Debiased Converted Measurement Kalman Filter Algorithm for Optic-electric Target Tracking

When using optic-electric devices for target tracking, due to target obscured, measurement equipment, and so on, so that the measurement data will be lost or singular value. Therefore, this paper designed an improved debiased converted measurement Kalman filter (RDCMKF). The idea of the method is that calculate out a scaling factor through the target measured values and predicted values. Then adding the scaling factor in status updates so only the data of the faulty sensor is scaled. Thus the algorithm has good robustness. And because of the scaling factor is associated with the measured value and predicted value of target, any unnecessary target information loss is prevented. The simulation results show that new debiased converted measurement Kalman filtering has a better robustness than the traditional debiased converted measurement Kalman filtering when the measurement data is missing or outliers. When the measurement data is outliers, the peak of the former's filtering position error reduced almost 90% than the latter.