Kalman filtering method in deep integrated navigation system

Abstract

The invention relates to a Kalman filtering method in deep integrated navigation system, comprising the following steps: (1) establishing filtering state equation and pseudo range / pseudo range based measurement equation; (2) selecting the first time period of a time-varying system; (3) calculating the observability matrix of current time period; (4) calculating current SOM matrix; (5) calculating the appearance measurement of the current time period; (6) calculating the singular value of the observability matrix in current time period; (7) figuring out state variable X(0) corresponding to each singular value, performing observability analysis and observability calculation; (8) circularly executing steps (3) to (8) until completing all the time periods of analysis; (9) performing improvement design and parameter selection of Kalman filter according to results of observability and observable degree analysis of state variable X(0). In the method, the filtering estimation result and navigation precision of the integrated navigation system are improved by performing observability and observable degree analysis to the integrated navigation system and performing improvement design to the Kalman filter according to the analysis result.

Description

technical field [0001] The invention belongs to the technical field of navigation, in particular to a Kalman filter processing method in a deep integrated navigation system. Background technique [0002] Global Positioning System (GPS) and Inertial Navigation System (INS) have strong functional complementarities, and the deep integrated navigation system composed of the two has been more and more widely used. The Kalman filter processing method is the most commonly used algorithm for data processing in integrated navigation systems. The effectiveness of the Kalman filter depends on the observability of the system. For an incompletely observable system, it is necessary not only to know which state variables are observable, but also to know the degree of observability of each state variable. Although the existing Kalman filter can qualitatively analyze whether all the system states can be observed by applying the PWCS method before performing Kalman filter processing on the t...

AI Technical Summary

Problems solved by technology

Although the existing Kalman filter can qualitatively analyze whether all the system states can be observed by applying the PWCS method before performing Kalman filter processing on the time-varying system, or judge which system states (or linear combinations of states) are possible Observation, which system states are unobservable, but it cannot quantitatively give the degree of observability of a certain system state at different time periods, the incomplete observability of the system makes the integrated navigation system unable to obtain higher positioning accuracy

Images