Condition Based Maintenance in Distributed Production Environment

Condition Based Maintenance (CBM) is part of the "on demand" response of the enterprise. Unlike other enterprise responses it contains a large amount of uncertain information, qualitative and numerical data. Distributed unmanned enterprises like electric power distribution network add to this task the demand for possibly uninterrupted users' service. The only reliable data are on-line data from the distribution network: transformer stations and switchyards. A multilevel condition evaluation framework is proposed for support of decision analysis on where to intervene in the system in order to ensure maximum system efficiency. Intelligent system monitoring is supplied with central knowledge processing and essential use of expert heuristics for detection of dubious maintenance scenarios. Dampster - Shaffer theory and Yang - Xu synthesis axioms are at the basis of maintenance object decomposition. The major obstacle is detected in one - to - many correspondences of real-time measurement data and object condition based decomposition. Results of CBM decomposition of distribution electric power grid are presented.