Volume 3B: 48th Design Automation Conference (DAC)最新文献_第4页

Experimental Investigation of Topology-Optimized Beams With Isotropic and Anisotropic Base Material Assumptions 各向同性和各向异性基材假设下拓扑优化梁的实验研究

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-89001

Hajin J. Kim, J. Carstensen

{"title":"Experimental Investigation of Topology-Optimized Beams With Isotropic and Anisotropic Base Material Assumptions","authors":"Hajin J. Kim, J. Carstensen","doi":"10.1115/detc2022-89001","DOIUrl":"https://doi.org/10.1115/detc2022-89001","url":null,"abstract":"\u0000 Additive Manufacturing (AM) technologies are promising fabrication methods with the potential to increase customizability and structural complexity. It is well established that the nature of AM typically results in base materials that exhibit an extent of anisotropy. Since topology optimization is a freeform approach that generally achieves high performing designs, it is often suggested as a powerful design-for-AM method. However, most topology optimization frameworks ignore anisotropic effects and assume isotropy of the base material. Although frameworks that consider anisotropy have been suggested, the influence anisotropy has on the physical behavior of fabricated designs is not well understood. Therefore, this work presents an experimental study of topology-optimized structures designed with both isotropic and anisotropic linear elastic material assumptions to explore how much anisotropic considerations matter when it comes to the discrepancy in numerical and experimental performance. The experimental investigation is conducted using a Fused Filament Fabrication print process that allows us to prescribe the anisotropy. The Young’s Modulus of the designated print setup is experimentally determined and used for design of 3D simply supported beams with various material volumes. Samples are fabricated and evaluated using 3-point bending tests. It is seen that the isotropic designs have a slightly better average performance at the design load (1.8–2.0%), but that inclusion of the anisotropic behavior significantly limits behavioral differences across samples (84.4–171.5% decrease in standard deviation) and improves the print success rate.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126822510","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}

引用次数: 0

Holistic Optimal Design of Face-Milled Hypoid Gearsets 面铣准双曲面齿轮组整体优化设计

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-89598

Eugeniu Grabovic, Alessio Artoni, M. Gabiccini

{"title":"Holistic Optimal Design of Face-Milled Hypoid Gearsets","authors":"Eugeniu Grabovic, Alessio Artoni, M. Gabiccini","doi":"10.1115/detc2022-89598","DOIUrl":"https://doi.org/10.1115/detc2022-89598","url":null,"abstract":"\u0000 The aim of this paper is to showcase a holistic approach to design optimized spiral bevel and hypoid gearsets. The first step is the definition of the gear and pinion blanks starting from the basic transmission data. The second step is to synthesize the basic machine-tool settings required to cut the two toothed members. The gear machine-tool settings are obtained first, whereas the basic pinion settings are identified by minimizing the deviations of the pinion tooth surface from the gear conjugate member accounting for the presence of misalignments. The proposed novel identification strategy can handle all the higher-order motions while offering a remarkable speedup with respect to existing techniques. The result of the macro-geometry design phase is a conjugate spiral bevel or hypoid gearset.\u0000 As a last step, the design of the pinion micro-geometry is formulated as a multi-objective optimization problem where the obtained optimal ease-off is guaranteed to be manufacturable. To this end, an original strategy is devised where the search for the optimal tooth surface is driven by the coefficients of a free-form polynomial representation of its micro-topography. However, the tooth geometry evaluated by the loaded tooth contact solver is actually its closest manufacturable analogue. A fully worked out numerical test case substantiates the whole method.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116250510","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}

引用次数: 2

Multi-Objective Bayesian Optimization Supported by Gaussian Process Classifiers and Conditional Probabilities 基于高斯过程分类器和条件概率的多目标贝叶斯优化

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-91343

H. Valladares, A. Tovar

{"title":"Multi-Objective Bayesian Optimization Supported by Gaussian Process Classifiers and Conditional Probabilities","authors":"H. Valladares, A. Tovar","doi":"10.1115/detc2022-91343","DOIUrl":"https://doi.org/10.1115/detc2022-91343","url":null,"abstract":"\u0000 In the last years, there has been an increasing effort to develop Bayesian methodologies to solve multi-objective optimization problems. Most of these methods can be classified in two groups: infilling criterion-based methods and aggregation-based methods. The first group employs an index that quantifies the gain that a new design can produce in the current Pareto front while the last group uses a (possibly non-linear) aggregation function and a weighting vector to identify a Pareto design. Most infilling-based methods have been developed to solve two-objective optimization problems. Aggregation-based methods enable the solution of many-objective optimization problems but their performance depends on the set of weighting vectors, which are often selected randomly. This study proposes a novel multi-objective Bayesian framework that exploits the rich probabilistic information that can be extracted from Gaussian process (GP) classifiers and the ability of conditional probabilities to capture design preferences. In the proposed framework, a GP classifier is trained to identify design zones that potentially contain Pareto designs. The training process involves the inference of a latent GP that encodes input-space interactions that describe a Pareto design. This latent GP enables the solution of many-objective optimization problems with any standard acquisition function and without the prescription of a weighting vector. Conditional probabilities are utilized to define design goals that promote a uniform expansion of the Pareto front. The proposed approach is demonstrated with two benchmark analytical problems and the design optimization of sandwich composite armors for blast mitigation, which involves expensive finite element simulations.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"299 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114853884","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}

引用次数: 0

Hall Effect Sensor Design Optimization With Multi-Physics Informed Gaussian Process Modeling 基于多物理场高斯过程建模的霍尔效应传感器设计优化

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-91196

Yanwen Xu, Zhuoyuan Zheng, Kanika Arora, D. Senesky, Pingfeng Wang

{"title":"Hall Effect Sensor Design Optimization With Multi-Physics Informed Gaussian Process Modeling","authors":"Yanwen Xu, Zhuoyuan Zheng, Kanika Arora, D. Senesky, Pingfeng Wang","doi":"10.1115/detc2022-91196","DOIUrl":"https://doi.org/10.1115/detc2022-91196","url":null,"abstract":"\u0000 Magnetic field sensor devices have been widely used to track changes in magnetic flux concentration, and the Hall sensors are promising in many engineering applications. Design optimization of the Hall effect sensor is required to ensure the quality and capability of the device when in service. Even though there has been empirical models established from experiments to guide the design of the Hall effect sensor, the underlying relationship in Hall effect sensor design parameters and corresponding performances has not been looked into thoroughly. This paper presents a physics-informed machine learning technique to optimize the geometry design of Hall magnetic sensors for a low offset and high sensitivity characteristic. Multi-physics based finite element models were first developed to simulate and predict the Hall voltage, offset voltage and sensor sensitivity of different Hall effect sensors with various geometries. In addition, to improve the design efficiency, Gaussian Process (GP) based surrogate models were constructed from multiphysics-based simulation results to effectively investigate the Hall sensor performances with an adaptive sampling strategy. Three types of geometries of Hall sensor were studied and optimized with the proposed physics-informed GP model, the obtained results were consistent with the empirical experimental result.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121639488","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}

引用次数: 1

Automated and Customized CAD Drawings by Utilizing Machine Learning Algorithms: A Case Study 自动化和定制CAD绘图利用机器学习算法:一个案例研究

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-88971

Javier Villena Toro, M. Tarkian

{"title":"Automated and Customized CAD Drawings by Utilizing Machine Learning Algorithms: A Case Study","authors":"Javier Villena Toro, M. Tarkian","doi":"10.1115/detc2022-88971","DOIUrl":"https://doi.org/10.1115/detc2022-88971","url":null,"abstract":"\u0000 This paper describes a methodology for automation of measurements in Computer-Aided Design (CAD) software by enabling the use of supervised learning algorithms. The paper presents a proof of concept of how dimensions are placed automatically in the drawing at predicted positions. The framework consists of two trained neural networks and a rule-based system. Four steps compound the methodology. 1. Create a data set of labeled images for training a pre-built convolutional neural network (YOLOv5) using CAD automatic procedures. 2. Train the model to make predictions on 2D drawing imagery, identifying their relevant features. 3. Reuse the information extracted from YOLOv5 in a new neural network to produce measurement data. The output of this model is a matrix containing measurement location and size data. 4. Convert the final data output into actual measurements of an unseen geometry using a rule-based system for automatic dimension generation. Although the rule-based system is highly dependent on the problem and the CAD software, both supervised learning models exhibit high performance and reusability. Future work aims to make the framework suitable for more complex products. The methodology presented is promising and shows potential for minimizing human resources in repetitive CAD work, particularly in the task of creating engineering drawings.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114222179","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}

引用次数: 1

The Impact of a Strategy of Deception About the Identity of an Artificial Intelligence Teammate on Human Designers 人工智能队友身份欺骗策略对人类设计师的影响

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-88535

Guanglu Zhang, A. Raina, Ethan Brownell, J. Cagan

{"title":"The Impact of a Strategy of Deception About the Identity of an Artificial Intelligence Teammate on Human Designers","authors":"Guanglu Zhang, A. Raina, Ethan Brownell, J. Cagan","doi":"10.1115/detc2022-88535","DOIUrl":"https://doi.org/10.1115/detc2022-88535","url":null,"abstract":"\u0000 Advances in artificial intelligence (AI) offer new opportunities for human-AI collaboration in engineering design. Human trust in AI is a crucial factor in ensuring an effective human-AI collaboration, and several approaches to enhance human trust in AI have been suggested in prior studies. However, it remains an open question in engineering design whether a strategy of deception about the identity of an AI teammate can effectively calibrate human trust in AI and improve human-AI joint performance. This research assesses the impact of the strategy of deception on human designers through a human subjects study where half of participants are told that they work with an AI teammate (i.e., without deception), and the other half of participants are told that they work with another human participant but in fact they work with an AI teammate (i.e., with deception). The results demonstrate that, for this study, the strategy of deception improves high proficiency human designers’ perceived competency of their teammate. However, the strategy of deception does not raise the average number of team collaborations and does not improve the average performance of high proficiency human designers. For low proficiency human designers, the strategy of deception does not change their perceived competency and helpfulness of their teammate, and further reduces the average number of team collaborations while hurting their average performance at the beginning of the study. The potential reasons behind these results are discussed with an argument against using the strategy of deception in engineering design.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115244915","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}

引用次数: 0

Am I Right? Investigating the Influence of Trait Empathy and Attitudes Towards Sustainability on the Accuracy of Concept Evaluations in Sustainable Design 我说的对吗?探讨特质共情和可持续性态度对可持续性设计概念评价准确性的影响

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-90029

M. Alzayed, Elizabeth Starkey, Sarah C. Ritter, Rohan Prabhu

{"title":"Am I Right? Investigating the Influence of Trait Empathy and Attitudes Towards Sustainability on the Accuracy of Concept Evaluations in Sustainable Design","authors":"M. Alzayed, Elizabeth Starkey, Sarah C. Ritter, Rohan Prabhu","doi":"10.1115/detc2022-90029","DOIUrl":"https://doi.org/10.1115/detc2022-90029","url":null,"abstract":"\u0000 Concept selection is an integral stage in the design process, in which designers evaluate and compare solutions to select a smaller subset of solutions for further development. In this stage, designers often rely on their self-evaluations of solutions — either independently or using design tools — to make design decisions. However, the reliance on self-evaluations among novice designers could lead to faulty decision-making, given the presence of numerous cognitive biases. Consequently, we aim to investigate the accuracy of novice designers’ self-evaluations of the sustainability of their solutions and the moderating role of (1) trait empathy and (2) their beliefs, attitudes, and intentions towards sustainability on this accuracy. Towards this aim, we conducted an experiment with first-year engineering students comprising a sustainable design lecture and a design activity. At the end of the design activity, participants were asked to evaluate the sustainability of their own solutions and these self-evaluations were compared against expert evaluations. From the results, we see that participants demonstrate some degree of accuracy in their self-evaluations, but only with the sustainable design heuristics of longevity, sharing for maximal use, and active repair of misuse. Second, we see that trait empathy moderated the accuracy of self-evaluations, with participants reporting lower levels of fantasy and empathic concern demonstrating more accurate self-evaluations. Finally, we see that beliefs, attitudes, and intentions towards sustainability also moderated the accuracy of their self-evaluations, and participants with lower levels demonstrated greater accuracy of self-evaluations. Taken together, these findings suggest that designers’ individual differences such as trait empathy could moderate the accuracy of the evaluation of their solutions, especially in the context of sustainability.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"318 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114698263","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}

引用次数: 2

Data Fusion as a Latent Space Learning Problem 作为潜在空间学习问题的数据融合

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-90233

Jonathan Tammer Eweis-Labolle, Nicholas Oune, R. Bostanabad

{"title":"Data Fusion as a Latent Space Learning Problem","authors":"Jonathan Tammer Eweis-Labolle, Nicholas Oune, R. Bostanabad","doi":"10.1115/detc2022-90233","DOIUrl":"https://doi.org/10.1115/detc2022-90233","url":null,"abstract":"Multi-fidelity modeling and calibration are two data fusion tasks that ubiquitously arise in engineering design. In this paper, we introduce a novel approach based on latent-map Gaussian processes (LMGPs) that enables efficient and accurate data fusion. In our approach, we convert data fusion into a latent space learning problem where the relations among different data sources are automatically learned purely based on the data. This conversion endows our approach with attractive advantages such as increased accuracy, reduced costs, and flexibility to jointly fuse any number of data sources. Additionally, the learned latent space in our approach compactly visualizes the correlations between data sources which allows designers and engineers to detect model form errors or determine the optimum strategy for high-fidelity emulation by only fusing correlated or sufficiently accurate data sources. We also develop a new correlation function that enables LMGPs to estimate calibration parameters with high accuracy and consistency even in the presence of unbalanced and noisy datasets. The implementation and use of our approach are considerably simpler and less prone to numerical issues compared to existing data fusion technologies. We demonstrate the benefits of LMGP-based data fusion on a wide range of analytic examples. by comparing its performance against existing technologies.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"12 6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133700110","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}

引用次数: 0

Digital Engineering Platform for Synergistic Decision-Making In Manufacturing Plant Operations: Research Questions 制造工厂运作中协同决策的数位工程平台:研究问题

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-91277

B. Gautham, Natarajan Swaminathan, R. Shukla, Trinath Gaduparthi, Chetan Malhotra

{"title":"Digital Engineering Platform for Synergistic Decision-Making In Manufacturing Plant Operations: Research Questions","authors":"B. Gautham, Natarajan Swaminathan, R. Shukla, Trinath Gaduparthi, Chetan Malhotra","doi":"10.1115/detc2022-91277","DOIUrl":"https://doi.org/10.1115/detc2022-91277","url":null,"abstract":"\u0000 Decision makers in the manufacturing industry ranging from plant operators to senior management make decisions based on a combination of defined procedures and rules, expert inputs and analysis, and their own knowledge and understanding of the problem context. Decision spaces are getting more complex, with the business paradigms shifting towards autonomous plants and servitization of products. With the advent of Internet of Things (IoT) and technologies such as machine learning and digital twins, the resources and capabilities available to decision-makers are expanding vastly. The gamut of concerns to be addressed is also expanding, with new challenges such as sustainability and its concomitant regulations and the pressure to make businesses more socially aware. Further, it would be ideal if decision-makers could easily draw upon the relevant knowledge, intuition, and experience of human experts, as well as knowledge currently buried in documents and data, and synthesize all the diverse inputs towards informed decision-making by integrating cyber, physical, and social systems. This motivates the question, “How do we create platforms that synergize these diverse knowledge sources and capabilities to facilitate better decision-making?” In this paper, we try to delve into identification of few key research questions and discuss opportunities and requirements around the same, that can aid in creating a digital platform to synergize all these diverse inputs and support decision-making. While this paper uses decision-making in manufacturing plant operations to explore the challenges and discuss one possible approach, the problem of enabling seamless synergy between the knowledge and capabilities of diverse human, IT and physical elements applies to all Cyber Physical Social Systems (CPSS).","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128722813","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}

引用次数: 0

A Scalable Graph Learning Approach to Capacitated Vehicle Routing Problem Using Capsule Networks and Attention Mechanism 基于胶囊网络和注意机制的可扩展图学习方法研究有能力车辆路径问题

Volume 3B: 48th Design Automation Conference (DAC) Pub Date : 2022-08-14 DOI: 10.1115/detc2022-90123

Steve Paul, Souma Chowdhury

{"title":"A Scalable Graph Learning Approach to Capacitated Vehicle Routing Problem Using Capsule Networks and Attention Mechanism","authors":"Steve Paul, Souma Chowdhury","doi":"10.1115/detc2022-90123","DOIUrl":"https://doi.org/10.1115/detc2022-90123","url":null,"abstract":"\u0000 This paper introduces a new graph neural network architecture for learning solutions of Capacitated Vehicle Routing Problems (CVRP) as policies over graphs. CVRP serves as an important benchmark for a wide range of combinatorial planning problems, which can be adapted to manufacturing, robotics and fleet planning applications. Here, the specific aim is to demonstrate the significant real-time executability and (beyond training) scalability advantages of the new graph learning approach over existing solution methods. While partly drawing motivation from recent graph learning methods that learn to solve CO problems such as multi-Traveling Salesman Problem (mTSP) and VRP, the proposed neural architecture presents a novel encoder-decoder architecture. Here the encoder is based on Capsule networks, which enables better representation of local and global information with permutation invariant node embeddings; and the decoder is based on the Multi-head attention (MHA) mechanism allowing sequential decisions. This architecture is trained using a policy gradient Reinforcement Learning process. The performance of our approach is favorably compared with state-of-the-art learning and non-learning methods for a benchmark suite of Capacitated-VRP (CVRP) problems. A further study on the CVRP with demand uncertainties is conducted to explore how this Capsule-Attention Mechanism architecture can be extended to handle real-world uncertainties by embedding them through the encoder.","PeriodicalId":394503,"journal":{"name":"Volume 3B: 48th Design Automation Conference (DAC)","volume":"128 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115044466","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}

引用次数: 4