C ~ 4isr System Target Recognition Theory And Technology Research

Because of lots of information sources in C4ISR system and the requirement for timely and accurate information in future war becoming more highly,the technology of identification fusion of C4ISR system in information synthesizing has become a research spot in area of information fusion. Because of the information provided by sources in C4ISR system is very uncertain and Dempster-Shafer evidential theory providing powerful and nature means for the express and combination of uncertain information,research in this thesis is aimed at modeling of evidence,testing of evidence consistency,evidential combination method and performance evaluation in identification fusion of C4ISR system by use of D-S evidential theory.By extending comparability relation of belief function,the process of founding a generic evidential model is described for kinds of uncertainty. This process of modeling has wider range of application than other modeling methods of evidence. Aiming at two typical cases which evidences are uncertain to which target they are referring,a evidential cluster algorithm based on minimizing superconflict rule and a evidential classification algorithm based on -nearest neighbour rule are proposed based on the conflict among evidences. The physical meaning and pertinence of these two algorithms are clear. And these are accessorial to conquer the ambiguous of data correlation. Based on analyzing evidential combination methods,a generic framework of evidential combination methods is built and a evidential combination method based on proportional belief assignment rule is proposed. Simulation results show that this method possesses obvious advantage in the case of steadiness. In order to improve the veracity of decision,truncated D-S method is ameliorated. It is proved that new method excels in the veracity of decision based on basic probability assignment function. Based on above academic research,performance of identification fusion algorithm is tested by the testing platform of navy. The result shows that the proposed evidence combination method can guide the design of identification fusion algorithm and meet the requirement of application.