Hwisang Jo, Byeong-uk Song, Joon-Yong Huh, Seungkyu Lee, Ikjin Lee
Multi-fidelity surrogate (MFS) modeling technology, which efficiently constructs surrogate models using low-fidelity (LF) and high-fidelity (HF) data, has been studied to enhance the predictive capability of engineering performances. In addition, several neural network (NN) structures for MFS modeling have been introduced, benefiting from recent developments in deep learning research. However, existing multi-fidelity (MF) NNs have been developed assuming identical sets of input variables for LF and HF data, a condition that is often not met in practical engineering systems. Therefore, this study proposes a new structure of composite NN designed for MF data with different input variables. The proposed network structure includes an input mapping network that connects the LF and HF data's input variables. Even when the physical relationship between these variables is unknown, the input mapping network can be concurrently trained during the process of training the whole network model. Customized loss functions and activation variables are suggested in this study to facilitate forward and backward propagation for the proposed NN structures when training MF data with different inputs. The effectiveness of the proposed method, in terms of prediction accuracy, is demonstrated through mathematical examples and practical engineering problems related to tire performances. The results confirm that the proposed method offers better accuracy than existing surrogate models in most problems. Moreover, the proposed method proves advantageous for surrogate modeling of nonlinear or discrete functions, a characteristic feature of NN-based methods.
多保真度代用（MFS）建模技术是一种利用低保真度（LF）和高保真度（HF）数据有效构建代用模型的技术，已被用于提高工程性能的预测能力。此外，得益于深度学习研究的最新发展，一些用于 MFS 建模的神经网络（NN）结构也被引入。然而，现有的多保真度（MF）神经网络是在假设低频和高频数据的输入变量集完全相同的情况下开发的，而这一条件在实际工程系统中往往无法满足。因此，本研究提出了一种新的复合网络结构，专为具有不同输入变量的 MF 数据而设计。建议的网络结构包括一个连接低频和高频数据输入变量的输入映射网络。即使这些变量之间的物理关系未知，也可以在训练整个网络模型的过程中同时训练输入映射网络。本研究提出了定制的损失函数和激活变量，以便在训练具有不同输入的中频数据时，为所提出的网络结构提供前向和后向传播。通过与轮胎性能相关的数学实例和实际工程问题，证明了所提方法在预测精度方面的有效性。结果证实，在大多数问题上，所提出的方法比现有的代用模型具有更高的准确性。此外，所提出的方法在非线性或离散函数的代用建模方面具有优势，这也是基于 NN 方法的一个特点。
{"title":"Modified structure of deep neural network for training multi-fidelity data with non-common input variables","authors":"Hwisang Jo, Byeong-uk Song, Joon-Yong Huh, Seungkyu Lee, Ikjin Lee","doi":"10.1115/1.4064782","DOIUrl":"https://doi.org/10.1115/1.4064782","url":null,"abstract":"\u0000 Multi-fidelity surrogate (MFS) modeling technology, which efficiently constructs surrogate models using low-fidelity (LF) and high-fidelity (HF) data, has been studied to enhance the predictive capability of engineering performances. In addition, several neural network (NN) structures for MFS modeling have been introduced, benefiting from recent developments in deep learning research. However, existing multi-fidelity (MF) NNs have been developed assuming identical sets of input variables for LF and HF data, a condition that is often not met in practical engineering systems. Therefore, this study proposes a new structure of composite NN designed for MF data with different input variables. The proposed network structure includes an input mapping network that connects the LF and HF data's input variables. Even when the physical relationship between these variables is unknown, the input mapping network can be concurrently trained during the process of training the whole network model. Customized loss functions and activation variables are suggested in this study to facilitate forward and backward propagation for the proposed NN structures when training MF data with different inputs. The effectiveness of the proposed method, in terms of prediction accuracy, is demonstrated through mathematical examples and practical engineering problems related to tire performances. The results confirm that the proposed method offers better accuracy than existing surrogate models in most problems. Moreover, the proposed method proves advantageous for surrogate modeling of nonlinear or discrete functions, a characteristic feature of NN-based methods.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"58 11","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139960792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Phillip Stevenson, Christopher A. Mattson, John Salmon, Nile Hatch
Engineered products are often optimized based on engineering requirements, user requirements, cost, and performance. This strategy is well suited for most applications, but products designed specifically to improve the lives of users and communities would benefit from an approach that helps engineers optimize a product based also on its social impacts. This paper introduces several optimization problem formulations adapted from multi-stakeholder strategies in the economics and business management literature. Each optimization problem formulation optimizes a product's social impact according to the ideology and principles inherent to the adapted multi-stakeholder strategy. A case study is presented for a cassava peeling machine that is being developed for farmers in the Brazilian Amazon. Finally, the resulting peeler design configurations and social impacts are discussed to illustrate the advantages and disadvantages of each strategy.
{"title":"Optimizing Engineered Products for Their Social Impacts On Multiple Stakeholders","authors":"Phillip Stevenson, Christopher A. Mattson, John Salmon, Nile Hatch","doi":"10.1115/1.4064694","DOIUrl":"https://doi.org/10.1115/1.4064694","url":null,"abstract":"\u0000 Engineered products are often optimized based on engineering requirements, user requirements, cost, and performance. This strategy is well suited for most applications, but products designed specifically to improve the lives of users and communities would benefit from an approach that helps engineers optimize a product based also on its social impacts. This paper introduces several optimization problem formulations adapted from multi-stakeholder strategies in the economics and business management literature. Each optimization problem formulation optimizes a product's social impact according to the ideology and principles inherent to the adapted multi-stakeholder strategy. A case study is presented for a cassava peeling machine that is being developed for farmers in the Brazilian Amazon. Finally, the resulting peeler design configurations and social impacts are discussed to illustrate the advantages and disadvantages of each strategy.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"25 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139854324","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Phillip Stevenson, Christopher A. Mattson, John Salmon, Nile Hatch
Engineered products are often optimized based on engineering requirements, user requirements, cost, and performance. This strategy is well suited for most applications, but products designed specifically to improve the lives of users and communities would benefit from an approach that helps engineers optimize a product based also on its social impacts. This paper introduces several optimization problem formulations adapted from multi-stakeholder strategies in the economics and business management literature. Each optimization problem formulation optimizes a product's social impact according to the ideology and principles inherent to the adapted multi-stakeholder strategy. A case study is presented for a cassava peeling machine that is being developed for farmers in the Brazilian Amazon. Finally, the resulting peeler design configurations and social impacts are discussed to illustrate the advantages and disadvantages of each strategy.
{"title":"Optimizing Engineered Products for Their Social Impacts On Multiple Stakeholders","authors":"Phillip Stevenson, Christopher A. Mattson, John Salmon, Nile Hatch","doi":"10.1115/1.4064694","DOIUrl":"https://doi.org/10.1115/1.4064694","url":null,"abstract":"\u0000 Engineered products are often optimized based on engineering requirements, user requirements, cost, and performance. This strategy is well suited for most applications, but products designed specifically to improve the lives of users and communities would benefit from an approach that helps engineers optimize a product based also on its social impacts. This paper introduces several optimization problem formulations adapted from multi-stakeholder strategies in the economics and business management literature. Each optimization problem formulation optimizes a product's social impact according to the ideology and principles inherent to the adapted multi-stakeholder strategy. A case study is presented for a cassava peeling machine that is being developed for farmers in the Brazilian Amazon. Finally, the resulting peeler design configurations and social impacts are discussed to illustrate the advantages and disadvantages of each strategy.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"21 1‐3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139794876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this work, the Efficient Robust Global Optimization (ERGO) method is revisited with the aim of enhancing and expanding its existing capabilities. The original objective of ERGO was to address the computational challenges associated with optimization-under-uncertainty through the use of Bayesian optimization (BO). ERGO tackles robust optimization problems which are characterized by sensitivity in the objective function due to stochasticity in the design space. It does this by concurrently minimizing the mean and variance of the objective in a multi-objective setting. To handle the computational complexity arising from the uncertainty propagation, ERGO exploits the analytical expression of the surrogate model underlying BO. In this study, ERGO is extended to accommodate multiple objectives, incorporate an improved predictive error estimation approach, investigate the treatment of failed function evaluations, and explore the handling of stochastic parameters next to stochastic design variables. To evaluate the effectiveness of these improvements, the enhanced ERGO scheme is compared with the original method using an analytical test problem with varying dimensionality. Additionally, the novel optimization technique is applied to an aerodynamic design problem to validate its performance.
在这项工作中，我们重新审视了高效稳健全局优化（ERGO）方法，旨在增强和扩展其现有能力。ERGO的最初目标是通过使用贝叶斯优化法（BO）解决与不确定性下优化相关的计算难题。ERGO 可解决稳健优化问题，这些问题的特点是目标函数因设计空间的随机性而具有敏感性。为此，它在多目标设置中同时最小化目标的均值和方差。为了处理不确定性传播带来的计算复杂性，ERGO 利用了 BO 基础代理模型的分析表达式。在本研究中，ERGO 进行了扩展，以适应多目标、采用改进的预测误差估计方法、研究函数评估失败的处理方法，并探索如何处理随机设计变量旁边的随机参数。为了评估这些改进的有效性，我们使用一个不同维度的分析测试问题，将增强型 ERGO 方案与原始方法进行了比较。此外，还将新型优化技术应用于空气动力学设计问题，以验证其性能。
{"title":"ERGO-II: An Improved Bayesian Optimization Technique for Robust Design with Multiple Objectives, Failed Evaluations and Stochastic Parameters","authors":"Jolan Wauters","doi":"10.1115/1.4064674","DOIUrl":"https://doi.org/10.1115/1.4064674","url":null,"abstract":"\u0000 In this work, the Efficient Robust Global Optimization (ERGO) method is revisited with the aim of enhancing and expanding its existing capabilities. The original objective of ERGO was to address the computational challenges associated with optimization-under-uncertainty through the use of Bayesian optimization (BO). ERGO tackles robust optimization problems which are characterized by sensitivity in the objective function due to stochasticity in the design space. It does this by concurrently minimizing the mean and variance of the objective in a multi-objective setting. To handle the computational complexity arising from the uncertainty propagation, ERGO exploits the analytical expression of the surrogate model underlying BO. In this study, ERGO is extended to accommodate multiple objectives, incorporate an improved predictive error estimation approach, investigate the treatment of failed function evaluations, and explore the handling of stochastic parameters next to stochastic design variables. To evaluate the effectiveness of these improvements, the enhanced ERGO scheme is compared with the original method using an analytical test problem with varying dimensionality. Additionally, the novel optimization technique is applied to an aerodynamic design problem to validate its performance.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"20 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139862316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Engineering designers are tasked with increasingly complex problems necessitating the use and development of various supports for navigating complexity. Prescriptive design process models are one such tool. However, little research has explored how engineering designers perceive these models' recommendations for engagement in design work. In this initial exploratory study, we analyzed data from 18 individual semi-structured interviews with mechanical engineering students to identify participant perceptions. As many design process model visualizations lack explicit attention to some social and contextual dimensions, we sought to compare perceptions among two drawn from engineering texts and one that was developed with the intent to emphasize social dimensions. We identified five salient areas of participant perceptions of the design process models. Perceptions of the process models related to what designers should do (starting and moving through a design process, gathering information, prototyping, and evaluating or testing) and what they should consider (aspects of focus). Our collection of participant perceptions across the three process models suggests different design process models make perceptions of certain recommendations more salient than others. However, participant perceptions also varied for the same process model. We suggest several implications for design education and training based on participant perceptions of these three process models, particularly the importance of leveraging multiple design process models. The comprehensive descriptions of participant perceptions across five areas of design work provided through our initial study provide a foundation for further investigations bridging designers' perceptions to intent to behavior and, ultimately, design outcomes.
{"title":"A Comparative Analysis of Student Perceptions of Recommendations for Engagement in Design Processes","authors":"K. Dugan, Shanna Daly","doi":"10.1115/1.4064671","DOIUrl":"https://doi.org/10.1115/1.4064671","url":null,"abstract":"\u0000 Engineering designers are tasked with increasingly complex problems necessitating the use and development of various supports for navigating complexity. Prescriptive design process models are one such tool. However, little research has explored how engineering designers perceive these models' recommendations for engagement in design work. In this initial exploratory study, we analyzed data from 18 individual semi-structured interviews with mechanical engineering students to identify participant perceptions. As many design process model visualizations lack explicit attention to some social and contextual dimensions, we sought to compare perceptions among two drawn from engineering texts and one that was developed with the intent to emphasize social dimensions. We identified five salient areas of participant perceptions of the design process models. Perceptions of the process models related to what designers should do (starting and moving through a design process, gathering information, prototyping, and evaluating or testing) and what they should consider (aspects of focus). Our collection of participant perceptions across the three process models suggests different design process models make perceptions of certain recommendations more salient than others. However, participant perceptions also varied for the same process model. We suggest several implications for design education and training based on participant perceptions of these three process models, particularly the importance of leveraging multiple design process models. The comprehensive descriptions of participant perceptions across five areas of design work provided through our initial study provide a foundation for further investigations bridging designers' perceptions to intent to behavior and, ultimately, design outcomes.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"100 s5","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139801283","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mingfei Jiang, Z. Ming, Chuanhao Li, J. Allen, F. Mistree
In this paper, we address the following question: How can multi-robot self-organizing systems be designed so that they show the desired behaviors and are able to perform tasks specified by designers? Multi-robot self-organizing systems, e.g., swarm robots, have great potential for adapting when performing complex tasks in a changing environment. However, such systems are difficult to design due to the stochasticity of the system performance and the non-linearity between the local actions/interaction and the desired global behavior. In order to address this, in this paper we propose a framework for designing self-organizing systems using Multi-Agent Reinforcement Learning (MARL) and the compromise Decision-Support Problem (cDSP) construct. In this paper we present a framework that consists of two stages, namely, preliminary design and design improvement. In the preliminary design stage, MARL is used to help designers train the robots so that they show stable group behavior for performing the task. In the design improvement stage, the cDSP construct is used to explore the design space and identify satisfactory solutions considering several performance indicators. Between the two stages, surrogate models are used to map the relationship between local parameters and global performance indicators utilizing the data generated in preliminary design. A multi-robot box-pushing problem is used as an example to test the efficacy of the framework. The framework is general and can be extended to design other self-organizing systems. Our focus in this paper is in describing the framework.
{"title":"DESIGN OF SELF-ORGANIZING SYSTEMS USING MULTI-AGENT REINFORCEMENT LEARNING AND THE COMPROMISE DECISION SUPPORT PROBLEM CONSTRUCT","authors":"Mingfei Jiang, Z. Ming, Chuanhao Li, J. Allen, F. Mistree","doi":"10.1115/1.4064672","DOIUrl":"https://doi.org/10.1115/1.4064672","url":null,"abstract":"\u0000 In this paper, we address the following question: How can multi-robot self-organizing systems be designed so that they show the desired behaviors and are able to perform tasks specified by designers? Multi-robot self-organizing systems, e.g., swarm robots, have great potential for adapting when performing complex tasks in a changing environment. However, such systems are difficult to design due to the stochasticity of the system performance and the non-linearity between the local actions/interaction and the desired global behavior. In order to address this, in this paper we propose a framework for designing self-organizing systems using Multi-Agent Reinforcement Learning (MARL) and the compromise Decision-Support Problem (cDSP) construct. In this paper we present a framework that consists of two stages, namely, preliminary design and design improvement. In the preliminary design stage, MARL is used to help designers train the robots so that they show stable group behavior for performing the task. In the design improvement stage, the cDSP construct is used to explore the design space and identify satisfactory solutions considering several performance indicators. Between the two stages, surrogate models are used to map the relationship between local parameters and global performance indicators utilizing the data generated in preliminary design. A multi-robot box-pushing problem is used as an example to test the efficacy of the framework. The framework is general and can be extended to design other self-organizing systems. Our focus in this paper is in describing the framework.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"12 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139859327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Amin Heyrani Nobari, Justin Rey, S. Kodali, Matthew Jones, Faez Ahmed
In many design automation applications, accurate segmentation and classification of 3D surfaces and extraction of geometric insight from 3D models can be pivotal. This paper primarily introduces a machine learning-based scheme that leverages Graph Neural Networks (GNN) for handling 3D geometries, specifically for surface classification. Our model demonstrates superior performance against two state-of-the-art models, PointNet++ and PointMLP, in terms of surface classification accuracy, beating both models. Central to our contribution is the novel incorporation of conformal predictions, a method that offers robust uncertainty quantification and handling with marginal statistical guarantees. Unlike traditional approaches, conformal predictions enable our model to ensure precision, especially in challenging scenarios where mistakes can be highly costly. This robustness proves invaluable in design applications, and as a case in point, we showcase its utility in automating the Computational Fluid Dynamics (CFD) meshing process for aircraft models based on expert guidance. Our results reveal that our automatically generated mesh, guided by the proposed rules by experts enabled through the segmentation model, is not only efficient but matches the quality of expert-generated meshes, leading to accurate simulations. For the community's benefit, we have made our code and data available at https://github.com/ahnobari/AutoSurf Upon paper acceptance.
{"title":"MeshPointNet: 3D Surface Classification Using Graph Neural Networks and Conformal Predictions on Mesh-Based Representations","authors":"Amin Heyrani Nobari, Justin Rey, S. Kodali, Matthew Jones, Faez Ahmed","doi":"10.1115/1.4064673","DOIUrl":"https://doi.org/10.1115/1.4064673","url":null,"abstract":"\u0000 In many design automation applications, accurate segmentation and classification of 3D surfaces and extraction of geometric insight from 3D models can be pivotal. This paper primarily introduces a machine learning-based scheme that leverages Graph Neural Networks (GNN) for handling 3D geometries, specifically for surface classification. Our model demonstrates superior performance against two state-of-the-art models, PointNet++ and PointMLP, in terms of surface classification accuracy, beating both models. Central to our contribution is the novel incorporation of conformal predictions, a method that offers robust uncertainty quantification and handling with marginal statistical guarantees. Unlike traditional approaches, conformal predictions enable our model to ensure precision, especially in challenging scenarios where mistakes can be highly costly. This robustness proves invaluable in design applications, and as a case in point, we showcase its utility in automating the Computational Fluid Dynamics (CFD) meshing process for aircraft models based on expert guidance. Our results reveal that our automatically generated mesh, guided by the proposed rules by experts enabled through the segmentation model, is not only efficient but matches the quality of expert-generated meshes, leading to accurate simulations. For the community's benefit, we have made our code and data available at https://github.com/ahnobari/AutoSurf Upon paper acceptance.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"14 6","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139799077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this work, the Efficient Robust Global Optimization (ERGO) method is revisited with the aim of enhancing and expanding its existing capabilities. The original objective of ERGO was to address the computational challenges associated with optimization-under-uncertainty through the use of Bayesian optimization (BO). ERGO tackles robust optimization problems which are characterized by sensitivity in the objective function due to stochasticity in the design space. It does this by concurrently minimizing the mean and variance of the objective in a multi-objective setting. To handle the computational complexity arising from the uncertainty propagation, ERGO exploits the analytical expression of the surrogate model underlying BO. In this study, ERGO is extended to accommodate multiple objectives, incorporate an improved predictive error estimation approach, investigate the treatment of failed function evaluations, and explore the handling of stochastic parameters next to stochastic design variables. To evaluate the effectiveness of these improvements, the enhanced ERGO scheme is compared with the original method using an analytical test problem with varying dimensionality. Additionally, the novel optimization technique is applied to an aerodynamic design problem to validate its performance.
在这项工作中，我们重新审视了高效稳健全局优化（ERGO）方法，旨在增强和扩展其现有能力。ERGO的最初目标是通过使用贝叶斯优化法（BO）解决与不确定性下优化相关的计算难题。ERGO 可解决稳健优化问题，这些问题的特点是目标函数因设计空间的随机性而具有敏感性。为此，它在多目标设置中同时最小化目标的均值和方差。为了处理不确定性传播带来的计算复杂性，ERGO 利用了 BO 基础代理模型的分析表达式。在本研究中，ERGO 进行了扩展，以适应多目标、采用改进的预测误差估计方法、研究函数评估失败的处理方法，并探索如何处理随机设计变量旁边的随机参数。为了评估这些改进的有效性，我们使用一个不同维度的分析测试问题，将增强型 ERGO 方案与原始方法进行了比较。此外，还将新型优化技术应用于空气动力学设计问题，以验证其性能。
{"title":"ERGO-II: An Improved Bayesian Optimization Technique for Robust Design with Multiple Objectives, Failed Evaluations and Stochastic Parameters","authors":"Jolan Wauters","doi":"10.1115/1.4064674","DOIUrl":"https://doi.org/10.1115/1.4064674","url":null,"abstract":"\u0000 In this work, the Efficient Robust Global Optimization (ERGO) method is revisited with the aim of enhancing and expanding its existing capabilities. The original objective of ERGO was to address the computational challenges associated with optimization-under-uncertainty through the use of Bayesian optimization (BO). ERGO tackles robust optimization problems which are characterized by sensitivity in the objective function due to stochasticity in the design space. It does this by concurrently minimizing the mean and variance of the objective in a multi-objective setting. To handle the computational complexity arising from the uncertainty propagation, ERGO exploits the analytical expression of the surrogate model underlying BO. In this study, ERGO is extended to accommodate multiple objectives, incorporate an improved predictive error estimation approach, investigate the treatment of failed function evaluations, and explore the handling of stochastic parameters next to stochastic design variables. To evaluate the effectiveness of these improvements, the enhanced ERGO scheme is compared with the original method using an analytical test problem with varying dimensionality. Additionally, the novel optimization technique is applied to an aerodynamic design problem to validate its performance.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"53 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139802380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mingfei Jiang, Z. Ming, Chuanhao Li, J. Allen, F. Mistree
In this paper, we address the following question: How can multi-robot self-organizing systems be designed so that they show the desired behaviors and are able to perform tasks specified by designers? Multi-robot self-organizing systems, e.g., swarm robots, have great potential for adapting when performing complex tasks in a changing environment. However, such systems are difficult to design due to the stochasticity of the system performance and the non-linearity between the local actions/interaction and the desired global behavior. In order to address this, in this paper we propose a framework for designing self-organizing systems using Multi-Agent Reinforcement Learning (MARL) and the compromise Decision-Support Problem (cDSP) construct. In this paper we present a framework that consists of two stages, namely, preliminary design and design improvement. In the preliminary design stage, MARL is used to help designers train the robots so that they show stable group behavior for performing the task. In the design improvement stage, the cDSP construct is used to explore the design space and identify satisfactory solutions considering several performance indicators. Between the two stages, surrogate models are used to map the relationship between local parameters and global performance indicators utilizing the data generated in preliminary design. A multi-robot box-pushing problem is used as an example to test the efficacy of the framework. The framework is general and can be extended to design other self-organizing systems. Our focus in this paper is in describing the framework.
{"title":"DESIGN OF SELF-ORGANIZING SYSTEMS USING MULTI-AGENT REINFORCEMENT LEARNING AND THE COMPROMISE DECISION SUPPORT PROBLEM CONSTRUCT","authors":"Mingfei Jiang, Z. Ming, Chuanhao Li, J. Allen, F. Mistree","doi":"10.1115/1.4064672","DOIUrl":"https://doi.org/10.1115/1.4064672","url":null,"abstract":"\u0000 In this paper, we address the following question: How can multi-robot self-organizing systems be designed so that they show the desired behaviors and are able to perform tasks specified by designers? Multi-robot self-organizing systems, e.g., swarm robots, have great potential for adapting when performing complex tasks in a changing environment. However, such systems are difficult to design due to the stochasticity of the system performance and the non-linearity between the local actions/interaction and the desired global behavior. In order to address this, in this paper we propose a framework for designing self-organizing systems using Multi-Agent Reinforcement Learning (MARL) and the compromise Decision-Support Problem (cDSP) construct. In this paper we present a framework that consists of two stages, namely, preliminary design and design improvement. In the preliminary design stage, MARL is used to help designers train the robots so that they show stable group behavior for performing the task. In the design improvement stage, the cDSP construct is used to explore the design space and identify satisfactory solutions considering several performance indicators. Between the two stages, surrogate models are used to map the relationship between local parameters and global performance indicators utilizing the data generated in preliminary design. A multi-robot box-pushing problem is used as an example to test the efficacy of the framework. The framework is general and can be extended to design other self-organizing systems. Our focus in this paper is in describing the framework.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"279 12","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139799380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The development of product platform is an effective strategy to respond to dynamic market demands, decrease lead-time and delay products differentiation. However, the traditional product platform configuration method can not satisfy the sustainability requirements for modern products. To solve this problem, a sustainable product multi-platform (SPMP) model for assembly/ disassembly technology is proposed in this paper. The proposed SPMP model measures the energy consumption of module instances during the installation based on the platform-based assembly index (PAI) and platform-based disassembly index (PDI), and provides a multi-platform solution for the assembly of product family. To demonstrate the effectiveness of the proposed method, two product family cases are discussed. Simplified case shows that multi-objective particle swarm optimisation (MOPSO) algorithm has stronger optimisation ability than linear programming method in reducing product processing cost. The hair dryer family case demonstrates that the proposed method reduces the energy consumption during assembly by linking sustainability to product design.
{"title":"A Sustainable Product Multi-platform Planning Model for Assembly and Disassembly Process","authors":"Guang-yu Zou, Zhongkai Li, Chao He","doi":"10.1115/1.4064675","DOIUrl":"https://doi.org/10.1115/1.4064675","url":null,"abstract":"\u0000 The development of product platform is an effective strategy to respond to dynamic market demands, decrease lead-time and delay products differentiation. However, the traditional product platform configuration method can not satisfy the sustainability requirements for modern products. To solve this problem, a sustainable product multi-platform (SPMP) model for assembly/ disassembly technology is proposed in this paper. The proposed SPMP model measures the energy consumption of module instances during the installation based on the platform-based assembly index (PAI) and platform-based disassembly index (PDI), and provides a multi-platform solution for the assembly of product family. To demonstrate the effectiveness of the proposed method, two product family cases are discussed. Simplified case shows that multi-objective particle swarm optimisation (MOPSO) algorithm has stronger optimisation ability than linear programming method in reducing product processing cost. The hair dryer family case demonstrates that the proposed method reduces the energy consumption during assembly by linking sustainability to product design.","PeriodicalId":506672,"journal":{"name":"Journal of Mechanical Design","volume":"28 8","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139798337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：