为快速有效地实现案例推理,提出了基于多EBP神经网的案例匹配技术,设计了改进的德尔菲法来确定输出样本集.与传统的最邻近法的结果相比,对于复杂系统,该技术能够提高案例匹配的精确性,更适于解决案例属性权重难以确定或存在着耦合与非线性关系的问题. In order to realize the case reasoning quickly and effectively, the case matching technology based on multi-EBP neural network is proposed and the improved Delphi method is designed to determine the output sample set. Compared with the traditional results of the nearest neighbor method, this technique Can improve the accuracy of case matching, and is more suitable for solving the problem that the weight of case attribute is difficult to be determined or the relationship between coupling and non-linearity exists.