摘要
规则表示和推理机制是计算机形成操作票过程中的关键问题,分析了在操作票专家系统中广泛采用的产生式规则表示方法的局限性,指出推理程序对规则表示的依赖是制约操作票专家系统通用性的根本原因。针对上述问题,首先研究了产生式规则的Petri网表示方法,提高规则表示的灵活性;其次,研究了基于Petri网规则表示的推理方法,使推理程序与规则表示完全分离,最后,结合应用实例说明了基于Petri网的规则表示方法和推理机制可以有效地提高操作票专家系统的适应能力。
Rules representation and inference mechanism are key problems in process of generating operation order sheets by computer. This paper analyses the disadvantage of production knowledge representation which has been widely used in operation order sheets expert system, and indicates that dependent relation between inference program and rules is the main reason which restricts the adaptability of operation order sheets expert system. To solve the above problem, firstly, this paper establishes a model of production knowledge based on Petri net, which improves flexibility of rules representation.Secondly, a new inference mechanism is brought forward, which can separate inference program from rules representation. In the end, an applied example is presented to show that Petri nets based rules representation and inference mechanism can imorove adaotabilitv of ooeration order sheets expert system.
出处
《继电器》
CSCD
北大核心
2007年第20期31-36,共6页
Relay
基金
国家电网公司科技项目(SGKJ[2003]01-16)
关键词
专家系统
PETRI网
操作票
规则表示
推理机制
expert system
Petri nets
operating command
rules representation
inference mechanism
