Multi-sensor Data Fusion On Obstacle Information Detection Of High Voltage Cable Deicing Robot

Multi-sensor for obstacle information detection and data fusion is one of the key techniques of autonomous deicing robot on high voltage transmission line. The obstacles such as counterweight, insulator, strain clamp and suspension on high voltage transmission line should be effectively measured for deicing robot to intelligently and autonomously avoid and cross obstacles. According to the characteristics and distribution of obstacles on 220KV transmission line, a series of methods for obstacle information detection and data fusion based on multi-sensor system are put forward in this thesis. The main studies in this paper are as follows:Firstly, a detailed analysis on key issues of methods for obstacle information detection and data fusion of autonomous deicing robot are put forward in this thesis, and a multi-sensor obstacle data fusion system is developed.Secondly, the model for multi-sensor obstacle data fusion system of robot on transmission line is built.The obstacle detection system and equation of system state are designed.Thirdly, a method of obstacle data fusion based on neural network is researched. Methods of feature extraction and data association for sensor data processing are given. A appropriate neural network data fusion model and its BP learning algorithm with adaptive convergence rate are proposed. good convergence results and a series of accurate data on Multi-sensor obstacle fusion are obtained.At last, another way of information fusion based on extended Kalman filterare(EKF) also is studied in this paper. Aimed at the nonlinear multi-sensor system, the model and state equation of obstacle data fusion based on EKF are put forward to receive new data filter and forecast process method. Multi-sensor obstacle fusion data with higher integration accuracy and more rapid convergence than neural network are obtained.