The On-Line Modeling and Filtering of Echo-Data Noise to Immobile Target

Abstract

Abstract Aiming at the echo-data noise of immobile target in radar data processing, the third order autoregressive model-AR(3) model of echo-data is studied and built by the time series method; a new and improved AR(3) model are designed, by which the echo-data model is set up in real time. According to this model and the Kalman filter arithmetic, the echo-data noise is filtered in real time. The echo-data that are filtered and not filtered are analyzed by the Allan variance analysis method. The experimental and simulating results show that variable range of distance is reduced from 2940-3060m to 2975-3025m and its measuring error is reduced from 25m to 21m by this modeling and filtering method. The correctness of this AR(3) model and the effectiveness of this method are also proved by the results of Allan variance analysis. This method of modeling and filtering has such advantages: on-line modeling, filtering in real time, updating of model and so on. It can be applied in target detection, positioning, imaging, radar's performance test and calibration when the target is immobile.