Consistent multi-sensor data fusion based on nearest statistical distance

To eliminate the mistake errors within measurements from multiple sensors, consistent data fusion method is widely studied. Based on the existing methods, a new statistical distance is defined, which is designed to measure the support degree between any two sensors' measurements, then based on this statistical distance, a new consistent multi-sensor data fusion method which is based on the nearest statistical distance is proposed. Numerical simulation results show that, compared with the existing methods, the new defined statistical distance can reflect the support degree between any two sensors' measurements more rationally, and computational complexity and load of the new method are greatly reduced.