基于混合搜寻法的车用涡轮增压器多目标优化

Multi-Objective Integrated Optimization of Vehicle Turbocharger Impeller Based on Hybrid Search Method

摘要:

选取压气机等熵效率、叶轮最大应力和最大形变为优化目标，基于ANSYS Workbench建立了涡轮增压器压气机流热固多物理场耦合仿真模型，利用试验设计方法中的正交矩阵法和2k（k表示试验有k个因子）法混合搜寻出压气机关键设计参数，运用径向基函数及非支配排序遗传算法解决全局寻优问题，最终获取叶轮结构设计最优方案.计算结果表明：叶轮最大形变优化效果最显著，降低28.09%，最大应力减小16.34%，压气机等熵效率提高2.83%.

Abstract:

Different from optimization study of the single performance of turbocharger impeller, this paper takes the compressor isentropic efficiency, the impeller maximum stress, and the impeller maximum deformation as the optimization objective. Based on ANSYS Workbench, the turbocharger compressor flowthermalsolid multiphysics coupling simulation model was established. Using the orthogonal matrix method and the 2k method, the key design parameters of the compressor were studied. The radial basis function and nondominated sorting genetic algorithm II were applied to solve the global optimization problem. Finally, the optimal scheme of impeller structure design were obtained. The results show that the most significant decrease is the impeller maximum deformation of 28.09%, the impeller maximum stress is decreased by 16.34%, and the compressor isentropic efficiency is increased by 2.83%.