Filter gain dynamic adjustment method for inhibiting GNSS (global navigation satellite system) information abnormality

Abstract

The invention discloses a filter gain dynamic adjustment method for inhibiting GNSS (global navigation satellite system) information abnormality. On the basis of a standard Kalman filter algorithm, an index quantity conforming to chi<2> distribution is constructed through an innovation vector and a chi<2> test is performed, and abnormal observation values are detected. According to parameters of the chi<2> test, a scaling factor is constructed, and the scaling factor is used for lowering down a corresponding filter gain matrix, so that a purpose of inhibiting influence of the abnormal observation values on a system is achieved. The filter gain dynamic adjustment method does not affect normal noise, mistaken detection and missing detection on the abnormal values are avoided, effects on single-moment abnormal values and continuous abnormal values are relatively efficient, the performance of the system can be improved by more than 90%, the acting time of the abnormal values is reduced by about 5 times, and the performance of an integrated navigation system is greatly improved.

Description

technical field [0001] The invention relates to an INS / GNSS integrated navigation system, in particular to a dynamic adjustment method of filter gain for suppressing abnormality of GNSS information. Background technique [0002] In the inertial navigation system (Inertial navigation system, INS) / global navigation satellite system (Global navigation satellite system, GNSS) integrated navigation system, the update rate of INS data information is high, the noise is low, and it has high accuracy in a short time, but its system principle determines In addition to the fact that navigation errors will accumulate over time, GNSS errors have nothing to do with time, and can provide high-precision navigation information around the clock and globally, but its information update rate is low. The INS / GNSS system performs information fusion through the filtering method, which can realize complementary advantages and obtain more accurate navigation results. [0003] Kalman filter (Kalma...

AI Technical Summary

Problems solved by technology

Particle filter can be applied to nonlinear systems and has good tolerance performance, but it has a large amount of calculation and poor real-time performance, so it is rarely used in the field of integrated navigation; adaptive filtering methods include LMS and RLS methods, and their tracking performance is poor; h ∞ Filtering analyzes the error information based on the filter performance index, the energy is the smallest, and it sacrifices the average estimation accuracy in exchange for the tolerance performance

Images