Modified unscented Kalman filter for X-ray pulsar-based navigation system in the presence of measurement outliers

In order to eliminate the influence of outlier in measurement to X-ray pulsar based navigation system, a modified unscented Kalman filter-based filtering algorithm is proposed. First, the outlier detection function is established based on the statistical information of measurement residuals to detect whether the fault occurs in the measurements. Then, the filter gain in regular unscented Kalman filter algorithm is modified so that the dramatic changes of measurement residual are avoided. The modified unscented Kalman filter filtering algorithm is tested on the existing high earth orbit to verify the validity in the presence of the measurement with outliers. Simulation results show that the proposed modified unscented Kalman filter can effectively reduce the impact of the outliers on X-ray pulsar-based navigation system and has higher accuracy than the unscented Kalman filter method.