Research on Emitter Recognition Model Based on Dempster-Shafer Evidence Theory

摘要

In the battlefield environment, emitter information detected by multisensor takes on temporal redundancy. In order to solve emitter recognition problems in such practical reconnaissance environment, D-S reasoning method based on information fusion is applied. The key problem to D-S reasoning is basic probability assignment function, which to a great extent limit its applications. For the special traits of emitter recognition, a new method of constructing basic probability assignment function based on gray correlation analysis is present. Gray correlation coefficient between reference data array and comparable data array is calculated first, and then gray correlation degree set of multi-observation samples is given. Under the frame of Bayes belief, the new belief function is deduced. Examples of identifying the emitter type have been selected to demonstrate the new method. If four sensors are used, the new constructed basic probability assignment function enables the correct recognition rates to reach 98% in the simulation environment 1. Experimental results show that this information fusion method is accurate and effective and the correct recognition rates are improved evidently with the increasing number of sensors, especially with the increasing measurement noise.