A reliability-based optimization of membrane-type total heat exchangers under uncertain design parameters

Membrane-type Total heat exchangers (Heat and moisture recovery ventilators) have become key components for building energy conservation. Various materials and duct structures have been proposed to find an improved energy performance and a better economic return. However most previous optimizations of total heat exchangers are limited to deterministic design parameters whereas operational flexibility and feasibility were less concerned. In the real industrial world, each system experiences various disturbances due to changes in outside weather conditions, flow rates, size errors in equipment manufacturing, as well as uncertainties in interest rates and material properties. In this research, an approach, named the SLGA (Single-loop Genetic Algorithm) is proposed for the optimization of membrane-based total heat exchangers under severe uncertain operating conditions, which are represented by the reliabilities of design guidelines. The optimization problem selects eight uncertain parameters such as material type, duct structure, exchanger size, interest rate, etc, as the input variables, and the exchanger performance like economic return, sensible and latent effectiveness are selected as the objective functions respectively. Then the probabilistic constraints are transformed into deterministic forms by a single-loop deterministic method. The discrete and non-linear optimization problem is thereafter solved by a direction-based GA (genetic algorithm). (C) 2016 Elsevier Ltd. All rights reserved.