The Design On The Filter Of INS/GPS Integrated Navigation System

The better data process method is needful to insure the INS/GPS integration system with higher position-precision and anti-jamming ability. In this paper, the research work focus on the data filter in the INS/GPS system. Selecting the appropriate integration mode and the reasonable filter method is significant and valueable to improve the performance, according to the principle of INS and GPS.Based on the analyzing the principle of INS/GPS, the mathematical model of integrated navigation system is established, and the four typical filters, i.e. the Kalman filter, the federated filter, the adaptive and the robust H_∞, filter, are analyzed and designed for the INS/GPS.First, the general Kalman filters are respectively designed for the position integration, velocity integration and position plus velocity integration; the simulations are made to analyzing the performance of the filters designed in the condition given. The federated Kalman filter is applied and realized to the system by means of the information fusion and distribution technologies, the failure detection and isolation are resolved when the GPS is in the problem. The adaptive Kalman filter is applied to the system, which can predict the uncertainty or unknown noise statistics. The filters just mentioned are all the Kalman filters in nature.Kalman filter requires the known statistical character of disturbance which can't be strictly satisfied in practice system usually. Robust H_∞, filter requires only the disturbance being energy bounded, other than known accurate statistic. Robust H_∞, filter has robustness both in uncertain noise and in uncertain system parameter, and it can be completed in the conditions above, this just supply a gap to Kalman filter. The performance of the robust H_∞filter is improved due to alterableγ.