A Combined Expert System/case-based Reasoning Approach for Cogeneration Plant Design

The paper deals with the design problem of natural gas cogeneration systems. Despite its complexity, this design problem subjected to several constraints is properly solved by humans, which suggests the use of artificial intelligence (AI). Moreover, the design process is strongly based on knowledge that experts have in a domain, so that the AI techniques known as knowledge-based systems (KBS) are a well-suited approach for the cogeneration plant design. Expert systems (ES) and case-based reasoning (CBR) are two of the most important KBS techniques. An ES prototype for natural gas cogeneration systems design and optimization has been developed by the authors and reported in previous articles. In this work, the basic concepts of CBR technique and a preliminar CBR prototype for cogeneration plant design are presented. The incorporation of CBR technique in the ES prototype is discussed as well. This combined approach allows the ES prototype to retain a solution in a case-base for future references, i.e., the prototype learns from its previous experiences. Quite advantageously the CBR technique can be implemented in the same shell used to develop the ES prototype, through the object-oriented language available in the shell.