Multi-objective optimization of variable altitude high-dimensional compression-ignition aviation piston engine based on Kriging model and NSGA-III

The operation of compression-ignition aviation piston engines in high-altitude environments is prone to critical issues such as power degradation and insufficient thrust. The research and optimization of the rapid coordination mechanism between fuel and air are crucial for preventing power loss during high-altitude operation and improving the service ceiling of the engine. Based on the constructed one-dimensional thermodynamic model of a compression-ignition aviation piston engine (CI APE), the Kriging surrogate model and non-dominated sorting genetic algorithm (NSGA) are utilized to optimize the brake specific fuel consumption (BSFC), maximum pressure rise rate (MPRR), and the maximum cylinder pressure (Pmax), exploring the optimal fuel-air combination at different altitudes. Firstly, the Pearson correlation coefficient analysis method is employed to confirm variables, and Latin hypercube sampling is used to generate training model samples. Secondly, a Kriging surrogate model of the engine with BSFC, MPRR, and Pmax as objective functions is constructed, and its accuracy is validated. Finally, the NSGA-III is employed for multi-objective optimization. The results indicate that injection timing, compression ratio, high-pressure stage blade opening, and low-pressure stage blade opening have the most significant impact on engine performance. The constructed surrogate models exhibit good predictive accuracy, with coefficient of determination (R2) values all greater than 0.9. At altitudes of 2000 m, 4000 m, 6000 m, and 8000 m, compared to before optimization, the BSFC decreased by 10.1 %, 11.5 %, 12.7 %, and 12 %, respectively. Compared to the power at 2000 m altitude before optimization, the optimized engine can achieve approximately 85 % of the power recovery target at 8000 m altitude.