Pub Date : 2025-10-29DOI: 10.1016/j.cor.2025.107322
Zhaofang Mao , Jiaxin Zhang , Yiting Sun , Dian Huang , Yida Xu
Multi-manned assembly lines play a vital role in modern manufacturing by enabling efficient management of complex assembly tasks and enhancing productivity, particularly for large-scale products. This paper investigates the integration of human–robot collaboration into multi-manned assembly lines, addressing a collaboration mode where workers and robots can flexibly choose to perform tasks in parallel or collaboratively execute the same task. This integration aims to enhance the efficiency of assembly resource utilization. We propose a mixed-integer programming model for this problem, with the primary objective of minimizing the number of workers and the secondary objective of minimizing the number of stations. Additionally, we establish a lower bound for the problem. To effectively solve medium and large-scale instances, we develop a matheuristic algorithm (MA) based on the mathematical model and heuristics, and we propose several variants. The experimental results demonstrate that our method effectively optimizes the application of collaborative robots in multi-manned assembly lines, outperforming the classical method presented in the literature.
{"title":"A matheuristic approach for the multi-manned assembly line balancing problem with collaborative robots","authors":"Zhaofang Mao , Jiaxin Zhang , Yiting Sun , Dian Huang , Yida Xu","doi":"10.1016/j.cor.2025.107322","DOIUrl":"10.1016/j.cor.2025.107322","url":null,"abstract":"<div><div>Multi-manned assembly lines play a vital role in modern manufacturing by enabling efficient management of complex assembly tasks and enhancing productivity, particularly for large-scale products. This paper investigates the integration of human–robot collaboration into multi-manned assembly lines, addressing a collaboration mode where workers and robots can flexibly choose to perform tasks in parallel or collaboratively execute the same task. This integration aims to enhance the efficiency of assembly resource utilization. We propose a mixed-integer programming model for this problem, with the primary objective of minimizing the number of workers and the secondary objective of minimizing the number of stations. Additionally, we establish a lower bound for the problem. To effectively solve medium and large-scale instances, we develop a matheuristic algorithm (MA) based on the mathematical model and heuristics, and we propose several variants. The experimental results demonstrate that our method effectively optimizes the application of collaborative robots in multi-manned assembly lines, outperforming the classical method presented in the literature.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107322"},"PeriodicalIF":4.3,"publicationDate":"2025-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145413841","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}
Pub Date : 2025-10-27DOI: 10.1016/j.cor.2025.107321
Soumen Atta
In this article, one of the most commonly encountered problems in manufacturing systems, known as the Tool Indexing Problem (TIP), is considered. TIP involves allocating cutting tools to different slots in a tool magazine of a Computer Numerically Controlled (CNC) machine to reduce the processing time of jobs on the machine. In this article, three mixed-integer linear programming formulations of single-objective TIP without tool duplication and lifespan are presented. A comparative study of these three linear formulations of TIP is also included in this article. During the study, it was found that the exact solver CPLEX with these linear formulations struggles to find optimal solutions in a reasonable time for larger instances. Therefore, a Hybrid Large Neighborhood Search with Local Search (HLNS-LS) algorithm, which is a metaheuristic approach, is proposed for solving TIP. The LNS phase iteratively destroys and repairs solutions to explore different regions of the solution space, while the LS phase intensifies the search by applying multiple neighborhood operators, such as swap, insert, shift, and customized 2-Opt and 3-Opt, to refine solutions further. This hybrid approach balances diversification and intensification, aiming to find high-quality solutions that minimize the total turret rotation cost associated with tool indexing. The performance of the proposed HLNS-LS algorithm is evaluated against an improved Harmony Search algorithm, a Weighted Superposition Attraction-based algorithm, and a Constraint Programming model across 85 small-, medium-, and large-sized benchmark instances from the existing TIP literature.
{"title":"Linear formulations and a hybrid large neighborhood search algorithm for the tool indexing problem","authors":"Soumen Atta","doi":"10.1016/j.cor.2025.107321","DOIUrl":"10.1016/j.cor.2025.107321","url":null,"abstract":"<div><div>In this article, one of the most commonly encountered problems in manufacturing systems, known as the Tool Indexing Problem (TIP), is considered. TIP involves allocating cutting tools to different slots in a tool magazine of a Computer Numerically Controlled (CNC) machine to reduce the processing time of jobs on the machine. In this article, three mixed-integer linear programming formulations of single-objective TIP without tool duplication and lifespan are presented. A comparative study of these three linear formulations of TIP is also included in this article. During the study, it was found that the exact solver CPLEX with these linear formulations struggles to find optimal solutions in a reasonable time for larger instances. Therefore, a Hybrid Large Neighborhood Search with Local Search (HLNS-LS) algorithm, which is a metaheuristic approach, is proposed for solving TIP. The LNS phase iteratively destroys and repairs solutions to explore different regions of the solution space, while the LS phase intensifies the search by applying multiple neighborhood operators, such as swap, insert, shift, and customized 2-Opt and 3-Opt, to refine solutions further. This hybrid approach balances diversification and intensification, aiming to find high-quality solutions that minimize the total turret rotation cost associated with tool indexing. The performance of the proposed HLNS-LS algorithm is evaluated against an improved Harmony Search algorithm, a Weighted Superposition Attraction-based algorithm, and a Constraint Programming model across 85 small-, medium-, and large-sized benchmark instances from the existing TIP literature.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107321"},"PeriodicalIF":4.3,"publicationDate":"2025-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145463568","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}
Pub Date : 2025-10-27DOI: 10.1016/j.cor.2025.107320
Zhengyang Li , Zhandong Xu , Guoyuan Li , Anthony Chen
The multiclass bi-criteria traffic assignment (MBTA) problem considers travelers’ bi-criteria (time and toll) route choice behaviors and user heterogeneity. The MBTA models can be classified as discrete or continuous based on whether the value of time (VOT) is modeled as a discrete or continuous variable. While both models have been suggested in the literature, their connections and differences remain underexplored. This study compares the discrete MBTA (DMBTA) and continuous MBTA (CMBTA) models and proposes a unified path-based gradient projection (GP) algorithm framework to solve both models. In the unified framework, three modules, including column generation, decomposition and equilibration, and convergence criteria, are customized for discrete and continuous models, respectively. With appropriate algorithmic designs, both problems can be solved effectively by the path-based GP algorithm. Extensive numerical experiments show that the equilibrium flow of the DMBTA model will fluctuate when the number of classes is small, and it will converge to the equilibrium flow of the CMBTA model with the increase in the number of classes. Additionally, in all test networks, the CMBTA model requires CPU time comparable to the DMBTA model with five classes and maintains a working path set smaller than that of the DMBTA model with three classes, demonstrating that the CMBTA model can achieve a refined solution with modest computational time and memory requirements.
{"title":"A unified gradient projection algorithm for solving both discrete and continuous multiclass bi-criteria traffic assignment problems","authors":"Zhengyang Li , Zhandong Xu , Guoyuan Li , Anthony Chen","doi":"10.1016/j.cor.2025.107320","DOIUrl":"10.1016/j.cor.2025.107320","url":null,"abstract":"<div><div>The multiclass bi-criteria traffic assignment (MBTA) problem considers travelers’ bi-criteria (time and toll) route choice behaviors and user heterogeneity. The MBTA models can be classified as discrete or continuous based on whether the value of time (VOT) is modeled as a discrete or continuous variable. While both models have been suggested in the literature, their connections and differences remain underexplored. This study compares the discrete MBTA (DMBTA) and continuous MBTA (CMBTA) models and proposes a unified path-based gradient projection (GP) algorithm framework to solve both models. In the unified framework, three modules, including column generation, decomposition and equilibration, and convergence criteria, are customized for discrete and continuous models, respectively. With appropriate algorithmic designs, both problems can be solved effectively by the path-based GP algorithm. Extensive numerical experiments show that the equilibrium flow of the DMBTA model will fluctuate when the number of classes is small, and it will converge to the equilibrium flow of the CMBTA model with the increase in the number of classes. Additionally, in all test networks, the CMBTA model requires CPU time comparable to the DMBTA model with five classes and maintains a working path set smaller than that of the DMBTA model with three classes, demonstrating that the CMBTA model can achieve a refined solution with modest computational time and memory requirements.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107320"},"PeriodicalIF":4.3,"publicationDate":"2025-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145413712","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}
Pub Date : 2025-10-21DOI: 10.1016/j.cor.2025.107306
Behrouz Mohammadi-Kordkheili , Rashed Sahraeian , Yasel Costa
Home healthcare centers provide medical services to patients at home, where timely delivery of medications and medical equipment is challenged by heavy traffic, limited accessibility, and supply chain coordination. New technologies like drones can reduce these problems by enabling faster and more flexible deliveries, but their performance is affected by environmental and operational conditions. To address these concerns, this paper proposes a two-echelon supply chain encompassing the main issues related to home healthcare delivery problems. In the first echelon, centers receive medical items from suppliers. In the second, a vehicle and drone routing problem is used to deliver items through multiple paths with a deadline. Drones face limitations including electrical energy consumption, collision risk with power lines, payload capacity, and wind speed. Along different paths, traffic conditions influence vehicle speed, while wind speed affects drone flight. We also present an improved Benders decomposition (IMBD I and IMBD II), using a genetic algorithm (GA) to generate initial subproblem inputs. Valid inequalities for subtour elimination and opening home healthcare centers are implemented in the master problem. To evaluate the proposed methods, we designed 50 instances and compared them with classical Benders decomposition, classical GA, a hybrid GA and variable neighborhood search, and particle swarm optimization. The results indicate that IMBD II outperforms other methods in solution quality and CPU time. Finally, sensitivity analyses on a case study demonstrate the high performance of the model in making accurate decisions amid changes in deadlines, drone constraints, fuel capacity, and supplier closure.
{"title":"Improved Benders decomposition for home healthcare vehicle and drone routing problem in two-echelon supply chains","authors":"Behrouz Mohammadi-Kordkheili , Rashed Sahraeian , Yasel Costa","doi":"10.1016/j.cor.2025.107306","DOIUrl":"10.1016/j.cor.2025.107306","url":null,"abstract":"<div><div>Home healthcare centers provide medical services to patients at home, where timely delivery of medications and medical equipment is challenged by heavy traffic, limited accessibility, and supply chain coordination. New technologies like drones can reduce these problems by enabling faster and more flexible deliveries, but their performance is affected by environmental and operational conditions. To address these concerns, this paper proposes a two-echelon supply chain encompassing the main issues related to home healthcare delivery problems. In the first echelon, centers receive medical items from suppliers. In the second, a vehicle and drone routing problem is used to deliver items through multiple paths with a deadline. Drones face limitations including electrical energy consumption, collision risk with power lines, payload capacity, and wind speed. Along different paths, traffic conditions influence vehicle speed, while wind speed affects drone flight. We also present an improved Benders decomposition (IMBD I and IMBD II), using a genetic algorithm (GA) to generate initial subproblem inputs. Valid inequalities for subtour elimination and opening home healthcare centers are implemented in the master problem. To evaluate the proposed methods, we designed 50 instances and compared them with classical Benders decomposition, classical GA, a hybrid GA and variable neighborhood search, and particle swarm optimization. The results indicate that IMBD II outperforms other methods in solution quality and CPU time. Finally, sensitivity analyses on a case study demonstrate the high performance of the model in making accurate decisions amid changes in deadlines, drone constraints, fuel capacity, and supplier closure.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107306"},"PeriodicalIF":4.3,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145360084","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}
Pub Date : 2025-10-20DOI: 10.1016/j.cor.2025.107308
José Coelho , Mario Vanhoucke
This paper solves the resource-constrained project scheduling problem (RCPSP) with a satisfiability problem (SAT) solver. This paper builds further on various existing SAT models for this well-known project scheduling problem and extends them with two methods to satisfy the resource constraints. Specifically, we use the well-known minimal forbidden sets and compare them with the so-called covers that are traditionally used in SAT implementations. Moreover, we also implement an existing binary decision trees approach under various settings and extend the model with networks with adders, so far never used for solving the RCPSP, to guarantee that resource constraints are satisfied.
The algorithms are tested under different settings on a set of 13,413 project instances with diverse network and resource structures, and the experiments demonstrate that a combination of these approaches help in finding better solutions within a reasonable time. Moreover, 393 new lower bounds, 62 new upper bounds, and 290 optimally solved instances (including 18 from the PSPLIB) have been discovered, which, to the best of our knowledge, had not been found before. The strong performance of the new algorithm motivated additional experiments, and the preliminary results suggest several promising directions for future research.
{"title":"Comparing and extending satisfiability solution methods for the resource-constrained project scheduling problem","authors":"José Coelho , Mario Vanhoucke","doi":"10.1016/j.cor.2025.107308","DOIUrl":"10.1016/j.cor.2025.107308","url":null,"abstract":"<div><div>This paper solves the resource-constrained project scheduling problem (RCPSP) with a satisfiability problem (SAT) solver. This paper builds further on various existing SAT models for this well-known project scheduling problem and extends them with two methods to satisfy the resource constraints. Specifically, we use the well-known <em>minimal forbidden sets</em> and compare them with the so-called <em>covers</em> that are traditionally used in SAT implementations. Moreover, we also implement an existing <em>binary decision trees</em> approach under various settings and extend the model with <em>networks with adders</em>, so far never used for solving the RCPSP, to guarantee that resource constraints are satisfied.</div><div>The algorithms are tested under different settings on a set of 13,413 project instances with diverse network and resource structures, and the experiments demonstrate that a combination of these approaches help in finding better solutions within a reasonable time. Moreover, 393 new lower bounds, 62 new upper bounds, and 290 optimally solved instances (including 18 from the PSPLIB) have been discovered, which, to the best of our knowledge, had not been found before. The strong performance of the new algorithm motivated additional experiments, and the preliminary results suggest several promising directions for future research.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107308"},"PeriodicalIF":4.3,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145360082","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}
Pub Date : 2025-10-16DOI: 10.1016/j.cor.2025.107307
Yulia Karpova, Fulgencia Villa, Eva Vallada
Ambulance relocation is a field of study focused on the repositioning of available ambulances in response to changing conditions to maintain or improve service levels. The scientific community has proposed numerous relocation strategies, often evaluated in environments that simulate the healthcare emergency response process. In order to properly model this process, it is essential to identify the relevant information. Despite efforts to improve realism, many approximations still hinder the accuracy of the results. This article analyzes the information about different ways of determining several aspects of the emergency response process. Furthermore, the importance of incorporating realistic aspects when modeling the emergency response process is shown and the instances generated for the case of Valencia (Spain) are offered based on these aspects.
{"title":"Modeling the process of responding to health emergencies: Main aspects and problems related to evaluating ambulance relocation strategies","authors":"Yulia Karpova, Fulgencia Villa, Eva Vallada","doi":"10.1016/j.cor.2025.107307","DOIUrl":"10.1016/j.cor.2025.107307","url":null,"abstract":"<div><div>Ambulance relocation is a field of study focused on the repositioning of available ambulances in response to changing conditions to maintain or improve service levels. The scientific community has proposed numerous relocation strategies, often evaluated in environments that simulate the healthcare emergency response process. In order to properly model this process, it is essential to identify the relevant information. Despite efforts to improve realism, many approximations still hinder the accuracy of the results. This article analyzes the information about different ways of determining several aspects of the emergency response process. Furthermore, the importance of incorporating realistic aspects when modeling the emergency response process is shown and the instances generated for the case of Valencia (Spain) are offered based on these aspects.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107307"},"PeriodicalIF":4.3,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145360081","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}
Pub Date : 2025-10-15DOI: 10.1016/j.cor.2025.107309
Sebastian Cáceres-Gelvez , Thu Huong Dang , Adam N. Letchford
The permutation flowshop scheduling problem with makespan objective, or PFM for short, is a classic NP-hard scheduling problem. At present, the most promising heuristics for the PFM are based on variations of local search. This led us to consider five new neighbourhoods for the PFM. Each neighbourhood is of exponential size, but can be explored quite quickly by solving a small mixed-integer program. We propose a matheuristic framework that incorporates our proposed neighbourhoods to evaluate and compare their effectiveness. Extensive computational experiments show that integrating our best neighbourhood to the proposed matheuristic reduces the makespan by over 60% on average, compared to the variant without it, on both the classical Taillard benchmark instances and the more recent instances proposed by Vallada, Ruiz and Framinan.
{"title":"MIP-based local search for permutation flowshop scheduling with makespan objective","authors":"Sebastian Cáceres-Gelvez , Thu Huong Dang , Adam N. Letchford","doi":"10.1016/j.cor.2025.107309","DOIUrl":"10.1016/j.cor.2025.107309","url":null,"abstract":"<div><div>The permutation flowshop scheduling problem with makespan objective, or PFM for short, is a classic NP-hard scheduling problem. At present, the most promising heuristics for the PFM are based on variations of local search. This led us to consider five new neighbourhoods for the PFM. Each neighbourhood is of exponential size, but can be explored quite quickly by solving a small mixed-integer program. We propose a <em>matheuristic</em> framework that incorporates our proposed neighbourhoods to evaluate and compare their effectiveness. Extensive computational experiments show that integrating our best neighbourhood to the proposed matheuristic reduces the makespan by over 60% on average, compared to the variant without it, on both the classical Taillard benchmark instances and the more recent instances proposed by Vallada, Ruiz and Framinan.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107309"},"PeriodicalIF":4.3,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145324017","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}
Pub Date : 2025-10-13DOI: 10.1016/j.cor.2025.107301
Sunkanghong Wang , Runqin Wang , Hao Zhang , Fengshi Jing , Qiang Liu , Lijun Wei
This study explores a specific variant of the classic two-dimensional bin-packing problem, known as the Circle Bin-Packing Problem with Rectangular Items (CBPP-RI). This problem involves the orthogonal packing of rectangular items into the fewest possible circular bins and has significant practical implications. We propose a novel and efficient Goal-Driven Iterated Local Search (GDILS) approach for solving CBPP-RI, which integrates a customized method that effectively addresses cold starts and prevents entrapment in local optima. To avoid unnecessary searches, we use lower bounds, which are improved by accounting for the inevitable waste produced by rectangular items at the edges of circular bins. To achieve good performance in single-bin packing, we propose a maximal-space-based heuristic, which introduces the widely used concept of maximal-space from other rectangle packing problems. The experimental results demonstrate that GDILS performs well and show that our method is not only applicable to CBPP-RI but also effective for other related packing problems. To establish a valid benchmark for future research, we also generate a new dataset for CBPP-RI and conduct extensive experiments.
{"title":"A goal-driven iterated local search approach based on the maximal-space for the circle bin-packing problem with rectangular items","authors":"Sunkanghong Wang , Runqin Wang , Hao Zhang , Fengshi Jing , Qiang Liu , Lijun Wei","doi":"10.1016/j.cor.2025.107301","DOIUrl":"10.1016/j.cor.2025.107301","url":null,"abstract":"<div><div>This study explores a specific variant of the classic two-dimensional bin-packing problem, known as the Circle Bin-Packing Problem with Rectangular Items (CBPP-RI). This problem involves the orthogonal packing of rectangular items into the fewest possible circular bins and has significant practical implications. We propose a novel and efficient Goal-Driven Iterated Local Search (GDILS) approach for solving CBPP-RI, which integrates a customized method that effectively addresses cold starts and prevents entrapment in local optima. To avoid unnecessary searches, we use lower bounds, which are improved by accounting for the inevitable waste produced by rectangular items at the edges of circular bins. To achieve good performance in single-bin packing, we propose a maximal-space-based heuristic, which introduces the widely used concept of maximal-space from other rectangle packing problems. The experimental results demonstrate that GDILS performs well and show that our method is not only applicable to CBPP-RI but also effective for other related packing problems. To establish a valid benchmark for future research, we also generate a new dataset for CBPP-RI and conduct extensive experiments.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107301"},"PeriodicalIF":4.3,"publicationDate":"2025-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145277974","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}
Pub Date : 2025-10-11DOI: 10.1016/j.cor.2025.107299
Lei Liu , Matthias Thürer , Shaohua He , Ting Qu , Lin Ma , Zhongfei Zhang
Production systems for customizable products are characterized by operational dynamics due to complicated job routings and machine failures, resulting in dynamic job shop scheduling problems. Dispatching rules are the most commonly used and effective solution approaches for dynamic job shop scheduling problems in practice, but individual rules can only provide short-sight solutions. Hence, adaptively choosing the right dispatching rule according to production system status of job shops is required. To improve the accuracy and robustness of data-driven dispatching rule decision, this paper first introduces a knowledge and data based dynamic job shop scheduling framework consisting of scheduling examples generation, knowledge-guided scheduling features selection and data-driven dispatching rule decision. Then a novel knowledge-guided estimation of distribution algorithm (KEDA) is proposed for the scheduling features selection to enhance the performance of the data-driven dispatching rule decision approaches, where KEDA adopts three types of knowledge-guided improvement strategies, namely mutual information guided population initialization, evolutionary fitness guided probability model update for offspring generation and feature selection ratio guided variable neighborhood search. Comprehensive experiments based on real-life job shop scenarios demonstrate the feasibility of the proposed approach framework and the superiority of KEDA over competitive algorithms for scheduling features selection problems.
{"title":"Scheduling features selection enhanced dispatching decision for dynamic job shop scheduling: a knowledge and data-based approach","authors":"Lei Liu , Matthias Thürer , Shaohua He , Ting Qu , Lin Ma , Zhongfei Zhang","doi":"10.1016/j.cor.2025.107299","DOIUrl":"10.1016/j.cor.2025.107299","url":null,"abstract":"<div><div>Production systems for customizable products are characterized by operational dynamics due to complicated job routings and machine failures, resulting in dynamic job shop scheduling problems. Dispatching rules are the most commonly used and effective solution approaches for dynamic job shop scheduling problems in practice, but individual rules can only provide short-sight solutions. Hence, adaptively choosing the right dispatching rule according to production system status of job shops is required. To improve the accuracy and robustness of data-driven dispatching rule decision, this paper first introduces a knowledge and data based dynamic job shop scheduling framework consisting of scheduling examples generation, knowledge-guided scheduling features selection and data-driven dispatching rule decision. Then a novel knowledge-guided estimation of distribution algorithm (KEDA) is proposed for the scheduling features selection to enhance the performance of the data-driven dispatching rule decision approaches, where KEDA adopts three types of knowledge-guided improvement strategies, namely mutual information guided population initialization, evolutionary fitness guided probability model update for offspring generation and feature selection ratio guided variable neighborhood search. Comprehensive experiments based on real-life job shop scenarios demonstrate the feasibility of the proposed approach framework and the superiority of KEDA over competitive algorithms for scheduling features selection problems.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107299"},"PeriodicalIF":4.3,"publicationDate":"2025-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145324035","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}
Pub Date : 2025-10-11DOI: 10.1016/j.cor.2025.107305
Yuting Yan , Frank (Youhua) Chen , Zhe Fu , Wenjie Bi
We consider the single-item, periodic-review inventory system with stochastic demand, an all-or-nothing random yield pattern, and non-zero leadtimes, where the unmet demand is backlogged. The firm’s objective is to find an optimal ordering policy to minimize its total expected discounted cost. As the optimal policy remains unknown, we first develop a new heuristic that takes advantage of the all-or-nothing yield pattern and then propose a deep reinforcement learning (DRL) algorithm based on a tailored neural network model. The proposed DRL algorithm combines the results of existing heuristics to improve performance and stabilize the training process. Moreover, the total cost to be computed is convex in the decision variable, which facilitates the search. Extensive numerical experiments show that our new solution methods outperform known heuristics and demonstrate near-optimal performance when leadtimes are relatively short and the minimal cost can be achieved numerically.
{"title":"Heuristics and deep reinforcement learning for the inventory problem with an all-or-nothing yield pattern and non-zero leadtimes","authors":"Yuting Yan , Frank (Youhua) Chen , Zhe Fu , Wenjie Bi","doi":"10.1016/j.cor.2025.107305","DOIUrl":"10.1016/j.cor.2025.107305","url":null,"abstract":"<div><div>We consider the single-item, periodic-review inventory system with stochastic demand, an all-or-nothing random yield pattern, and non-zero leadtimes, where the unmet demand is backlogged. The firm’s objective is to find an optimal ordering policy to minimize its total expected discounted cost. As the optimal policy remains unknown, we first develop a new heuristic that takes advantage of the all-or-nothing yield pattern and then propose a deep reinforcement learning (DRL) algorithm based on a tailored neural network model. The proposed DRL algorithm combines the results of existing heuristics to improve performance and stabilize the training process. Moreover, the total cost to be computed is convex in the decision variable, which facilitates the search. Extensive numerical experiments show that our new solution methods outperform known heuristics and demonstrate near-optimal performance when leadtimes are relatively short and the minimal cost can be achieved numerically.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"186 ","pages":"Article 107305"},"PeriodicalIF":4.3,"publicationDate":"2025-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145360083","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号