Research on multi-sensor fusion of underwater robot navigation system

A general design method of federated Kalman filter based on multi-sensor fusion is introduced, and applied in integrated navigation system including SINS, electronic compass and depth gauge for underwater robot integrated navigation system. We adopts federal filtering NR structure of integrated navigation which establishes a SINS / electronic compass sub-filter, SINS / depth gauge sub-filter and main filter. Simulation results show that the federal filter is feasible for integrated navigation. It ensures not only fault tolerance and real-time at the same time, but also meets the system accuracy.