Combined navigation method based on Kalman filtering estimation variance constraint

Abstract

The invention provides a combined navigation method based on Kalman filtering estimation variance constraint. The combined navigation method based on Kalman filtering estimation variance constraint comprises the following steps: performing coarse alignment on an inertial navigation system; carrying out inertial navigation solution according to the angular rate of the gyroscope and the specific force of the accelerometer, and updating the attitude quaternion, the position and the speed; establishing a system error model; selecting Kalman filtering quantity measurement, and setting an initial value of a Kalman filter; and performing Kalman filtering calculation to obtain integrated navigation data until the combined navigation is finished, and constraining an error covariance value of a Kalman filter according to an error of the inertial device in the Kalman filtering calculation process. By applying the technical scheme provided by the invention, the technical problems that the navigation precision is limited due to the fact that a filter cannot track in an integrated navigation process in the prior art and filtering divergence exists in long-time integrated navigation can be solved.

Description

technical field [0001] The invention relates to the technical field of inertial navigation, in particular to an integrated navigation method based on Kalman filter estimation variance constraints. Background technique [0002] During the use of the integrated navigation system, due to the existence of gravity on the celestial channel, when the Kalman filter is estimated, there is often a slowly increasing deviation between the altitude channel and the satellite altitude when the system performs long-term integrated navigation , which has a great impact on high-precision fixed-point strikes or precision drone hovering. [0003] According to the analysis and summary of the Kalman filtering process, it is found that when a certain state quantity of the system is highly observable, the variance of the Kalman filter will quickly converge, and the system will be in the subsequent integrated navigation process, the accelerometer zero or When the calibration factor changes, the fil...

AI Technical Summary

Problems solved by technology

[0002] During the use of the integrated navigation system, due to the existence of gravity on the celestial channel, when the Kalman filter is estimated, there is often a slowly increasing deviation between the altitude channel and the satellite altitude when the system performs long-term integrated navigation , which has a great impact on high-precision fixed-point strikes or precision drone hovering

[0003] According to the analysis and summary of the Kalman filtering process, it is found that when a certain state quantity of the system is highly observable, the variance of the Kalman filter will quickly converge, and the system will be in the subsequent integrated navigation process, the accelerometer zero or When the calibration factor changes, the filter can no longer track, resulting in a deviation in the sky direction channel of the integrated navigation, which is manifested in the gradual deviation between the altitude and the satellite altitude during the long-term integrated navigation; when the system does not perform Kalman for a long time When the filtering measurement is updated, due to the sky-direction channel and its easy divergence, the variance of the corresponding height and sky-speed state variables in the Kalman filtering process becomes extremely large, for example, the height diverges to 10000000m, and the gyro drift setting value is 0.01° / h, converges to 0.001° / h in the initial filtering process, and converts to the standard unit of 4.85e-9rad / s. When the system meets the conditions for measurement update again, due to the limited accuracy of the computer during the calculation process, It is prone to the phenomenon of "big numbers eat small numbers", which leads to abnormal filtering

Images