Optimized design of aero-engine high temperature rise combustion chamber based on 'kriging-NSGA-II'

摘要

This article proposes a kriging-based surrogate model for the combustion performance of aero-engine high-temperature combustion chambers, and applies the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) method to carry out multi-objective optimization of the design of aerodynamics. The main purpose of this is to augment the NSGA-II algorithm using the kriging surrogate model to optimize the combustion performance of the combustor when the computing resources are limited. The experimental results show that compared with the traditional one-dimensional calculation method, the proposed method can quickly and accurately predict the combustion efficiency and total pressure loss coefficient. Compared with the results based on cubic polynomial and artificial neural network, the root mean square errors predicted by the kriging model are the smallest, at 0.0045 and 0.0878, respectively; this confirms the effectiveness of the kriging model in predicting the combustion performance of the aero-engine high temperature rise combustor designed in this paper. Using a global sensitivity analysis method, the design variables are sorted, and a multi-objective optimization method is studied based on the NSGA-II algorithm. The kriging model predicts a set of optimal Pareto non-dominated solution sets. Based on these results, it is concluded that the kriging method based on the surrogate model is more accurate in the prediction of the combustion performance of an aero-engine combustion chamber and can accelerate the optimal design of the combustion chamber.