提出了一种基于第一原理的定性诊断探测法，该方法将经验知识纳入算法中，与国际上典型的ＧＤＥ系统中的Ｓｈａｎｎｏｎ熵预测法相比更快速、简单、有效，并且对探测值与新候选的关系进行了仔细分析，利用所得结论可大大地减少诊断系统的计算时间。 A qualitative diagnosis detection method based on the first principle is proposed, which incorporates empirical knowledge into the algorithm, which is faster, simpler and more efficient than the Shannon entropy prediction method in the typical GDE system in the world. The detection value and the new Candidate relationships have been carefully analyzed, and the conclusions drawn from the study can greatly reduce the computational time for diagnostic systems.