Optimizing a Rolling Piston Compressor Using a Genetic Algorithm

摘要

The paper presents a geometrical optimization of a rolling piston compressor using a genetic algorithm.The compressor was designed for household air-conditioning running on HCFC-22.The search domain is made up of a highly constrained 12 design variables.The case under examined has two objectives;maximising cycle refrigerating capacity while minimising the compressor indicated power input.The initial test suggests that a population number of 10 is sufficient for the optimisation of this particular problem.When compared with the conventional single objective optimisation technique,the genetic algorithm (GA) is generally required a larger number of model executions and hence a larger amount of computational time.However,the benefits of the GA are two folds;not only it is capable of handling multiple conflicting objective functions in its original format,the final results from GA provide the whole spectrum of optimised designs,the final selection of which dependent on the specified required compromised of each of the mutually conflicting objective functions.