Research on multi-fidelity aerodynamic optimization methods

 Constructing high approximation accuracy surrogate model with lower computational cost has great engineering significance. In this paper, using co-Kriging method, an efficient multifidelity surrogate model is constructed based on two independent high and low fidelity samples. Co-Kriging method can use a greater quantity of low-fidelity information to enhance the accuracy of a surrogate of the high-fidelity model by modeling the correlation between high and low fidelity model, thus computational cost of building surrogate model can be greatly reduced. A wing-body problem is taken as an example to compare characteristics of co-Kriging multi-fidelity (CKMF) model with traditional Kriging based multi-fidelity (KMF) model. A sampling convergence of the CKMF model and the KMF model is conducted, and an appropriate sampling design is selected through the sampling convergence analysis. The results indicate that CKMF model has higher approximation accuracy with the same high-fidelity samples, and converges at less high-fidelity samples. A wing-body drag reduction optimization design using genetic algorithm is implemented. Satisfying design results are obtained, which validate the feasibility of CKMF model in engineering design.     

Efficient aerodynamic shape optimization using variable-fidelity surrogate models and multilevel computational grids

A variable-fidelity method can remarkably improve the efficiency of a design optimization based on a high-fidelity and expensive numerical simulation, with assistance of lower-fidelity and cheaper simulation(s). However, most existing works only incorporate “two” levels of fidelity, and thus efficiency improvement is very limited. In order to reduce the number of high-fidelity simulations as many as possible, there is a strong need to extend it to three or more fidelities. This article proposes a novel variable-fidelity optimization approach with application to aerodynamic design. Its key ingredient is the theory and algorithm of a Multi-level Hierarchical Kriging (MHK), which is referred to as a surrogate model that can incorporate simulation data with arbitrary levels of fidelity. The high-fidelity model is defined as a CFD simulation using a fine grid and the lower-fidelity models are defined as the same CFD model but with coarser grids, which are determined through a grid convergence study. First, sampling shapes are selected for each level of fidelity via technique of Design of Experiments (DoE). Then, CFD simulations are conducted and the output data of varying fidelity is used to build initial MHK models for objective (e.g. C_D) and constraint (e.g. C_L, C_m) functions. Next, new samples are selected through infill-sampling criteria and the surrogate models are repetitively updated until a global optimum is found. The proposed method is validated by analytical test cases and applied to aerodynamic shape optimization of a NACA0012 airfoil and an ONERA M6 wing in transonic flows. The results confirm that the proposed method can significantly improve the optimization efficiency and apparently outperforms the existing single-fidelity or two-level-fidelity method.     

Kriging模型及代理优化算法研究进展

代理模型方法由于能显著提高工程优化设计问题的效率,在航空航天及其他领域得到了广泛重视,并逐渐发展成为一类优化算法,本文称其为代理优化（SBO）算法。在现有的代理模型方法中,如多项式响应面、径向基函数、神经网络、支持向量回归、多变量插值/回归、多项式混沌展开等,源于地质统计学的Kriging模型具有代表性,是一种非常具有应用潜力的代理模型方法。以飞行器设计领域的优化问题为背景,介绍了Kriging代理模型及应用于优化设计的理论和算法的最新研究进展。首先,概述了Kriging模型的基本理论和算法,并讨论了影响Kriging模型鲁棒性和效率的几个关键性问题。其次,回顾了Kriging模型理论和算法研究的3个最新研究进展,包括梯度增强型Kriging、CoKriging和分层Kriging模型。而后,分析提炼了基于Kriging模型的代理优化算法的优化机制和优化框架,给出了“优化加点准则”和“子优化”的概念,并介绍了目前常用的几种优化加点准则及其相应子优化问题的求解与约束处理;同时,还介绍了最新提出的局部EI加点准则以及代理优化的终止条件。最后,介绍了代理优化在标准测试函数算例验证、飞行器气动与多学科优化设计典型算例确认方面的研究进展,并对当前存在的一些关键科学问题以及未来研究方向进行了讨论。     

Application of a PCA-DBN-based surrogate model to robust aerodynamic design optimization

An efficient method employing a Principal Component Analysis (PCA)-Deep Belief Network (DBN)-based surrogate model is developed for robust aerodynamic design optimization in this study. In order to reduce the number of design variables for aerodynamic optimizations, the PCA technique is implemented to the geometric parameters obtained by parameterization method. For the purpose of predicting aerodynamic parameters, the DBN model is established with the reduced design variables as input and the aerodynamic parameters as output, and it is trained using the k-step contrastive divergence algorithm. The established PCA-DBN-based surrogate model is validated through predicting lift-to-drag ratios of a set of airfoils, and the results indicate that the PCA-DBN-based surrogate model is reliable and obtains more accurate predictions than three other surrogate models. Then the efficient optimization method is established by embedding the PCA-DBN-based surrogate model into an improved Particle Swarm Optimization (PSO) framework, and applied to the robust aerodynamic design optimizations of Natural Laminar Flow (NLF) airfoil and transonic wing. The optimization results indicate that the PCA-DBN-based surrogate model works very well as a prediction model in the robust optimization processes of both NLF airfoil and transonic wing. By employing the PCA-DBN-based surrogate model, the developed efficient method improves the optimization efficiency obviously.     

Multi-disciplinary design optimization with variable complexity modeling for a stratosphere airship

This paper proposes a hybrid architecture based on Multi-disciplinary Design Optimization(MDO) with the Variable Complexity Modeling(VCM) method, to solve the problem of general design optimization for a stratosphere airship. Firstly, MDO based on the Concurrent SubSpace Optimization(CSSO) strategy is improved for handling the subsystem coupling problem in stratosphere airship design which contains aerodynamics, structure, and energy. Secondly, the VCM method based on the surrogate model is presented for reducing the computational complexity in high-fidelity modeling without loss of accuracy. Moreover, the global-to-local optimization strategy is added to the architecture to enhance the process. Finally, the result gives a prominent stratosphere airship general solution that validates the feasibility and efficiency of the optimization architecture. Besides, a sensitivity analysis is conducted to outline the critical impact upon stratosphere airship design.     

基于改进CST参数化方法和转捩模型的翼型优化设计

为提高翼型优化设计效率,增大设计空间,采用B样条基函数替代传统的形状类别函数(CST)方法中的Bezier多项式,增强了对翼型参数化表达的局部控制能力并提高了翼型局部表达精度。为了确保翼型在优化设计过程中的几何光顺特性和代理模型的准确性,采用小波分解技术提出了多分辨率翼型的局部光顺方法。采用基于本征正交分解(POD)的流场数值代理模型,并结合γ-Reθt转捩模型实现了快速准确的气动力与流动转捩预测。采用小波技术光顺的CST翼型参数化建模、POD流场数值计算代理模型以及γ-Reθt转捩模型,结合遗传算法建立了完整的翼型气动优化设计系统。针对低速层流翼型与超临界翼型进行优化设计,优化设计后的翼型升阻比分别提高了47.42%和45.85%,且对改进前后参数化建模方法的优化性能进行了对比,结果表明本文构建的翼型气动优化设计系统具备很高的优化效率。     

A high-fidelity design methodology using LES-based simulation and POD-based emulation: A case study of swirl injectors

Engineering design is undergoing a paradigm shift from design for performance to design for affordability, operability, and durability, seeking multi-objective optimization. To facilitate this transformation, significantly extended design freedom and knowledge must be available in the early design stages. This paper presents a high-fidelity framework for design and optimization of the liquid swirl injectors that are widely used in aerospace propulsion and power-generation systems. The framework assembles a set of techniques, including Design Of Experiment (DOE), high-fidelity Large Eddy Simulations (LES), machine learning, Proper Orthogonal Decomposition (POD)-based Kriging surrogate modeling (emulation), inverse problem optimization, and uncertainty quantification. LES-based simulations can reveal detailed spatiotemporal evolution of flow structures and flame dynamics in a high-fidelity manner, and identify important injector design parameters according to their effects on propellant mixing, flame stabilization, and thermal protection. For a given a space of design parameters, DOE determines the number of design points to perform LES-based simulations. POD-based emulations, trained by the LES database, can effectively explore the design space and deduce an optimal group of design parameters in a turn-around time that is reduced by three orders of magnitude. The accuracy of the emulated results is validated, and the uncertainty of prediction is quantified. The proposed design methodology is expected to profoundly extend the knowledge base and reduce the cost for initial design stages.     

Adaptive optimization methodology based on Kriging modeling and a trust region method

Surrogate-Based Optimization (SBO) is becoming increasingly popular since it can remarkably reduce the computational cost for design optimizations based on high-fidelity and expensive numerical analyses. However, for complicated optimization problems with a large design space, many design variables, and strong nonlinearity, SBO converges slowly and shows imperfection in local exploitation. This paper proposes a trust region method within the framework of an SBO process based on the Kriging model. In each refinement cycle, new samples are selected by a certain design of experiment method within a variable design space, which is sequentially updated by the trust region method. A multi-dimensional trust-region radius is proposed to improve the adaptability of the developed methodology. Further, the scale factor and the limit factor of the trust region are studied to evaluate their effects on the optimization process. Thereafter, different SBO methods using error-based exploration, prediction-based exploitation, refinement based on the expected improvement function, a hybrid refinement strategy, and the developed trust-region-based refinement are utilized in four analytical tests. Further, the developed optimization methodology is employed in the drag minimization of an RAE2822 airfoil. Results indicate that it has better robustness and local exploitation capability in comparison with those of other SBO methods.     

通用飞行器气动优化设计数字化集成平台——DIPasda

未来航空工业的发展，需要解决多学科综合设计的关键问题，为新型高性能飞行器的设计提供有力的设计方法和设计工具。DIPasda作为复杂外形设计的通用飞行器多学科优化设计平台，研制目的主要是提供一套新型通用、鲁棒、高效的优化设计架构，应用于通用飞行器工业设计环境，改善传统设计耗时低效的状况，提高新型飞行器设计的效率和精度。DIPasda平台系统包含了优化设计过程中所需用到的各类方法，主要包括数值优化算法、几何模型参数化方法、代理模型方法、高精度的学科分析工具等。通过详细介绍平台的系统架构、主要的功能模块、伴随优化设计和多目标优化设计流程，展现了DIPasda平台系统架构设计的灵活性和功能模块的完备性。最后通过优化算例展示了系统的综合优化设计能力。     

临近空间固体动力飞行器发动机与轨迹一体化设计优化

针对临近空间多级固体动力飞行器发动机与轨迹一体化设计优化问题,提出一种基于序列代理优化的高效设计方法。为了准确计算发动机的性能特性,对发动机进行了几何参数化建模,并针对复杂装药的燃面计算,提出了基于移动四面体的燃面计算算法。为了准确评估飞行器的最大航程能力,采用自适应Legendre-Gauss-Radau伪谱法获得给定发动机设计方案下的最大航程。为了提高发动机与轨迹一体化设计优化效率,提出了基于Kriging代理模型的多采样点高效全局代理优化算法,并进行了数值验证。计算结果表明：该优化方法收敛速度快,相比传统参数优化算法可以显著减少耗时目标函数和约束函数的计算次数,并能够有效地实现临近空间多级固体动力飞行器发动机与轨迹一体化设计优化。     

基于代理模型的高效全局气动优化设计方法研究进展

基于高可信度计算流体力学的数值优化设计方法，在提高飞行器气动与综合性能方面正发挥着越来越重要的作用。基于代理模型的优化算法（SBO），由于能够实现高效全局优化，逐渐成为了气动优化设计领域的研究热点之一。近20年来，代理优化算法研究已取得了长足进步，多种先进的新型代理模型被提出，优化理论和算法也不断完善和发展。以飞行器精细化气动优化设计为背景，综述了基于代理模型的高效全局气动优化设计方法研究进展。首先，介绍了基于变可信度代理模型的气动优化设计方法、结合代理模型和伴随方法的气动优化设计方法以及基于非生物进化的并行气动优化设计方法的研究现状和最新进展。然后，针对飞行器气动优化设计学科领域的前沿问题，介绍了基于代理模型的多目标气动优化设计方法、混合反设计/优化设计方法、稳健气动优化设计方法的研究进展，以及基于代理模型的多学科优化设计方法的研究进展。文献综述表明，代理优化算法在设计效率、全局性以及鲁棒性等方面性能优良，已经发展到可以解决100维（100个设计变量）以内的气动优化设计问题，具有良好的工程应用前景。最后，探讨了基于代理模型的高效全局气动优化设计在理论、方法及飞行器设计应用方面所面临的问题和挑战，给出了未来研究方向的建议。     

一种适用于气动优化的高效自适应全局优化方法

随着设计空间的增大和优化问题非线性程度的提高，基于代理模型的优化（SBO）过程收敛越来越慢，并且在局部勘测上呈现不足。本文发展了一种高效自适应全局优化方法，在整个样本细化迭代过程中采用变设计空间取样：即在每一步样本细化迭代过程中，利用当前设计空间中的样本建立代理模型，并且根据样本的内部特征，利用模糊聚类算法将该设计空间分割成几个子空间，然后在每个子空间内通过最大化目标函数的期望提高函数和最小化模型预测目标来增加新的样本，之后对子空间进行融合更新设计空间。6个解析测试算例的结果表明，所发展的方法相比于一般的代理模型优化方法，具有更好的鲁棒性以及全局探索和局部勘测能力，更适用于具有强非线性和多极值的优化问题。RAE2822气动优化实例表明，所发展的方法在处理工程实际问题时，仍然能够保持很好的效率、鲁棒性和自适应性。     

Robust design of NLF airfoils

 A robust optimization design approach of natural laminar airfoils is developed in this paper. First, the non-uniform rational B-splines (NURBS) free form deformation method based on NURBS basis function is introduced to the airfoil parameterization. Second, aerodynamic characteristics are evaluated by solving Navier-Stokes equations, and the γ-Re_θt transition model coupling with shear-stress transport (SST) turbulent model is introduced to simulate boundary layer transition. A numerical simulation of transition flow around NLF0416 airfoil is conducted to test the code. The comparison between numerical simulation results and wind tunnel test data approves the validity and applicability of the present transition model. Third, the optimization system is set up, which uses the separated particle swarm optimization (SPSO) as search algorithm and combines the Kriging models as surrogate model during optimization. The system is applied to carry out robust design about the uncertainty of lift coefficient and Mach number for NASA NLF-0115 airfoil. The data of optimized airfoil aerodynamic characteristics indicates that the optimized airfoil can maintain laminar flow stably in an uncertain range and has a wider range of low drag.     

Interactive Multi-objective Optimization Design for the Pylon Structure of an Airplane

The pylon structure of an airplane is very complex, and its high-fidelity analysis is quite time-consuming. If posterior preference optimization algorithm is used to solve this problem, the huge time consumption will be unacceptable in engineering practice due to the large amount of evaluation needed for the algorithm. So, a new interactive optimization algorithm-interactive multi-objective particle swarm optimization （IMOPSO） is presented. IMOPSO is efficient, simple and operable. The decision-maker can expediently determine the accurate preference in IMOPSO. IMOPSO is used to perform the pylon structure optimization design of an airplane, and a satisfactory design is achieved after only 12 generations of IMOPSO evolutions. Compared with original design, the maximum displacement of the satisfactory design is reduced, and the mass of the satisfactory design is decreased for 22%.     

Sequential ensemble optimization based on general surrogate model prediction variance and its application on engine acceleration schedule design

The Efficient Global Optimization (EGO) algorithm has been widely used in the numerical design optimization of engineering systems. However, the need for an uncertainty estimator limits the selection of a surrogate model. In this paper, a Sequential Ensemble Optimization (SEO) algorithm based on the ensemble model is proposed. In the proposed algorithm, there is no limitation on the selection of an individual surrogate model. Specifically, the SEO is built based on the EGO by extending the EGO algorithm so that it can be used in combination with the ensemble model. Also, a new uncertainty estimator for any surrogate model named the General Uncertainty Estimator (GUE) is proposed. The performance of the proposed SEO algorithm is verified by the simulations using ten well-known mathematical functions with varying dimensions. The results show that the proposed SEO algorithm performs better than the traditional EGO algorithm in terms of both the final optimization results and the convergence rate. Further, the proposed algorithm is applied to the global optimization control for turbo-fan engine acceleration schedule design.     

Aeroacoustic and aerodynamic optimization of propeller blades

It is of great significance to develop a high-efficiency and low-noise propeller optimization method for new-generation propeller aircraft design. Coupled with free form deformation method, dynamic mesh interpolation technology, optimization algorithm, surrogate model, aerodynamic calculation and aeroacoustic prediction model module, the integrated aerodynamic and aeroacoustic design method of propeller is built. The optimization design for the six-blade propeller is carried out. The non-reduction in efficiency, thrust coefficient and the minimum of aerodynamic noise is treated as the optimization design objective. The spatial vorticity distribution of the propeller before and after the design is also analyzed by using unsteady computational fluid dynamics method. The results show that the optimized propeller can effectively reduce the aerodynamic noise level. The maximum total sound pressure level can be reduced by 5 dB without reducing its aerodynamic performance. The developed method has good application potential in low-noise optimization design of propeller and other rotating machinery.     

Adaptive data fusion framework for modeling of non-uniform aerodynamic data

Multi-fidelity Data Fusion(MDF) frameworks have emerged as a prominent approach to producing economical but accurate surrogate models for aerodynamic data modeling by integrating data with different fidelity levels. However, most existing MDF frameworks assume a uniform data structure between sampling data sources; thus, producing an accurate solution at the required level, for cases of non-uniform data structures is challenging. To address this challenge, an Adaptive Multi-fidelity Data Fusion(...     

考虑不确定性的飞机总体参数优化方法

由于在各种设计问题包括飞机概念设计中都存在一定的不确定性,因此在总体参数优化时有必要考虑这种不确定性。以大型客机总体参数优化设计为例,定义了考虑不确定性的飞机总体参数优化问题,该问题与传统飞机总体参数优化的区别是要进行不确定性分析。而不确定性分析的计算量过大,为此提出了一种渐进代理模型方法来解决这一难题。在建立代理模型时,通过连续成批地在设计空间的全局和局部均加入新样本点,不断提高代理模型的全局拟合精度,直至获得满意的代理模型为止。然后在优化过程中使用计算量小的代理模型。大型客机总体参数优化问题中含有5个设计变量,目标函数为起飞重量最轻,并需满足4个性能约束。考虑了不确定性后,不仅使目标值（起飞重量）对总体参数变化的敏感度有所减小,而且满足约束（设计要求）概率显著提高。     

Surrogate role of machine learning in motor-drive optimization for more-electric aircraft applications

Motor drives form an essential part of the electric compressors, pumps, braking and actuation systems in the More-Electric Aircraft (MEA). In this paper, the application of Machine Learning (ML) in motor-drive design and optimization process is investigated. The general idea of using ML is to train surrogate models for the optimization. This training process is based on sample data collected from detailed simulation or experiment of motor drives. However, the Surrogate Role (SR) of ML may vary for different applications. This paper first introduces the principles of ML and then proposes two SRs (direct mapping approach and correction approach) of the ML in a motor-drive optimization process. Two different cases are given for the method comparison and validation of ML SRs. The first case is using the sample data from experiments to train the ML surrogate models. For the second case, the joint-simulation data is utilized for a multi-objective motor-drive optimization problem. It is found that both surrogate roles of ML can provide a good mapping model for the cases and in the second case, three feasible design schemes of ML are proposed and validated for the two SRs. Regarding the time consumption in optimizaiton, the proposed ML models can give one motor-drive design point up to 0.044 s while it takes more than 1.5 mins for the used simulation-based models.     

Surrogate-based analysis and optimization

A major challenge to the successful full-scale development of modern aerospace systems is to address competing objectives such as improved performance, reduced costs, and enhanced safety. Accurate, high-fidelity models are typically time consuming and computationally expensive. Furthermore, informed decisions should be made with an understanding of the impact (global sensitivity) of the design variables on the different objectives. In this context, the so-called surrogate-based approach for analysis and optimization can play a very valuable role. The surrogates are constructed using data drawn from high-fidelity models, and provide fast approximations of the objectives and constraints at new design points, thereby making sensitivity and optimization studies feasible. This paper provides a comprehensive discussion of the fundamental issues that arise in surrogate-based analysis and optimization (SBAO), highlighting concepts, methods, techniques, as well as practical implications. The issues addressed include the selection of the loss function and regularization criteria for constructing the surrogates, design of experiments, surrogate selection and construction, sensitivity analysis, convergence, and optimization. The multi-objective optimal design of a liquid rocket injector is presented to highlight the state of the art and to help guide future efforts.