A robust recursive least squares passive location algorithm
To solve the problem that airborne passive location is susceptible to outliers, a robust recursive least squares passive location (RRLS) algorithm based on the angle information is proposed. First of all, we discusses source of outliers, establish the airborne single passive location model and the least squares (LS) solution is obtained. Secondly, the concept of pollution distribution is introduced to construct a robust extremal function, and we establish an appropriate equivalent weight function, so that the robust least squares location algorithm is obtained. Then the algorithm is transformed into recursive form based on the real-time requirements of the location. Theoretical analysis show that the RRLS algorithm is the promotion of the LS algorithm and the influence function value of RRLS algorithm will decrease with a strong robust performance as the error increases while the LS algorithm is not able to resist outliers. The simulation results show that the two algorithms are able to better converge if there are no outliers. When outliers appear, the estimated result of LS will be distorted while the RRLS algorithm can automatically identify outliers and reduce the influence of the outliers by reducing weight or removing the abnormal data to effectively improve the robustness of location.
Author's Name: Wu, H., Chen, S., Hou, Z., Zhang, H.
Volume: Volume 9
Issues: Issue 4
Keywords: Equivalent weight, Least squares estimation, Passive location, Robust estimation