of the location system and designing of working process of location nodes. Combined with experimental data, the model which indicates attenuation of radio signal is revised. Because of the fact that the radio signal is severely affected by the stochastic disturbance, clipping and filtering algorithm according to Maximum Likelihood Estimation(MLE), and the arrangement form of WSN nodes in tunnel positioning system via research in the field of Zigbee communication technology and working process of different functional nodes in information collection procedure are proposed. The programming of Gateway reference and blind nodes are designed via research and learning of Z-stack and comprehending of discrete working process of each node. In the process of model revising and filtering algorithm proposing, CC2430chip which supports zigbee technology is put into use in collection of signal intensity in the real tunnel environment. After completing the model revising, the real experiment was conducted to test the the accuracy of model and the effetiveness of filtering algorithm.