Robust Adaptive Filtering and State Estimation Method Based on Huber Estimation

Abstract

The present invention provides a robust adaptive filtering and state estimation method based on Huber estimation. Firstly, an estimation value is calculated through estimating an upper bound of a covariance matrix of a one-step prediction dual-uncertainty system in an online way, only a time updating process is modified, at the same time a method of a linear matrix inequality is used to carry out adaptive parameter adjustment, and an adaptive traceless information filter for an uncertainty model is obtained. Then the matrix inversion theorem is applied and a Huber estimation method is used to modify a measurement update process, and a final estimation value is obtained. The precision of filtering and state estimation is high, and the stability is good. According to the robust adaptive filtering and state estimation method, a filtering and state estimation problem of a dual-uncertainty unmanned ship is solved, the advantages of an information filter structure are retained, the precision and stability of filtering and state estimation are improved, and the calculation is simple and the application is convenient.

Description

technical field [0001] The invention relates to a robust adaptive filtering and state estimation method, in particular to a robust adaptive filtering and state estimation method based on Huber estimation. Background technique [0002] The Unscented Information Filter (UIF) is an emerging filtering and state estimation method. It is essentially a Kalman filter updated by the inverse of the covariance matrix. In view of the fact that the unscented information filter has no requirement for the initial value of the covariance matrix , the update equation that is easier to calculate than the ordinary Kalman filter, and the extension of multi-source fusion, etc., have attracted widespread attention. However, for the dual uncertain system where the uncertain motion model of the UAV, the ocean disturbance model and the measurement noise of the uncertain sensor coexist, the filtering accuracy of the conventional unscented information filter will drop significantly or even fail to con...

AI Technical Summary

Problems solved by technology

However, for the dual uncertain system where the uncertain motion model of the UAV, the ocean disturbance model and the measurement noise of the uncertain sensor coexist, the filtering accuracy of the conventional unscented information filter will drop significantly or even fail to converge.

Moreover, most of the existing wave filtering and state estimation technologies are researched on a certain uncertainty problem and put forward a better solution. For the above-mentioned dual uncertain system containing two or more uncertain factors, no effective solution has been given. Filtering and State Estimation Methods

Images