Abstract:

On the basis of mathematical description for H ∞ estimation problems, an iterative equation to H ∞ sub-optimal filtering algorithm is an alyzed. The relationship between H ∞ filtering algorithm and traditional Kalman filter is discussed, and at the same time, some performance indexes of these two methods, e.g. accuracy, robustness, are compared by using a GPS/SINS integrated system. The simulation results indicate that Kalman filter has better accuracy in the ideal condition. But H ∞ filtering algorithm has better robustness and accuracy to model uncertainties and statistics characteristics of disturbance variations, which overcomes some limitations of Kalman filter, e.g., stability, robustness, and so on.

Key words: H &infin, filter ing;Kalman filter ing;ro bustness;GPS/ SINS;integ rat ed navigat ion