模糊神经网络应用于数据融合中实现目标的身份识别

An Application of Fuzzy Neural Network to Targets Identity Identification of Data Fusion

结合“情报侦察与数据融合专家系统”的需求 ,对基于产生式规则的模糊知识表示和推理、神经网络的结构和机理进行了研究 ,提出了将模糊规则与神经网络相结合 ,实现一种新的推理方法。在此基础上 ,实现了一种将模糊规则转化为神经网络的算法以及神经网络的学习算法 ,并具体讨论了这种新的推理方法在“情报侦察与数据融合专家系统”中实现飞机的身份识别的应用。实验证明这种网络结构巧妙地将模糊集理论和神经网络结合起来 ,充分利用了两者各自的优势 。

Combined with the need of 'Intelligence reconnaissance and data fusion Expert System', studied the knowledge representation and inference of production?mechanism and construction of neural network. A new method of inference is proposed by the combination of fuzzy rule and neural network. On the basis of above discussion, a method of translating fuzzy rules into neural network and a learning algorithm are put forword. In addition, applications of the new method at identity identification of airplane in 'Intelligence reconnaissance and data fusion Expert System' a re discussed. The experiment results indicate that this neural strucure by the felicitous combination of fuzzy rule and neural network fulfills the target identity identification by the fullness utilization of their predominances.