Library Subscription: Guest

ISSN Online: 2379-1748

ISBN Flash Drive: 978-1-56700-431-1

ISBN Online: 978-1-56700-430-4

First Thermal and Fluids Engineering Summer Conference
August, 9-12, 2015 , New York City, USA


Get access (open in a dialog) pages 2195-2204
DOI: 10.1615/TFESC1.tdp.012932


Rising from dynamically updating challenges of sustainable development the issue of energy security and consequent energy technology improvement is always on the agenda. Robust application of ejector technology systems (ETS) requires application of working fluids covered by number of contradicting criteria. The selection of working fluids with desirable combination of such properties as contribution to greenhouse effect, flammability, toxicity, thermodynamic behavior, performance specifications, and the others is one of the most important stages in ETS simulation and design.

In this study, we propose trade-off working fluids for application in the ETS based on the 'tailored' working fluids concept. The artificial intelligence methods are applied to perform evaluation of criteria used in the feasibility description of different ETS configurations. In case of lack of data for new candidates the artificial neural network correlation are applied. The networks for coefficient of performance (COP), entertainment ratio, and pressure ratio (output) as functions of critical temperature, critical pressure and normal boiling temperature (input) are trained using data from thermodynamic databases. The same approach is applied for forecasting of flammability and toxicity of candidates. The accuracy of neural network prediction for the cycle performances generally does not exceed 4%.

Criteria of sustainable development cannot be formulated on a strict mathematical basis and always have subjective character. Finding the compromise actually is a non-trivial decision-making problem and cannot be formalized. The selection criteria for trade-off working fluids of Pareto optimal solutions set involve the fuzzy criteria mappings formalized as an intersection of membership functions.