Simultaneous structure design optimization of multiple car models using the K computer

Oyama, Akira; Kohira, Takehisa; Kemmotsu, Hiromasa; Tatsukawa, Tomoaki; Watanabe, Takeshi

doi:10.1109/ssci.2017.8285350

Cited by 27 publications

(13 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) are not experts in multi-objective evolutionary computation methods and they apply MOEA solvers using fixed (i.e., recommended) parameterizations; (2) tackle complicated industrial problems where even state-ofthe-art surrogate-based modeling techniques [19] are not fully applicable when wishing to reduce the reliance on intensive simulations [18] and thus the total number of optimization runs and of individuals that can be evaluated during these runs is rather limited; (3) often adopt a form of interactive usage of MOEA-based optimization where (for computationally-intensive problems) the solver is stopped as soon as it has discovered a sufficiently good/interesting PN [21].…”

Section: Motivation and Approachmentioning

confidence: 99%

“…During the past two decades, in light of their intrinsic ability to produce entire PNs after a single run, multi-objective optimization algorithms (MOEAs) have emerged as one of the most successful methods for solving MOOPs [1]. Although numerous valuable contributions have been made over the years, three main evolutionary GECCO '18, July [15][16][17][18][19]2018, Kyoto, Japan A.-C. Zăvoianu et al…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Two enhancements for improving the convergence speed of a robust multi-objective coevolutionary algorithm

Zăvoianu

Saminger‐Platz

Lughofer

et al. 2018

Proceedings of the Genetic and Evolutionary Computation Conference

View full text Add to dashboard Cite

We describe two enhancements that significantly improve the rapid convergence behavior of DECMO2 -a previously proposed robust coevolutionary algorithm that integrates three different multiobjective space exploration paradigms: differential evolution, twotier Pareto-based selection for survival and decomposition-based evolutionary guidance. The first enhancement is a refined active search adaptation mechanism that relies on run-time sub-population performance indicators to estimate the convergence stage and dynamically adjust and steer certain parts of the coevolutionary process in order to improve its overall efficiency. The second enhancement consists in a directional intensification operator that is applied in the early part of the run during the decomposition-based search phases. This operator creates new random local linear individuals based on the recent historically successful solution candidates of a given directional decomposition vector. As the two efficiencyrelated enhancements are complementary, our results show that the resulting coevolutionary algorithm is a highly competitive improvement of the baseline strategy when considering a comprehensive test set aggregated from 25 (standard) benchmark multi-objective optimization problems.

show abstract

Section: Motivation and Approachmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Two enhancements for improving the convergence speed of a robust multi-objective coevolutionary algorithm

Zăvoianu

Saminger‐Platz

Lughofer

et al. 2018

Proceedings of the Genetic and Evolutionary Computation Conference

View full text Add to dashboard Cite

show abstract

“…Because of their ability to discover high-quality PS approximations called Pareto non-dominated sets (PNs) in single runs, multi-objective evolutionary algorithms (MOEAs) have emerged as some of the most successful MOOP solvers [3]. As an increasing number of MOEA practitioners (e.g., mechatronic engineers [19], industrial designers [16], quality assurance analysts [28]) are tackling ever more challenging real-world problems, difficulties stemming from relying on experimentation/simulation-driven F (x) values are being brought to the forefront. A costly evaluation of solution candidate quality (i.e., fitness) greatly reduces the number of fitness evaluations (nfe) that can be computed during an optimisation and runs might be stopped prematurely.…”

Section: Introductionmentioning

confidence: 99%

Comparative Run-Time Performance of Evolutionary Algorithms on Multi-objective Interpolated Continuous Optimisation Problems

Zăvoianu

Lacroix

McCall

2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

We propose a new class of multi-objective benchmark problems on which we analyse the performance of four well established multiobjective evolutionary algorithms (MOEAs) -each implementing a different search paradigm -by comparing run-time convergence behaviour over a set of 1200 problem instances. The new benchmarks are created by fusing previously proposed single-objective interpolated continuous optimisation problems (ICOPs) via a common set of Pareto non-dominated seeds. They thus inherit the ICOP property of having tunable fitness landscape features. The benchmarks are of intrinsic interest as they derive from interpolation methods and so can approximate general problem instances. This property is revealed to be of particular importance as our extensive set of numerical experiments indicates that choices pertaining to (i) the weighting of the inverse distance interpolation function and (ii) the problem dimension can be used to construct problems that are challenging to all tested multi-objective search paradigms. This in turn means that the new multi-objective ICOPs problems (MO-ICOPs) can be used to construct well-balanced benchmark sets that discriminate well between the run-time convergence behaviour of different solvers.

show abstract

“…Evolutionary algorithms (EAs) have been applied to a wide range of real-world optimization problems such as engineering [23], pathfinding [25], data mining [29], T. Harada 6-6 Asahigaoka, Hino, Tokyo, Japan Faculty of System Design, Tokyo Metropolitan University E-mail: harada@tmu.ac.jp nanoscience [28], power system [32], and so on because of their high search capability without any problemspecific knowledge. When applying EAs to real-world optimization problems, solution evaluations may take much computational time due to physical simulation or actual consumption time measurement.…”

Section: Introductionmentioning

confidence: 99%

Search Progress Dependent Parent Selection for Avoiding Evaluation Time Bias in Asynchronous Parallel Multi-Objective Evolutionary Algorithms

Harada

2020

2020 IEEE Symposium Series on Computational Intelligence (SSCI)

View full text Add to dashboard Cite

This paper proposes a new parent selection method for reducing the effect of evaluation time bias in asynchronous parallel evolutionary algorithms (APEAs). APEAs have the advantage of increasing computational efficiency even when the evaluation times of solutions differ. However, APEAs have a problem that their search direction is biased toward the search region with a short evaluation time. The proposed parent selection method considers the search frequency of solutions to reduce such an adverse influence of APEAs while maintaining their computational efficiency. We conduct experiments on toy problems that reproduce the evaluation time bias on multi-objective optimization problems to investigate the effectiveness of the proposed method. The experiments use NSGA-III, a well-known multi-objective evolutionary algorithm. In the experiments, we compare the proposed method with the synchronous and asynchronous methods. The experimental results reveal that the proposed method can reduce the effect of the evaluation time bias while reducing the computing time of the parallel NSGA-III.

show abstract

Simultaneous structure design optimization of multiple car models using the K computer

Cited by 27 publications

References 1 publication

Two enhancements for improving the convergence speed of a robust multi-objective coevolutionary algorithm

Two enhancements for improving the convergence speed of a robust multi-objective coevolutionary algorithm

Comparative Run-Time Performance of Evolutionary Algorithms on Multi-objective Interpolated Continuous Optimisation Problems

Search Progress Dependent Parent Selection for Avoiding Evaluation Time Bias in Asynchronous Parallel Multi-Objective Evolutionary Algorithms

Contact Info

Product

Resources

About