Computers & Operations Research最新文献

{"title":"Learning-based column generation approach for the vehicle routing problem with release dates and incompatible loading constraints","authors":"Mujin Gao ,&nbsp;Yanru Chen ,&nbsp;Zongcheng Zhang ,&nbsp;M.I.M. Wahab","doi":"10.1016/j.cor.2025.107152","DOIUrl":"10.1016/j.cor.2025.107152","url":null,"abstract":"<div><div>This study introduces a variant of classical distribution problems, vehicle routing problems with release dates, and incompatible loading constraints (VRPR-ILC). The VRPR-ILC is derived from the practical application of a pharmaceutical distribution company based in China. It incorporates various real-world constraints, including release dates, product weights and volumes, incompatible loading, and time windows. The objective of the VRPR-ILC is to minimize the total travel distance. This variant can also find applications in diverse domains, such as e-commerce. Integrating the above constraints introduces the challenge of optimizing in both the temporal and spatial dimensions. To tackle this issue, we propose a learning-based column generation (LCG) approach. The LGG provides a new framework combining the deep learning (DL) technique with the column generation (CG) algorithm. By utilizing DL, the LCG effectively guides the CG in concentrating on the search space containing high-quality integer solutions. It helps to narrow the gap between linear and integer solutions and significantly enhances the convergence of the algorithm. Additionally, to address the challenges posed by the pricing problem of the VRPR-ILC, we develop the heuristic pricing, the dummy label dominance rule, and a lower bound evaluation strategy for labels. Computational results show that the LCG achieves competitive results compared with the GUROBI solver, the existing exact algorithm, and the heuristic algorithm. The results further indicate that utilizing the DL leads to improved solutions while reducing the time by 20%.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"183 ","pages":"Article 107152"},"PeriodicalIF":4.1,"publicationDate":"2025-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A bi-cooperative parallel evolutionary algorithm for co-scheduling of distributed production and distribution considering shared transportation resources","authors":"Zijiang Liu ,&nbsp;Hongyan Sang ,&nbsp;Quanke Pan ,&nbsp;Ling Wang","doi":"10.1016/j.cor.2025.107157","DOIUrl":"10.1016/j.cor.2025.107157","url":null,"abstract":"<div><div>This paper investigates the distributed permutation flowshop co-scheduling problem with sequence-dependent setup times (DPFCSP-SDST). A mixed-integer linear programming model is established to minimize the number of AGVs, the total AGV travel distance, and the makespan. The model is validated using the Gurobi solver. The DPFCSP-SDST contains several coupled sub-problems, including transport task assignment, transport task sequencing, processing task assignment, and processing task sequencing. We make the first attempt to propose a bi-cooperative parallel evolutionary algorithm (BCPEA) for solving such a strong NP-hard problem. The proposed BCPEA integrates a hybrid heuristic initialization and two cooperative strategies: knowledge-and-learning cooperative neighborhood search and parallel cooperative co-evolution. The hybrid heuristic initialization incorporates two initialization strategies that compete to generate a superior initial solution. The knowledge-and-learning cooperative neighborhood search employs a learning mechanism to select from 11 pre-designed neighborhood structures using empirical knowledge, with accelerated evaluation methods designed to speed up the insertion and swapping operations. In the parallel cooperative co-evolutionary algorithm, two multi-thread groups are assigned to be responsible for global and local search, respectively, and both use the iterated greedy algorithm as the optimization engine. Cooperation is achieved through elite information transfer between the multi-thread groups and useful information collection by the host process. Computational experiments demonstrate that the proposed BCPEA exhibits superior effectiveness and efficiency in solving the DPFCSP-SDST, outperforming existing methods.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"183 ","pages":"Article 107157"},"PeriodicalIF":4.1,"publicationDate":"2025-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144239796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A greedy hypervolume polychotomic scheme for multiobjective combinatorial optimization","authors":"Gonçalo Lopes ,&nbsp;Kathrin Klamroth ,&nbsp;Luís Paquete","doi":"10.1016/j.cor.2025.107140","DOIUrl":"10.1016/j.cor.2025.107140","url":null,"abstract":"<div><div>The usual goal in multiobjective combinatorial optimization is to find the complete set of nondominated points. However, in general, the nondominated set may be too large to be enumerated under a tight time budget. In these cases, it is preferable to rapidly obtain a concise representation of the nondominated set that satisfies a given property of interest. This work describes a generic greedy approach to compute a representation of the nondominated set for multiobjective combinatorial optimization problems that <em>approximately</em> maximizes the dominated hypervolume. The representation is built iteratively by solving a sequence of hypervolume scalarized problems, each of which with respect to <span><math><mi>k</mi></math></span> reference points, which is a parameter of our approach. We present a mixed-integer formulation of the hypervolume scalarization function for <span><math><mi>k</mi></math></span> reference points as well as a combinatorial branch-and-bound for the <span><math><mi>m</mi></math></span>-objective knapsack problem. We empirically analyse the functional relationship between <span><math><mi>k</mi></math></span> and its running-time and representation quality. Our results indicate that the branch-and-bound is a much more efficient approach and that increasing <span><math><mi>k</mi></math></span> does not directly translate into much better representation quality.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"183 ","pages":"Article 107140"},"PeriodicalIF":4.1,"publicationDate":"2025-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196122","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Online algorithms with foresight for radiotherapy patient scheduling","authors":"Roberto Aringhieri ,&nbsp;Davide Duma ,&nbsp;Giuseppe Squillace","doi":"10.1016/j.cor.2025.107132","DOIUrl":"10.1016/j.cor.2025.107132","url":null,"abstract":"<div><div>A radiotherapy treatment consists of a given number of radiation sessions, which should start before a specified due date and have a duration that varies based on the patient category. Waiting time is the main critical issue in the management of a radiotherapy health system: the time elapsed between the first consultation and the first treatment is typically rather long and delays have the potential to damage the health status of the patients. Then, an efficient use of available linacs is crucial to ensure the success of the treatment. In this paper, we deal with a Radiotherapy Patient Scheduling (RPS) problem at the pure-online level under the blocking policy, which divides work shifts into time slots of equal duration. We propose online algorithms that enable the scheduling of a sequence of appointments for each patient, all at the same time slot each day, whenever a patient needs to commence a series of radiotherapy sessions. Considering a realistic and patient-centred operational context, the problem becomes highly challenging, even in its offline setting. We address this problem by introducing the concept of online optimisation with foresight, which is the common framework of the proposed approaches. The rationale behind foresight is to take real-time decisions under uncertainty by exploiting the partial knowledge of the optimal solution structure. A quantitative analysis shows that the proposed algorithms outperform two competitor algorithms inspired by the literature. Furthermore, the exploitation of a pattern observed in the offline solutions of the problem (implicit foresight) results to be more flexible and effective than using a solution structure given by integer linear programs on the most likely scenario (explicit foresight).</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107132"},"PeriodicalIF":4.1,"publicationDate":"2025-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144169292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Flexible Job Shop Scheduling Problem using graph neural networks and reinforcement learning","authors":"Xi Liu ,&nbsp;Xin Chen ,&nbsp;Vincent Chau ,&nbsp;Jedrzej Musial ,&nbsp;Jacek Blazewicz","doi":"10.1016/j.cor.2025.107139","DOIUrl":"10.1016/j.cor.2025.107139","url":null,"abstract":"<div><div>The Flexible Job Shop Scheduling Problem (FJSP) is an important research topic in the field of manufacturing. Many studies have used Deep Reinforcement Learning (DRL) to learn Priority Dispatching Rules (PDR) to address the FJSP. However, compared to exact methods, there is still significant room for improvement in the quality of solutions. This paper proposes a new end-to-end DRL framework that utilizes Graph Attention Networks (GAN) to extract relevant information from the disjunctive graph. In this framework, we introduce adaptive weights when calculating attention scores, allowing the model to dynamically adjust the attention scores based on the characteristics of the input data. This helps the model more effectively capture key information within the data. To alleviate the network degradation issue and enhance model performance, the features extracted by the aforementioned model are input into a Residual Connection (RC) module for further deep feature extraction. Finally, our model is validated on generated datasets and public benchmarks, with experimental results indicating that the proposed method outperforms traditional PDR methods and the latest DRL approaches.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107139"},"PeriodicalIF":4.1,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144134892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhanced open-source scatter search algorithm for solving quadratic unconstrained binary optimization problems","authors":"Donghao Liu ,&nbsp;Wei Yang ,&nbsp;Hongpu Wang ,&nbsp;Yu Du ,&nbsp;Yang Wang ,&nbsp;Zhipeng Lü ,&nbsp;Jin-Kao Hao","doi":"10.1016/j.cor.2025.107137","DOIUrl":"10.1016/j.cor.2025.107137","url":null,"abstract":"<div><div>In recent years, quantum computing has driven significant excitement and innovation, with the Quadratic Unconstrained Binary Optimization (QUBO) model at its core. This paper introduces SATPR, a new open-source quantum-inspired metaheuristic algorithm that combines scatter search, adaptive tenure tabu search, and path-relinking. The adaptive nature of the tabu tenure, achieved through the integration of various heuristic components, enables SATPR to effectively solve different types of QUBO problem instances. Additionally, SATPR utilizes parallelism to fully leverage multi-threading capabilities, further enhancing its computational efficiency. We conducted extensive evaluations on large and challenging problem instances from four benchmark sets, including well-known QUBO and Max-Cut instances, as well as less explored random graph structures. Our results demonstrate that SATPR is highly competitive in both solution quality and computational efficiency when compared with leading metaheuristic QUBO solvers and the quantum-inspired Fixstars Amplify Annealing Engine.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107137"},"PeriodicalIF":4.1,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144139301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"New lower bounds on the cutwidth of graphs","authors":"Jean-Claude Bermond,&nbsp;Michel Cosnard,&nbsp;David Coudert,&nbsp;Frédéric Havet","doi":"10.1016/j.cor.2025.107130","DOIUrl":"10.1016/j.cor.2025.107130","url":null,"abstract":"<div><div>Cutwidth is a parameter used in many layout problems. Determining the cutwidth of a graph is an NP-complete problem, but it is possible to design efficient branch-and-bound algorithms if good lower bounds are available for cutting branches during exploration. Knowing how to quickly evaluate good bounds in each node of the search tree is therefore crucial.</div><div>In this article, we give new lower bounds based on different graph density parameters such as the minimum, the average and the maximum average degree. Our main result is a new bound using the notion of traffic grooming on a path network, which appear to be in many cases better than bounds in the literature. Furthermore, the bound based on grooming can be computed quickly, in <span><math><mrow><mi>O</mi><mrow><mo>(</mo><mo>log</mo><mi>n</mi><mo>)</mo></mrow></mrow></math></span> time, and so is of interest to design faster branch-and-bound algorithms. Through extensive experiments, we show that this bound behaves very well compared to other bounds. Furthermore, we show how to obtain even better results when combining it with heuristics for finding dense subgraphs.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107130"},"PeriodicalIF":4.1,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144131074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Two-stage robust optimization for public health emergency project scheduling with uncertain activity durations","authors":"Yuchen Dong ,&nbsp;Weibo Zheng ,&nbsp;Zhiqiang Ma ,&nbsp;Zhengwen He","doi":"10.1016/j.cor.2025.107135","DOIUrl":"10.1016/j.cor.2025.107135","url":null,"abstract":"<div><div>The utilization of emergency projects, which can facilitate the coordinated scheduling of medical resources and rescue activities, offers a promising approach for absorbing disturbances, mitigating damage and achieving restoration from public healthcare catastrophes. However, the complex implementation environment in public health emergencies significantly impacts the preventive effectiveness and rescue efficiency of emergency projects. Considering the substantial uncertainty in activity durations, this study models resource allocation and schedule generation of an emergency project as a two-stage robust optimization formulation to reduce economic expenditure and minimize delayed rescue. Specifically, the preparedness stage is to minimize the pre-deployment cost of emergency resources, and a max–min objective is introduced in the response stage to optimize the deprivation cost affected by random activity durations. Then, as the robust optimization model could be reformulated in a master-submodel framework, we develop a customized column-and-constraint generation (C&amp;CG) algorithm with enhancements based on the structure of the problem and decision variables for rapid problem-solving. Besides, the algorithms are tested on randomly generated datasets, and the influences of key parameters on the algorithm performance, the resource cost and the deprivation cost of the emergency project are analyzed. Based on the computational results, the customized C&amp;CG algorithm with enhancements outperforms others and sensitivity analysis of key parameters is presented. This research provides effective decision support for public health emergency scheduling and draws management insights, validated through a real-world case study that demonstrates the model’s practical effectiveness in enhancing emergency response resilience and efficiency.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107135"},"PeriodicalIF":4.1,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144146881","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Neural combinatorial optimization: A tutorial","authors":"Davide Angioni ,&nbsp;Claudia Archetti ,&nbsp;M. Grazia Speranza","doi":"10.1016/j.cor.2025.107102","DOIUrl":"10.1016/j.cor.2025.107102","url":null,"abstract":"<div><div>Recent advancements in deep reinforcement learning have sparked a growing interest in the application of this approach to solve combinatorial optimization (CO) problems. This paper presents neural combinatorial optimization (NCO) as a framework for constructing functions that work as heuristics for CO problems. Given the rapid expansion of the field and the increasing interest in the topic, this tutorial introduces the main techniques utilized in NCO and explores the current open issues in the field.</div><div>We define key terms and concepts related to NCO and present the latest developments, using the Knapsack Problem as a running example to complement theoretical explanations. Finally, we analyze prominent works in the field of NCO, with a focus on their application to the Traveling Salesman Problem, which serves as the most extensively studied problem in this domain.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107102"},"PeriodicalIF":4.1,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144139302","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Coordinated scheduling optimization of quay cranes and AGVs in automated container terminals","authors":"Yuzhen Hu ,&nbsp;Min Wang ,&nbsp;Rui Min ,&nbsp;Jianxia Liu ,&nbsp;Valery F. Lukinykh ,&nbsp;Shulin Tang ,&nbsp;Da Zhao","doi":"10.1016/j.cor.2025.107147","DOIUrl":"10.1016/j.cor.2025.107147","url":null,"abstract":"<div><div>In recent years, smart ports have attracted widespread attention because of their great potential to improve port operational efficiency. Automated container terminals (ACTs), which are equipped with quay cranes (QCs) and automated guided vehicles (AGVs), play important roles in smart ports. Their coordination directly affects berthing time and port congestion, thereby influencing overall working efficiency. In this paper, we present a novel mixed-integer programming model for the coordinated scheduling of QCs and AGVs. The model aims to minimize the makespan while considering factors such as the safe working intervals of QCs, the flexible working strategy of QCs, and the relationships between adjacent tasks. In contrast to scheduling QCs and AGVs sequentially, we integrate the operations as a whole and design a two-stage solution framework for the integration. The initial solution is generated through a newly built integer programming model and dichotomy in the first stage. Variable neighborhood search (VNS) and adaptive large neighborhood search (ALNS) are integrated to improve the solution in the second stage. Optimality analysis and numerical results show the followings. (i) The coordinated scheduling schemes obtained by our method outperform the sequential scheduling schemes, reducing the makespan by an average of 15.84 time units. (ii) Considering the safe working intervals and flexible working strategies of QCs can lead to better scheduling schemes than realistic scenarios can achieve. (iii) The port’s working efficiency is maximized, and the equipment idle time is minimized when the number of AGVs is three times that of QCs in our cases. (iv) Changes in the number of AGVs have a greater effect on the makespan of coordinated operation than do changes in the quantity of QCs. Consequently, our study aims to provide a quantitative reference for the development of ACTs, which could contribute to operational efficiency improvement and sustainable development in the smart port industry.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"182 ","pages":"Article 107147"},"PeriodicalIF":4.1,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144146944","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}