Large-scale multidisciplinary design optimization—review and recommendations

Hwang, John T.; Jain, Arnav; Ha, Tae Hyon

doi:10.2514/6.2019-3106

Cited by 12 publications

(8 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This requires new optimizers and modeling frameworks that can facilitate a new paradigm where the solvers in the model could be instructed by the optimizer to partially converge to specified tolerances to speed up optimization. to be an active application area for large-scale MDO, both with OpenMDAO [42] and without it, where algorithmic di erentiation is used instead of the adjoint method [43].…”

Section: Optimizer Modelmentioning

confidence: 99%

A hybrid architecture for large-scale system design optimization of PDE-based models

Joshy

Yan

Hwang

2022

AIAA SCITECH 2022 Forum

View full text Add to dashboard Cite

Large-scale system design optimization has recently started being used in many fields as a tool in the design of new concepts. This has exposed many of its limitations, especially in regard to its usability and computational efficiency. State-of-the-art optimization algorithms can solve problems with tens of thousands of design variables in just hundreds of model evaluations. However, this can be computationally expensive for complex systems with costly simulations. A new, hybrid algorithm called SURF, which unifies the full-space and reduced-space architectures, can accelerate optimization of computational models of engineering systems involving large numbers of implicitly defined state variables. However, the theory and numerical results for SURF were restricted to equality-constrained problems solved using the method of Newton-Lagrange. This paper investigates the extension of SURF to the inequality-constrained optimization setting. We first extend the theory of SURF to bring sequential quadratic programming under the scope of SURF. Based on this extended theory, we propose an implementation of the SURF algorithm for inequality-constrained problems. We then test the hypothesis that this algorithm unifies the full-space and reduced-space algorithms, using a nonlinear topology optimization problem. Finally, we measure the improvement provided by SURF over the reduced-space method, using the same optimization problem. The results from the investigation suggest that the proposed new algorithm unifies the full-space and reduced-space methods for any constrained optimization setting. We also observe a reduction in the total number of linear system solutions required in solving the nonlinear systems within the model. However, we expect an actual improvement in the overall optimization time only with an industrial quality sequential quadratic programming optimizer implemented with SURF since the optimizer used in this study did not scale well with the number of design variables thus inherently favoring the reduced-space approach. Nomenclature= partial derivative of a function F with respect to a variable x d d = total derivative of a function F with respect to a variable x ∇ ( ) = gradient of a function F with respect to a variable x

show abstract

Section: Optimizer Modelmentioning

confidence: 99%

A hybrid architecture for large-scale system design optimization of PDE-based models

Joshy

Yan

Hwang

2022

AIAA SCITECH 2022 Forum

View full text Add to dashboard Cite

show abstract

“…12 Later, this approach was extended to multi-disciplinary optimization, which considers systems that involve several disciplines and subsystems. 13,14 Consideration of the assembly process during design optimization attracts significantly less attention compared to performance-based optimization, and hardly any widely used software for assembly process analysis exists. Assembly process consideration and related research, initiated in the 1960s, resulted in design for assembly (DFA) guidelines, which were manually used by companies and designers when necessary.…”

Section: Introductionmentioning

confidence: 99%

“…12 Later, this approach was extended to multi-disciplinary optimization, which considers systems that involve several disciplines and subsystems. 13,14…”

Section: Introductionmentioning

confidence: 99%

Framework for metamodel-based design optimization considering product performance and assembly process complexity

Eremeev,

Cock,

Devriendt

et al. 2024

SIMULATION

View full text Add to dashboard Cite

This paper proposes a method for simultaneous evaluation of the assembly process complexity together with the performance of the future product. It allows for product design optimization, considering different aspects of the future design at the early stage of the development process. The proposed method, embodied in a fully automated framework, substitutes the traditional sequential development process with a more efficient and rapid combined procedure, which addresses multiple design aspects simultaneously. Design for assembly (DFA) rules, used as quantitative metrics of the ease-of-assembly of the whole product and individual assembly operations, are automatically evaluated together with performance metrics, estimated based on finite element (FE) simulations. The direct solution to this optimization problem might be inefficient or impossible since it requires the recurrent evaluation of computationally expensive discrete and continuous functions with unknown behavior that represent the optimization objectives and constraints. For that reason, the proposed framework employs regression models based on the Gaussian process and artificial neural networks, thus achieving the optimal design of a product as a result of metamodel-based design optimization (MBDO). The suggested approach is demonstrated in the optimization of a gearbox assembly, considering its mechanical performance and assembly process. Comparing the results of the metamodel-based and direct design optimization shows that MBDO allows finding a better solution using a three times smaller computational budget. In addition, analysis of the results obtained using stationary sampling data sets of different sizes highlighted the limitations of the employed sampling procedure.

show abstract

“…Note that the GT is only performed on the globally best individual and is just carried out as a simple operator that is added to the standard DE. Therefore, compared with other large-scale optimization algorithms [35]- [39], GTDE not only maintains the simple algorithm structure of standard DE, but also avoids the sensitive decomposition strategy and the complicated learning strategy or local search technique, which is simpler to understand and easier to implement.…”

Section: Introduction Ifferential Evolution (De) Proposed By Storn An...mentioning

confidence: 99%

Gene Targeting Differential Evolution: A Simple and Efficient Method for Large-Scale Optimization

Wang

Jian

Zhan

et al. 2023

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

Large-scale optimization problems (LSOPs) are challenging because the algorithm is difficult in balancing too many dimensions and in escaping from trapped bottleneck dimensions. To improve solutions, this paper introduces targeted modification to the certain values in the bottleneck dimensions. Analogous to gene targeting (GT) in biotechnology, we experiment on targeting the specific genes in candidate solution to improve its trait in differential evolution (DE). We propose a simple and efficient method, called GT-based DE (GTDE), to solve LSOPs. In the algorithm design, a simple GT-based modification is developed to perform on the best individual, comprising probabilistically targeting the location of bottleneck dimensions, constructing a homologous targeting vector, and inserting the targeting vector into the best individual. In this way, all the bottleneck dimensions of the best individual can be probabilistically targeted and modified to break the bottleneck and to provide global guidance for more optimal evolution. Note that the GT is only performed on the globally best individual and is just carried out as a simple operator that is added to the standard DE. Experimental studies compare the GTDE with some other state-of-the-art large-scale optimization algorithms, including the winners of CEC2010, CEC2012, CEC2013, and CEC2018 competitions on large-scale optimization. The results show that the GTDE is efficient and performs better than or at least comparable to the others in solving LSOPs.Manuscript received….revised….accepted….

show abstract

Large-scale multidisciplinary design optimization—review and recommendations

Cited by 12 publications

References 41 publications

A hybrid architecture for large-scale system design optimization of PDE-based models

A hybrid architecture for large-scale system design optimization of PDE-based models

Framework for metamodel-based design optimization considering product performance and assembly process complexity

Gene Targeting Differential Evolution: A Simple and Efficient Method for Large-Scale Optimization

Contact Info

Product

Resources

About