A Modified Unbiased Converted Measurement Target Tracking Algorithm Based on Expectation Maximization

摘要

Aiming at the problem of radar measurement system tracking accuracydegradation affected by wild values, this paper proposes an ExpectationMaximization Modified Unbiased Converted Measurement Kalman Filter(EM-MUCMKF). Firstly, the target measurement information is performedan unbiased conversion. Then a time prediction and a measurement updateare performed under the Kalman filter framework. Next the updated targetstate is regarded as a new measurement to perform an unbiased conversionagain, for the covariance revaluated at the updated target state is moreaccurate and much less noisy. Finally, under the framework of expectationmaximization, the adaptive factor matrix of the measurement noisecovariance is calculated which is used to correct the measurement noisecovariance. The simulation results show that compared with traditionalalgorithms, the proposed algorithm can get more accurate target stateestimation in the environment affected by wild values.