产生式知识库的不动点计算建模方法

A modeling method based on fixed point computation for a production knowledge base

提出了利用产生式规则的过程性特征对产生式知识库进行建模的方法.该方法在给出条件迁移定义的基础上引入不动点计算构建条件迁移系统,将模型构建过程转换为求解状态集合构造函数最小不动点的过程,并给出了求解算法的正确性与可终止性证明;该方法的时间复杂度相对静态建模方法降低了一个量级,构造的条件迁移系统包含了状态迁移过程的完整信息,同时解决了静态建模方式构建的迁移系统信息丢失问题,提高了错误诊断的效率.

A modeling method based on the procedural nature of production rules was proposed to build a system model.The definition of the conditional transition system（CTS） was given by the concept of conditional transitions.Small fix point computing was introduced to build the conditional transition system.The system model represented by the conditional transition system was built by solving the small fix point of the constructing function of the states set,and the correctness and termination of this method was proven.The proposed dynamic modeling method improves the efficiency of the production knowledge base modeling process.The time complexity reduced one order of magnitude compared with the static modeling method.The conditional transition system built by this method contains all information about the state transition process,and solves the information loss problem of the transition system constructed by the static modeling method,increasing the efficiency of error diagnosis.