Study On Uwb Wireless Localization Algorithm And Navigation Contorl System Of Inspection Robot

For the special requirements and complex environment in substation, ultra-wide-band(UWB) –based inspection robot system is designed. The localization and navigation accuracies are important for inspection robot system. In order to improve the accuracy of patrol robot, robot system is set up. Algorithms including ranging filtering algorithms, localization algorithms and navigation algorithms are studied in this thesis.Basic requirements for substation patrol robot system are introduced, and then the subsystems robot systems including network systems, motion control systems, detection systems, and positioning and navigation systems are briefly described respectively.Research work focuses on NLOS(non-light-of-sight) wireless localization methods. The improved algorithms based on LLS(linear least squares) method and KNN-based algorithm are introduced. Simulation tests are carried out on Matlab platform to verify the performances of those algorithms. The TOA-based ranging system model in dynamic system has been modified and the error models have been established in this thesis. It makes more clear to find out the error sources and more easy to deal with the ranging errors in different occasions. At last, STF(self-adaptive thresholds filter) algorithm with rapid noise reduction and tracking(RNRT), and real-time noise elimination(RTNE) algorithms, have been proposed to deal with the noise of PRM in the dynamic system as well as in the static system.Mechanical structures of inspection robot are analyzed. The kinematics and dynamic modeling of inspection robot are established in this thesis. Then, we design motion control system to realize inspection work. Stm32-based CanOpen master station is designed to provide a stable control system. Finally, a fuzzy PID algorithm for inspection robot navigation is proposed.A test platform is built for the UWB wireless ranging experiment on some specific conditions, such as indoors and outdoors, LOS(line-of-sight) and NLOS conditions. Then, parameters of algorithms are tuned. For different algorithms, localization accuracy is analyzed. Moreover, tests are carried out to analyze the navigation accuracy of inspection robot. Finally, the comparison between UWB and laser in ranging and localization is performed.