A study on multi-parameter and multi-objective optimization for centrifugal compressor

In this paper, the Optimal Latin Hypercube Sampling (OLHS) method is employed to sample nine parameters that define the blade profile of a 75kW centrifugal compressor, and numerical simulations are carried out to identify the five parameters that have the most impact on its performance. With the objectives of maximizing the compressor's outlet pressure and multivariable efficiency, these five key parameters are simultaneously optimized to determine the optimal combination, using the Kriging approximation model combined with the NSGA-II optimization algorithm. As a result, the optimized compressor demonstrates a 1.03% increase in outlet pressure and a 3.18% improvement in efficiency at the design condition.