检索结果-内蒙古大学图书馆

Leveraging trust for joint multi-objective and multi-fidelity optimization

MACHINE LEARNING-SCIENCE AND TECHNOLOGY 2024年第1期5卷 015056页

作者： Irshad, Faran Karsch, Stefan Doepp, Andreas Ludwig Maximilian Univ Munich Fac Phys Coulombwall 1 D-85748 Garching Germany

In the pursuit of efficient optimization of expensive-to-evaluate systems, this paper investigates a novel approach to Bayesian multi-objective and multi-fidelity (MOMF) optimization. Traditional optimization methods, while effective, often encounter prohibitively high costs in multi-dimensional optimizations of one or more objectives. multi-fidelity approaches offer potential remedies by utilizing multiple, less costly information sources, such as low-resolution approximations in numerical simulations. However, integrating these two strategies presents a significant challenge. We propose the innovative use of a trust metric to facilitate the joint optimization of multiple objectives and data sources. Our methodology introduces a modified multi-objective (MO) optimization policy incorporating the trust gain per evaluation cost as one of the objectives of a Pareto optimization problem. This modification enables simultaneous MOMF optimization, which proves effective in establishing the Pareto set and front at a fraction of the cost. Two specific methods of MOMF optimization are presented and compared: a holistic approach selecting both the input parameters and the fidelity parameter jointly, and a sequential approach for benchmarking. Through benchmarks on synthetic test functions, our novel approach is shown to yield significant cost reductions-up to an order of magnitude compared to pure MO optimization. Furthermore, we find that joint optimization of the trust and objective domains outperforms sequentially addressing them. We validate our findings with the specific use case of optimizing particle-in-cell simulations of laser-plasma acceleration, highlighting the practical potential of our method in the Pareto optimization of highly expensive black-box functions. Implementation of the methods in existing Bayesian optimization frameworks is straightforward, with immediate extensions e.g. to batch optimization possible. Given their ability to handle various continuous

关键词： Bayesian optimization multi-objective optimization computational physics

来源：评论

学校读者我要写书评

暂无评论

multi-Stage Battery Charging Scheduling optimization at Battery Swapping Stations 16

Multi-Stage Battery Charging Scheduling Optimization at Batt...

引用

16th IEEE PES Asia-Pacific Power and Energy Engineering Conference, APPEEC 2024

作者： Yv, Hao Han, Weiji China-UK Low Carbon College Shanghai Jiao Tong University Shanghai201306 China

ISBN: (纸本)9798350386127

Battery swapping has become a prominent method of energy supply for electric vehicles, featuring short service time and streamlined charging management. This paper explores efficient battery charging scheduling (BCS) strategies at battery swapping stations (BSSs). By analyzing the operational characteristics of batteries, a multi-stage BCS model is proposed. Unlike the conventional constant-power single-stage BCS model, the proposed multi-stage BCS model allows batteries to be charged to different expected state of energy (SoE) levels in different stages, and the charging power rate can be adaptively adjusted to avoid over-voltage or accelerated degradation when the battery approaches its highest SoE level. Considering both the battery SoEs and real-time swapping demands and taking the batteries to be charged in each stage as control variables, a multi-objective integer programming problem is formulated to minimize the average charging cost as well as to maximize the average SoE of charged batteries. This problem is solved using the Gurobi solver, and the effectiveness and advantages of the proposed BCS strategy are demonstrated through a case study of a practical BSS for heavy-duty electric trucks. © 2024 IEEE.

关键词： Charging stations

来源：评论

学校读者我要写书评

暂无评论

Incomplete risk-preference information in portfolio decision analysis

引用

EUROPEAN JOURNAL OF OPERATIONAL RESEARCH 2023年第3期304卷 1084-1098页

作者： Liesio, Juuso Kallio, Markku Argyris, Nikolaos Aalto Univ Dept Informat & Serv Management Sch Business Espoo Finland Loughborough Univ Sch Business & Econ Loughborough Leics England

Portfolio decision analysis models support decisions on the allocation of resources among assets with un-certain outcomes (e.g., investments, projects or stocks). These models require information on the decision maker's risk-preferences which can be difficult to obtain in practice. Stochastic dominance criteria show promise in this regard as they can compare portfolios without exact specification of risk-preferences, but the current literature lacks practical approaches for generating the efficient frontier, i.e., the set of those portfolios that are not stochastically dominated by any other portfolio. We address this gap by develop-ing models to identify sets of portfolios that are efficient in the sense of second-or third-order stochastic dominance (SSD, TSD). These models provide novel insights into the composition of portfolios belong-ing to the efficient frontier by, e.g., identifying those assets that are included in all efficient portfolios. Moreover, the identification of the efficient frontier makes it possible to utilize additional information on the decision maker's risk-preferences to further reduce the set of admissible portfolio alternatives, and to analyze the implications this information has on the amount of capital that should be allocated to each individual asset. We illustrate the usefulness of these models with applications in project portfolio selection and financial portfolio diversification. (c) 2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( http://***/licenses/by/4.0/ )

关键词： Decision support systems Portfolio decision analysis Stochastic dominance multi -objective programming optimization

来源：评论

学校读者我要写书评

暂无评论

Metamodeling in Evolutionary multi-objective optimization for Constrained and Unconstrained Problems

Metamodeling in Evolutionary Multi-Objective Optimization fo...

引用

作者： Hussein, Rayan Michigan State University

学位级别：Ph.D., Doctor of Philosophy

One of the main difficulties in applying an optimization algorithm to a practical problem is that the evaluation of objectives and constraints often involve computationally expensive procedures. To handle such problems, a metamodel (or surrogate model, or response surface approximations) is first formed from a few exact (high-fidelity) solution evaluations, and then optimized by an algorithm in a progressive manner. However, there has been lukewarm interest in finding multiple trade-off solutions for multi-objective optimization problems using surrogate models. The literature on surrogate modeling for constrained optimization problems is also rare. The difficulty lies in the requirement of building and solving multiple surrogate models, one for each Pareto-optimal solution. In this study, we propose a taxonomy of different possible metamodeling frameworks for multi-objective optimization and provide a comparative study by discussing advantages and disadvantages of each framework. Also, we argue that it is more efficient to use different metamodeling frameworks at different stages of the optimization process. Thereafter, we propose a novel adaptive method for switching among different metamodeling frameworks. Moreover, we observe the convergence behavior of the proposed approaches is better with a trust regions method applied within the metamodeling frameworks. The results presented in this study are obtained using the well-known Kriging metamodeling approach. Based on our extensive simulation studies on proposed frameworks, we report new and interesting observations about the behavior of each metamodeling framework, which may provide salient guidelines for further studies in this emerging area within evolutionary multi-objective optimization. Results of this study clearly show the efficacy and efficiency of the proposed adaptive switching approach compared to three recently-proposed other metamodeling algorithms on challenging multi-objective optimization problems u

关键词： Metamodeling optimization algorithm Surrogate models

来源：评论

学校读者我要写书评

暂无评论

Enhancing Robotics Online 3D Bin Packing: A Comparative Study of Conventional Heuristic and Deep Reinforcement Learning approaches 20

Enhancing Robotics Online 3D Bin Packing: A Comparative Stud...

引用

IEEE 20th International Conference on Automation Science and Engineering (CASE)

作者： Xiong, Heng Ding, Kai Ding, Wan Qiu, Xuchong Janschek, Klaus Xu, Jianfeng Huazhong Univ Sci & Technol Sch Mech Sci & Engn State Key Lab Intelligent Mfg Equipment & Technol Wuhan 430074 Peoples R China BOSCH Corp Res Beijing Peoples R China Tech Univ Dresden Dresden Germany

ISBN: (纸本)9798350358513;9798350358520

Robotic object packing holds a wide array of practical applications across logistics and manufacturing sectors. The online 3D Bin Packing Problem (BPP) is a challenging task that involves online packing of three-dimensional boxes into a container while considering constraints and objectives. Unlike the offline version, where all boxes are known in advance, the online variant requires decisions about how to pack each box as it arrives, without prior knowledge of upcoming boxes. To maximize space utilization, our study explores two distinct strategies: conventional heuristics-based algorithms and a deep reinforcement learning (DRL)-based approach. For the heuristic strategy, we propose four heuristic criteria alongside two variants of multi-objective optimization (MOO) algorithms. Quantitative experiments reveal that MOO outperforms singleobjective approaches for online bin packing. In our DRL-based approach, we introduce a framework that leverages a candidate map that indicates the potentially feasible placements, ensuring a balanced exploration and exploitation in the considerable discrete action space. Experiments demonstrate the superior performance of our DRL-based approach compared to both DRL-based baseline methods and conventional approaches. Additionally, we discuss the limitations of DRL-based methods and offer practical recommendations for real-world applications.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

multi-objective topology optimization of pin-fin heat exchangers using spectral and finite-element methods

引用

STRUCTURAL AND multiDISCIPLINARY optimization 2021年第4期64卷 2075-2095页

作者： Ghasemi, Ali Elham, Ali Tech Univ Carolo Wilhelmina Braunschweig Inst Aircraft Design & Lightweight Struct Braunschweig Germany

Forced convective pin-fin heat exchangers, due to the high wet surface area per volume and the hindered thermal boundary layers, feature low thermal resistances. However, the considerable coolant pressure drop, particularly for densely packed fin arrays, imposes operational costs for pumping power supply. This paper develops a multi-objective topology optimization approach to optimize sink geometries in order to minimize thermal resistance and pressure loss, concurrently. In accordance to the pin-fin geometrical characteristics, a dedicated pseudo-3D conjugate heat transfer model is utilized, by assuming periodic flow and fin design pattern, to reasonably reduce the high cost of full-3D model optimization. For the solution of flow part, a pseudo-spectral scheme is used, which is intrinsically periodic and features a high spectral accuracy, and the finite element method for the non-periodic conjugate heat transfer model. Exact partial derivatives of the discrete solutions are obtained analytically by hand-differentiation. This task is rather convenient for the flow part, due to the simplicity of the pseudo-spectral implementation;however, the MATLAB symbolic toolbox is selectively utilized for the finite element code to ensure an error-free implementation. Finally, the sensitivities are computed directly or via a discrete adjoint method, for the flow and heat models, respectively. To examine the present approach, two approaches are used for optimization of a practical cooling task: constrained and unconstrained multi-objective formulations, where in all cases more emphasis is placed on thermal resistance minimization. A series of optimized heat sink geometries via constrained or unconstrained multi-objective optimizations are obtained to examine practical utility of the present approach. The optimized topologies demonstrated superior cooling performances at lower costs of pressure losses compared to conventional (circular) in-line and staggered fins, and confirmed th

关键词： Topology optimization Pin-fin sink multi-objective Spectral method Heat-exchanger optimization

来源：评论

学校读者我要写书评

暂无评论

Dynamic User Association in Scalable Ultra-Dense LEO Satellite Networks

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2022年第8期71卷 8891-8905页

作者： Jiang, Hao Wang, Hai Hu, Yulin Wu, Jing Wuhan Univ Wuhan 430072 Peoples R China

Ultra-dense low earth orbit (LEO) satellite networks (LEO-UDSNs) are becoming a foundational component of future 5G and beyond. LEO-UDSNs bring unprecedented high-speed mobility, broadband capacity, and ultra-density qualities, introducing a significant challenge to traditional user association (UA) methods. In this article, we first discuss the difficulties in sequential UA optimization for scalable LEO-UDSNs. A specific multi-objective UA problem was formulated for the overall service efficiency. We initially proposed a practical Iterative Subgradient Algorithm (MOISA) decoupling potential trade-offs between limited equivalent power and availability relationship, rapidly obtaining solutions in stable cases. Second, we proposed a multi-objective Dynamic User Association Approach (MODUAA), inheriting the MOISA with stability enhancements for sequential optimizations, ensuring the performance requirements, and alleviating the iteration redundancy in dynamic scenarios. Finally, the outperforming of the MOISA and the applicability of the MODUAA were elaborated in various comparisons.

关键词： Satellites Vehicle dynamics Low earth orbit satellites optimization Resource management Broadband communication Broadband antennas Load balancing resource allocation ultra-dense LEO satellite networks user association

来源：评论

学校读者我要写书评

暂无评论

Parametric optimization of friction stir welding process parameters of dissimilar welded joints using grey relational analysis and desirability function approach

引用

WORLD JOURNAL OF ENGINEERING 2023年第3期20卷 609-617页

作者： Nadikudi, Bhanodaya Kiran Babu Sreenidhi Inst Sci & Technol Hyderabad India

Purpose The main purpose of the present work is to study the multi response optimization of dissimilar friction stir welding (FSW) process parameters using Taguchi-based grey relational analysis and desirability function approach (DFA). Design/methodology/approach The welded sheets were fabricated as per Taguchi orthogonal array design. The effects of tool rotational speed, transverse speed and tool tilt angle process parameters on ultimate tensile strength and hardness were analyzed using grey relational analysis, and DFA and optimum parameters combination was determined. Findings The tensile strength and hardness values were evaluated from the welded joints. The optimum values of process parameters were estimated through grey relational analysis and DFA methods. Similar kind of optimum levels of process parameters were obtained through two optimization approaches as tool rotational speed of 1150 rpm, transverse speed of 24 mm/min and tool tilt angle of 2 degrees are the best process parameters combination for maximizing both the tensile strength and hardness. Through these studies, it was confirmed that grey relational analysis and DFA methods can be used to find the multi response optimum values of FSW process parameters. Research limitations/implications In the present study, the FSW is performed with L9 orthogonal array design with three process parameters such as tool rotational speed, transverse speed and tilt angle and three levels. practical implications Aluminium alloys are widely using in automotive and aerospace industries due to holding a high strength to weight property. Originality/value Very limited work had been carried out on multi objective optimization techniques such as grey relational analysis and DFA on friction stir welded joints made with dissimilar aluminium alloys sheets.

关键词： Aluminium alloys Process parameters Mechanical properties Grey relational analysis Desirability function analysis

来源：评论

学校读者我要写书评

暂无评论

Exploring trade-offs in agro-ecological landscapes: Using a multi-objective land-use allocation model to support agroforestry research

引用

BASIC AND APPLIED ECOLOGY 2022年 64卷 103-119页

作者： Reith, Esther Gosling, Elizabeth Knoke, Thomas Paul, Carola Tech Univ Munich Inst Forest Management TUM Sch Life Sci Weihenstephan Hans Carl Von Carlowitz Pl 2 D-85354 Freising Weihenstephan Germany Univ Gottingen Dept Forest Econ & Sustainable Land Use Planning Busgenweg 1 D-37077 Gottingen Germany Ctr Biodivers & Sustainable Land Use Busgenweg 1 D-37077 Gottingen Germany

Finding the optimal land allocation for providing ecosystem services, conserving biodiversity and maintaining rural liveli-hoods is a key challenge of agricultural management and land-use planning. Agroforestry has been widely discussed as a sus-tainable land-use solution and as one strategy to improve the provision of multiple ecological and economic functions in agricultural landscapes. In this study, we use the backdrop of agroforestry research to evaluate a method from the multi-criteria decision analysis toolbox: robust multi-objective optimization. The key feature of this modelling approach is its capacity to integrate uncertain ecological and socio-economic data. We illustrate the optimization model with a case study from eastern Panama, showing how the model can bring together scientific and practical knowledge to provide potentially desirable land-scape compositions from the perspective of farmers, a public perspective, and a compromise solution. Example results of our case study show how to assess whether agroforestry is a desirable component in a landscape composition to satisfy multiple objectives of different interest groups. Furthermore, we use the model to demonstrate how different objectives influence the optimal area share and type of agroforestry. Due to its parsimonious nature, the model could be used as a starting point of an interactive co-learning process with decision-makers, researchers and other stakeholders. The model, however, is not yet suit-able for an exact prediction of future land-use dynamics, for questions of spatially explicit land-use configuration, studies going beyond the regional scale or for socio-economic interactions of agents. Therefore, we outline future research needs and recom-mendations for other types of models or hybrid approaches.(c) 2022 The Author(s). Published by Elsevier GmbH on behalf of Gesellschaft fur Okologie. This is an open access article under the CC BY license (http://***/licenses/by/4.0/

关键词： Agroforestry Decision-support tools Ecosystem services Integrated landscape management multi-objective optimization Sim-ulation models

来源：评论

学校读者我要写书评

暂无评论

optimization of end-of-life product recycling supply chain using PSO algorithm and R language 27

Optimization of end-of-life product recycling supply chain u...

引用

27th World multi-Conference on Systemics, Cybernetics and Informatics, WMSCI 2023

作者： Ellili, Walid Marheel, Mahfoud Louati, Mahdi Loukil, Taïcir Moalla Computer Science Department University of Sfax Sfax3018 Tunisia

ISBN: (纸本)9781713880325

This article presents an optimization approach for the end-of-life product recycling supply chain using the Particle Swarm optimization (PSO) algorithm and the R programming language. The objective is to minimize the total cost associated with transportation, facility opening, and variable costs, while meeting the capacity constraints of collection, recycling, and landfill sites over multiple times. The PSO algorithm is utilized to find the optimal solution for the problem. The results demonstrate the effectiveness of the proposed approach in optimizing the end-of-life product recycling supply chain. The optimal solution yields a significant cost reduction compared to previous approaches. The article concludes with insights into the benefits and limitations of using the PSO algorithm and R language, and provides recommendations for future research in this area. Overall, this study contributes to the field of sustainable supply chain management by providing a practical and efficient solution for optimizing the end-of-life product recycling process. © 2023 Proceedings of World multi-Conference on Systemics, Cybernetics and Informatics, WMSCI. All rights reserved.

关键词： Optimal systems

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：