Research on State Estimation Algorithm of Maze Robot Based on Multi-Sensor Fusion in Complex Environment

In order to further improve the efficiency of the maze robot searching the maze and the accuracy of state information, this paper proposes a state estimation algorithm for maze robot based on multi-sensor fusion, Firstly, the algorithm uses an adaptive weighted batch estimation algorithm to measure the distance of the infrared sensors Fusion, get more accurate distance information. Secondly, the accurate distance information is combined with the sensory data of the gyroscope and the encoder, and the extended Kalman filter is used for fusion to obtain the position, angle, speed, angular velocity and other information of the maze robot. Finally, the experimental results show that the algorithm can effectively improve the positioning accuracy and robustness of the maze robot, and accurately feed back the state information of the maze robot. The problem of low positioning accuracy caused by the large error of the state information of the maze robot is solved.