An Underwater Autonomous Robot Based on Multi-sensor Data Fusion

A kind of new autonomous robot can excavate hole in the mud underwater, which is called move-in-mud robot. It can be used in sunken wreck salvage to improve the efficienty of excavating the hole. In this paper, analyse the working principle of move-in-mud robot according to planned trajectory. The kinematics modelling of move-in-mud robot is established. Data fusion can combine multi-sources information from one object, and form a more precise and complete estimate than the sole sensor. The paper analyse several methods of multi-sensor data fusion such as Bayesian estimate, Kalman filter, Dempster-Shafer evidence theory etc, and discuss general structure of multi-sensor data fusion. Location system of move-in-mud robot is designed in the paper and location principle of move-in-mud robot is analyzed. Kalman filter is applied to fuse redundant information of the robot. The data fusion method is simulated in location system of move-in-mud robot, the effect of simulation of data fusion shows the data fusion method can improve location accuracy