Abstract: Current remote expert system does not make any control of redundancy when knowledge is stored.That makes the knowledge base redundant，and makes the management and utilization of the knowledge more difficult.In order to solve the knowledge redundancy problem during the process of remote real-time diagnosis，the non-redundant storage technology of dynamic knowledge is proposed.The concrete algorithm of this technology is realized on the platform of remote fault collaborative diagnosis of aircraft.And through data instance，the effectiveness and correctness of the algorithm are analyzed and demonstrated.The results show that the non-redundant storage technology of dynamic knowledge makes knowledge storage accurate and non-redundant，and helpful to improve the reasoning efficiency of remote real-time diagnosis.

Key words: dynamic knowledge storage, remote real-time diagnosis, expert system, redundancy control, knowledge base

摘要： 当前的远程诊断专家系统在知识存储时并未做任何冗余控制，使得知识库中的信息出现较大冗余，给知识的管理和利用带来较大难度。为了解决上述问题，对无冗余的动态知识存储技术进行了研究，并基于飞机远程故障协同诊断平台，对该技术的具体算法进行了实现，通过数据实例对算法的有效性和正确性进行了分析论证。结果表明：无冗余动态知识存储技术可以精确地对知识信息进行无冗余存储，从而有效地提高远程实时诊断的推理效率。

关键词: 动态知识存储, 远程实时诊断, 专家系统, 冗余控制, 知识库