Organic Rankine Cycle Systems Design Using a Case-Based Reasoning Approach

Over the past several decades, the organic Rankine cycle (ORC) has been considered to be a promising technology for converting low-grade heat sources into electricity. ORC design is a knowledge-intensive procedure with a variety of design variables and operational constraints that normally needs large-scale and expert interventions. Computer-aided tools for systematic ORC design have been emerging for decision support in recent years. However, it remains an open question regarding how to systematically exploit existing ORC plant designs to provide more valuable information for new ORC systems design and operation. Case-based reasoning (CBR) technology is feasible and suitable to facilitate this exploitation. In this paper, a new CBR-based approach to ORC design is proposed. The complete workflow is described in detail along with the essential processes of case representation, case base establishment, similarity measurement, attribute weighting, and basic CBR steps. The leave-one-out cross-validation method is employed to evaluate the performance of this CBR model. Moreover, an example is conducted to demonstrate how this approach works. The approach results in an improved design scheme with noticeable increases in the net power output, thermal efficiency, and exergy efficiency, up to 3.17%, 5.88%, and 6.80%, respectively. Therefore, the new approach is a feasible and effective way to make use of existing ORC plant designs systematically. The implementation of the CBR approach is a meaningful attempt to support decision-making and facilitate the development of ORC systems.