基于案例推理方法在惯性导航故障诊断专家系统的应用

摘要

以测量船惯性导航系统为例,为解决其故障诊断专家系统知识瓶颈问题,特别是突发、原因不详的故障,利用专家知识和以往的工作经验积累,在故障诊断专家系统中融入了基于案例的推理方法.这种推理方法把知识获取简化为领域内经验知识的收集,并以此为基础构造案例库;把基于规则的推理转化为基于案例推理,即在案例库中检索相似的经验知识.采用基于案例推理的方法,可以简化知识获取的过程,求解效率高、质量高,具有持续的学习能力,能有效提高故障诊断系统的智能化程度.文中重点介绍了构建该故障诊断专家系统的案例知识表示方法、案例的索引和检索算法等关键技术.%Taking an inertia! navigation system of survey ship as an example, to cope with the knowledge bottleneck of the expert system, especially to the unexpected, unexplainable fault, a fault diagnosis expert system is adopted based on the Case-Based Reasoning of the Inertial Navigation System using expert knowledge and a plenty of work experience. The learning mechanism of CBR could simplify the knowledge acquisition through constructing case-based database based on collecting the experience knowledge in the field, and the Rule-Based Reasoning is translated into Case-based Reasoning through retrieving the similar experience in the case-based database. With the advantage of simplified knowledge acquisition mechanism and high quality solution efficiency, the method can effectively improve the performance and the intelligence of the fault diagnosis system. In addition, several key technologies for constructing the fault diagnosis expert system were emphatically introduced, such as knowledge express method, case index, search arithmetic and so on.