基于伴随方法、改进文化基因算法和kriging代理模型的涡扇发动机短舱减噪优化方法研究

项目名称： 基于伴随方法、改进文化基因算法和kriging代理模型的涡扇发动机短舱减噪优化方法研究

项目编号： No.51506189

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 能源与动力工程

项目作者： 邱昇

作者单位： 中国航空研究院

项目金额： 21万元

中文摘要： 宽频风扇噪声是现代涡扇发动机的主要噪声源之一，预测和降低其大小成为工业界的重要需求和挑战。通过优化发动机短舱外形来降低其噪声是一种行之有效的方法。本项目基于Kriging 代理模型，提出一种仅需少量初始样本点的快速收敛的混合优化方法。其中，首先提出基于伴随方法的多参数、宽频风扇噪声减噪设计理论，通过使用伴随方法进行局部寻优，并将该局部最优点加到样本中进行下一次优化，显著提高代理模型的精度; 通过改进的高效文化基因算法（memetic），在响应面上快速找到满足气动约束下的全局最优设计点。该混合优化方法可大大降低多参数降噪设计计算时间。本项目研究成果为涡扇发动机短舱宽频风扇噪声减噪设计提供了全新的途径，可以高效处理含大量设计变量的降噪设计问题，在满足气动约束下，显著降低宽频风扇噪声，为我国低噪声涡扇发动机短舱研制提供科学依据，并探索出一条高效可靠的新路。

中文关键词： 风扇噪声；伴随方法；文化基因算法；代理模型；发动机短舱

英文摘要： Since fan broadband noise is a major source of annoyance in modern aeroengines, prediction and reduction of this component is nowadays an important industrial need and probably one of the most challenging issues in turbomachinery aeroacoustics. It’s an effective method to reduce the fan noise by optimizing the shape of engine nacelle.. A fast convergent hybrid aeroacoustic optimization method based on kriging surrogate model is proposed, in which only little initial samples are needed. In this project, the aeroacoustic optimization design theory based on the adjoint method has been proposed first for broadband fan noise. The adjoint method is used to obtain the local optimum design point and the obtained local optimum design point is add to samples of kriging model, which significantly improve the precision of the kriging surrogate model; the global optimum design point satisfying the aerodynamic constraint can be quickly found on the response surface through using the developed improved memetic algorithm. Based on those, this hybrid aeroacoustic optimization method can outstandingly reduct the total cost of multi-parameter optimization design for low-noise. The achievements of this project not only provide the new way to carry out the aeroacoustic optimization of turbofan nacelle and to deal with the multi-parameters noise reduction problems with high efficiency, but also offer solid scientific evidence for the development of our low noise turbofan nacelle under aerodynamic constraint.

英文关键词： fan noise;adjoint method;memetic algorithm;surrogate model;turbofan nacelle