Multi-objective design optimization of a rotary compressor for household air-conditioning

Abstract

This paper presents a design optimization of a rolling piston compressor using a multi-objective optimization technique that uses a genetic and evolutionary algorithm. The procedures begin with a pool of compressor designs called population, which were generated pseudo-randomly based on the preset constraints, to arrive at a set of optimum trade-off solutions from the various multiple objective function sets. The optimum solution set allow designers to choose a particular optimum solution that best suited their needs. The cases under examination attempt to optimize combinations of some nine objective functions: the coefficient of performance, refrigerating capacity, motor input power, friction power, indicated work, discharge valve loss, suction valve loss, compressor overall size, and machine cost. There are 18 compressor design variables that are allowed to vary during the optimization process bounded by 23 preset constraints. Results show an effective employment of the multi-objective optimization technique in compressor design.