Pseudo linear estimator developed for passive target tracking is extended to track the targets using measurements from active sonar. Here there is no need of initializing covariance matrix of target state vector and hence its performance is better than that of Kalman filter. Though this offers a biased estimate in certain scenarios, it has an advantage as it hardly diverges. It offers the features of Extended Kalman filter viz., sequential processing, flexibility to adopt the variance of each measurement etc. The measurements of a fixed length in sliding window are used to track a maneuvering target. The results of this estimator are compared with that of Kalman filter with input estimation. The Monte-Carlo simulation results are presented for two typical scenarios.
展开▼