Architecture of a distributed system for acquiring knowledge in supporting decision making in operation/maintenance of electric power plants

This paper presents a new approach to acquire a knowledge base (MKB) of continuous processes through a distributed system. The proposed methodology combines MKB and case-based reasoning (CBR) for adaptive knowledge base creation, maintenance and updating trough the use of an architecture with the following aspects: i) distributed architecture network; ii) real-time operation; iii) advanced visual interface. Although originally developed for use in hydroelectric power plants it is useful for other industrial processes in mission-critical with high cost of interruption. This work presents the system requirements, the aspects of the case base formation, the CBR cycle and the believe network.