摘要： 在异类多传感器信息融合目标识别中，不同传感器对系统提供的证据等级不同。为此，提出一种模糊信息融合目标识别方法。将各证据按证据权进行转化，用Dempster-Shafer(D-S)证据理论进行合成，利用模糊数学模型对传感器测量值和数据库中的数据进行建模，根据证据距离得到各证据的相互支持度，进而获得传感器对系统提供信息量的权重。分析结果表明，该方法具有较高的精度和可靠性。

关键词: 异类传感器, 模糊信息, 证据理论, 信息融合, 目标识别

Abstract: Different sensors provide different evidence importance in a target recognizing system with heterogeneous multi-sensor data fusion method. This paper proposes a fuzzy information fusion target recognition method. The evidences are transformed according to their weights before fused together using the Dempster-Shafer(D-S) evidence theory. The sensor measurements and data in the database are simulated by using the fuzzy mathematical model, and the mutual support degree among evidences is obtained from the evidence distances in order that the information weight of the evidence to the system is obtained. Analysis results show that this method has higher accuracy and reliability.

Key words: heterogeneous sensor, fuzzy information, evidence theory, information fusion, target recognition

中图分类号: 

• TP391