基于熵权优属度的模糊传感器目标识别方法  

Object Recognition Method Based on Entropy Weight Optimal Membership for Fuzzy Sensor

作　　者：万树平[1] 

机构地区：[1]江西财经大学信息管理学院,南昌330013

出　　处：《计算机工程》2009年第6期26-28,共3页Computer Engineering

基　　金：国家自然科学基金资助项目(10626029);江西省自然科学基金资助项目(0611082);江西省教育厅科技基金资助项目(GJJ08350)

摘　　要：针对具有多个特征指标的模糊多传感器目标识别问题,提出一种新的模糊多传感器数据融合方法。该方法根据信息熵理论,引入不均衡度定义熵权矢量,通过求解数学规划问题,得到各目标类别的优属度,并给出目标识别规则。实验结果表明,该方法能提高目标识别结果的客观性和可信度,具有可操作性。Aiming at the object recognition problem with multiple characteristic indexes for the fuzzy multi-sensor, a new fusion method for the fuzzy multi-sensor data is proposed. According to the theory of information entropy, the method introduces the disequilibrium index to define the vector of entropy weight. By solving the mathematical programming, the optimal membership degree for each target type is obtained, and the rule of object recognition is given. The method improves the objectivity and trustworthy degree of recognition result. The applied example proves that the method is exercisable.

关 键 词：模糊传感器 数据融合 目标识别 熵权 优属度 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]