Evaluating Measurement Quality in Kalman Filtering Using Second-Order Mutual Differences of Redundant Measurements

Abstract

Advanced signal detection and state parameter estimation are critical across radar, sonar, communications, video processing, and fault diagnosis. Sensor measurements inherently contain noise. Kalman filtering is effective for noise reduction and state estimation; however, outliers adversely affect adaptive Kalman filter accuracy. This paper introduces a second-order mutual difference (SOMD) algorithm using redundant measurements to evaluate measurement quality and mitigate the impact of abnormal observations.