在一些复杂控制系统中 ,专家系统作为一个决策控制器 ,满足整个系统的性能需要 ,因此研究专家系统的推理时间是必要的。专家系统的推理时间与其推理模式、知识库结构等因素有关。针对知识库结构 ,通过专家系统的推理时间研究了实时性问题 ,利用时间齐次马尔可夫链为专家系统知识库进行建模 ,并给出了相应的时间估计模型及其排列准则。 In some complex control systems, expert system as a decision-making controller to meet the performance requirements of the entire system, so it is necessary to study the reasoning time of expert system. The reasoning time of expert system is related to its reasoning mode, knowledge base structure and other factors. Aiming at the structure of knowledge base, the problem of real-time is researched through the reasoning time of expert system. The homogeneous time-dependent Markov chain is used to model the expert system knowledge base, and the corresponding time estimation model and its arrangement criterion are given.