NEW FEATURES OF KNOWLEDGE REPRESENTATION IN MAINTENANCE EXPERT SYSTEMS
One of the primary tasks of maintenance is to define, i.e. to make a decision on what and when should be replaced to avoid a breakdown. Until today, reaching of such decisions has mostly been the concern of a long experience and skills of a system maintainer who is able to detect the possible failures during a certain period of time. This problem is particularly manifested in different and complex maintenance systems where several dozen or several interdependent values are monitored. Practically, human monitoring of such maintenance systems is impossible, as e.g. in the cases when in a real time and on the basis of the values monitored a decision should be made on stopping the machinery. Therefore, in recent times apply intelligent information systems that solve these problems successfully. In the paper a new generation of graphical expert system toolkit -VisiRule based on Prolog programming language and its possibility of application in maintenance technology is described. Some of the main characteristics of the toolkit for visual development and building of complex expert systems are also presented. The possibility of building the traditional expert system based on the rules and the expert system on fuzzy logic is demonstrated by examples.
maintenance ezpert systems CMMS
Dragutin Lisjak Ivo Cala Vesna Alic Kostesic
University of Zagreb Faculty of Mechanical Engineering and Naval Architecture Department of Industri Polytechnic of Zagreb Vice Dean for Science and External Affairs Department of Mechanical Engineerin Polytechnic of Zagreb Department of Mechanical Engineering Vrbik 8, 10000 Zagreb, Croatia
国际会议
World Maintenance Forum(世界维修论坛)
广州
英文
1-8
2009-11-27（万方平台首次上网日期，不代表论文的发表时间）