Juan Wang;Guanghui Zhang;Xiaoling Li;Yanxiang Feng
In this research, a novel dynamic and heterogeneous identity based cooperative co-evolutionary algorithm (DHICCA) is proposed for addressing the distributed lot-streaming flowshop scheduling problem (DLSFSP) with the objective to minimize the makespan. A two-layer-vector representation is devised to bridge the solution space of DLSFSP and the search space of DHICCA. In the evolution of DHICCA, population individuals are endowed with heterogeneous identities according to their quality, including superior individuals, ordinary individuals, and inferior individuals, which serve local exploitation, global exploration, and diversified restart, respectively. Because individuals with different identities require different evolutionary mechanisms to fully unleash their respective potentials, identity-specific evolutionary operators are devised to evolve them in a cooperative co-evolutionary way. This is important to use limited population resources to solve complex optimization problems. Specifically, exploitation is carried out on superior individuals by devising three exploitative operators with different intensities based on techniques of variable neighborhood, destruction-construction, and gene targeting. Exploration is executed on ordinary individuals by a newly constructed discrete Jaya algorithm and a probability crossover strategy. In addition, restart is performed on inferior individuals to introduce new evolutionary individuals to the population. After the cooperative co-evolution, all individuals with different identities are merged as a population again, and their identities are dynamically adjusted by new evaluation. The influence of parameters on the algorithm is investigated based on design-of-experiment and comprehensive computational experiments are used to evaluate the performance of all algorithms. The results validate the effectiveness of special designs and show that DHICCA performs more efficient than the existing state-of-the-art algorithms in solving the DLSFSP.
{"title":"Dynamic and Heterogeneous Identity-Based Cooperative Co-Evolution for Distributed Lot-Streaming Flowshop Scheduling Problem","authors":"Juan Wang;Guanghui Zhang;Xiaoling Li;Yanxiang Feng","doi":"10.23919/CSMS.2024.0025","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0025","url":null,"abstract":"In this research, a novel dynamic and heterogeneous identity based cooperative co-evolutionary algorithm (DHICCA) is proposed for addressing the distributed lot-streaming flowshop scheduling problem (DLSFSP) with the objective to minimize the makespan. A two-layer-vector representation is devised to bridge the solution space of DLSFSP and the search space of DHICCA. In the evolution of DHICCA, population individuals are endowed with heterogeneous identities according to their quality, including superior individuals, ordinary individuals, and inferior individuals, which serve local exploitation, global exploration, and diversified restart, respectively. Because individuals with different identities require different evolutionary mechanisms to fully unleash their respective potentials, identity-specific evolutionary operators are devised to evolve them in a cooperative co-evolutionary way. This is important to use limited population resources to solve complex optimization problems. Specifically, exploitation is carried out on superior individuals by devising three exploitative operators with different intensities based on techniques of variable neighborhood, destruction-construction, and gene targeting. Exploration is executed on ordinary individuals by a newly constructed discrete Jaya algorithm and a probability crossover strategy. In addition, restart is performed on inferior individuals to introduce new evolutionary individuals to the population. After the cooperative co-evolution, all individuals with different identities are merged as a population again, and their identities are dynamically adjusted by new evaluation. The influence of parameters on the algorithm is investigated based on design-of-experiment and comprehensive computational experiments are used to evaluate the performance of all algorithms. The results validate the effectiveness of special designs and show that DHICCA performs more efficient than the existing state-of-the-art algorithms in solving the DLSFSP.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"5 1","pages":"86-106"},"PeriodicalIF":0.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10934759","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work addresses bi-objective hybrid flow shop scheduling problems considering consistent sublots (Bi-HFSP_CS). The objectives are to minimize the makespan and total energy consumption. First, the Bi-HFSP_CS is formalized, followed by the establishment of a mathematical model. Second, enhanced version of the artificial bee colony (ABC) algorithms is proposed for tackling the Bi-HFSP_CS. Then, fourteen local search operators are employed to search for better solutions. Two different O-learning tactics are developed to embed into the ABC algorithm to guide the selection of operators throughout the iteration process. Finally, the proposed tactics are assessed for their efficacy through a comparison of the ABC algorithm, its three variants, and three effective algorithms in resolving 95 instances of 35 different problems. The experimental results and analysis showcase that the enhanced ABC algorithm combined with O-learning (QABC1) demonstrates as the top performer for solving concerned problems. This study introduces a novel approach to solve the Bi-HFSP_CS and illustrates its efficacy and superior competitive strength, offering beneficial perspectives for exploration and research in relevant domains.
{"title":"Scheduling Bi-Objective Lot-Streaming Hybrid Flow Shops with Consistent Sublots via an Enhanced Artificial Bee Colony Algorithm","authors":"Benxue Lu;Kaizhou Gao;Peiyong Duan;Adam Slowik","doi":"10.23919/CSMS.2024.0022","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0022","url":null,"abstract":"This work addresses bi-objective hybrid flow shop scheduling problems considering consistent sublots (Bi-HFSP_CS). The objectives are to minimize the makespan and total energy consumption. First, the Bi-HFSP_CS is formalized, followed by the establishment of a mathematical model. Second, enhanced version of the artificial bee colony (ABC) algorithms is proposed for tackling the Bi-HFSP_CS. Then, fourteen local search operators are employed to search for better solutions. Two different O-learning tactics are developed to embed into the ABC algorithm to guide the selection of operators throughout the iteration process. Finally, the proposed tactics are assessed for their efficacy through a comparison of the ABC algorithm, its three variants, and three effective algorithms in resolving 95 instances of 35 different problems. The experimental results and analysis showcase that the enhanced ABC algorithm combined with O-learning (QABC1) demonstrates as the top performer for solving concerned problems. This study introduces a novel approach to solve the Bi-HFSP_CS and illustrates its efficacy and superior competitive strength, offering beneficial perspectives for exploration and research in relevant domains.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"5 1","pages":"46-67"},"PeriodicalIF":0.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10934126","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abdulmajeed Abdullah Mohammed Mokbel;Fei Yu;Yumba Musoya Gracia;Bohong Tan;Hairong Lin;Herbert Ho-Ching Iu
This paper proposes a novel 5D hyperchaotic memristive system based on the Sprott-C system configuration, which greatly improves the complexity of the system to be used for secure communication and signal processing. A critical aspect of this research work is the introduction of a flux-controlled memristor that exhibits chaotic behavior and dynamic responses of the system. To this respect, detailed mathematical modeling and numerical simulations about the stability of the system's equilibria, bifurcations, and hyperchaotic dynamics were conducted and showed a very wide variety of behaviors of great potential in cryptographic applications and secure data transmission. Then, the flexibility and efficiency of the real-time operating environment were demonstrated, and the system was actually implemented on a field-programmable gate array (FPGA) hardware platform. A prototype that confirms the theoretical framework was presented, providing new insights for chaotic systems with practical significance. Finally, we conducted National Institute of Standards and Technology (NIST) testing on the proposed 5D hyperchaotic memristive system, and the results showed that the system has good randomness.
{"title":"Exploring Dynamics and Hardware Implementation of an Enhanced 5D Hyperchaotic Memristive System Inspired by Sprott-C System","authors":"Abdulmajeed Abdullah Mohammed Mokbel;Fei Yu;Yumba Musoya Gracia;Bohong Tan;Hairong Lin;Herbert Ho-Ching Iu","doi":"10.23919/CSMS.2024.0024","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0024","url":null,"abstract":"This paper proposes a novel 5D hyperchaotic memristive system based on the Sprott-C system configuration, which greatly improves the complexity of the system to be used for secure communication and signal processing. A critical aspect of this research work is the introduction of a flux-controlled memristor that exhibits chaotic behavior and dynamic responses of the system. To this respect, detailed mathematical modeling and numerical simulations about the stability of the system's equilibria, bifurcations, and hyperchaotic dynamics were conducted and showed a very wide variety of behaviors of great potential in cryptographic applications and secure data transmission. Then, the flexibility and efficiency of the real-time operating environment were demonstrated, and the system was actually implemented on a field-programmable gate array (FPGA) hardware platform. A prototype that confirms the theoretical framework was presented, providing new insights for chaotic systems with practical significance. Finally, we conducted National Institute of Standards and Technology (NIST) testing on the proposed 5D hyperchaotic memristive system, and the results showed that the system has good randomness.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"5 1","pages":"34-45"},"PeriodicalIF":0.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10934125","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
After-sale service plays an essential role in the electronics retail industry, where providers must supply the required repair parts to consumers during the product warranty period. The rapid evolution of electronic products prevents part suppliers from maintaining continuous production, making it impossible to supply spare parts consistently during the warranty periods and requiring the providers to purchase all necessary spare parts on Last Time Buy (LTB). The uncertainty of customer demand in spare parts brings out difficulties to maintain optimal spare parts inventory. In this paper, we address the challenge of forecasting spare parts demand and optimizing the purchase volumes of spare parts during the regular monthly replenishment period and LTB. First, the problem is well defined and formulated based on the dynamic economic lotsize model. Second, a transfer function model is constructed between historical demand values and product sales, aiming to identify the length of warranty period and forecast the spare part demands. In addition, the linear Model Predictive Control (MPC) scheme is adopted to optimize the purchase volumes of spare part considering the inaccuracy in the demand forecasts. A real-world case considering different categories of spare parts consumption is studied. The results demonstrate that our proposed algorithm outperforms other algorithms in terms of forecasting accuracy and the inventory cost.
{"title":"Spare Part Replenishment Strategy for Electronic Product Based on Model Predictive Control","authors":"Xingchang Fu;Chu-ge Wu;Bo Fu;Yuanqing Xia","doi":"10.23919/CSMS.2024.0027","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0027","url":null,"abstract":"After-sale service plays an essential role in the electronics retail industry, where providers must supply the required repair parts to consumers during the product warranty period. The rapid evolution of electronic products prevents part suppliers from maintaining continuous production, making it impossible to supply spare parts consistently during the warranty periods and requiring the providers to purchase all necessary spare parts on Last Time Buy (LTB). The uncertainty of customer demand in spare parts brings out difficulties to maintain optimal spare parts inventory. In this paper, we address the challenge of forecasting spare parts demand and optimizing the purchase volumes of spare parts during the regular monthly replenishment period and LTB. First, the problem is well defined and formulated based on the dynamic economic lotsize model. Second, a transfer function model is constructed between historical demand values and product sales, aiming to identify the length of warranty period and forecast the spare part demands. In addition, the linear Model Predictive Control (MPC) scheme is adopted to optimize the purchase volumes of spare part considering the inaccuracy in the demand forecasts. A real-world case considering different categories of spare parts consumption is studied. The results demonstrate that our proposed algorithm outperforms other algorithms in terms of forecasting accuracy and the inventory cost.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"5 1","pages":"1-15"},"PeriodicalIF":0.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10934124","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kang Liu;Yang Yu;Yuanjiang Li;Tao Lang;Ruochen Liu
In the field of fault diagnosis for rolling bearings under variable working conditions, significant progress has been made using methods based on unsupervised domain adaptation (UDA). However, most existing UDA methods primarily achieve identification by directly aligning the distributions of the source and target domains, often overlooking the relevance of samples between different domains, which may result in incomplete extraction of deep features and alignment of feature distributions. Therefore, this study proposes a novel domain adaptation network based on Gaussian prior distributions, aiming at solving the challenges of cross working conditions bearing fault diagnosis. The method consists of a feature mining module and an adversarial domain adaptation module. The former effectively extracts deep features by stacking multiple residual networks (Resnet), while the latter employs an indirect latent alignment strategy, using Gaussian prior distributions in the latent feature space to indirectly align the feature distributions of the source and target domains, achieving more precise feature alignment. In addition, an adaptive factor is introduced to dynamically assess the method's transfer and discriminative capabilities. Experimental data from two bearing systems validate that the method can effectively transfer source domain knowledge to the target domain, confirming its effectiveness.
{"title":"RAAN: A Gaussian Prior Domain Adaptive Network for Rolling Bearing Fault Diagnosis Under Variable Working Conditions","authors":"Kang Liu;Yang Yu;Yuanjiang Li;Tao Lang;Ruochen Liu","doi":"10.23919/CSMS.2024.0026","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0026","url":null,"abstract":"In the field of fault diagnosis for rolling bearings under variable working conditions, significant progress has been made using methods based on unsupervised domain adaptation (UDA). However, most existing UDA methods primarily achieve identification by directly aligning the distributions of the source and target domains, often overlooking the relevance of samples between different domains, which may result in incomplete extraction of deep features and alignment of feature distributions. Therefore, this study proposes a novel domain adaptation network based on Gaussian prior distributions, aiming at solving the challenges of cross working conditions bearing fault diagnosis. The method consists of a feature mining module and an adversarial domain adaptation module. The former effectively extracts deep features by stacking multiple residual networks (Resnet), while the latter employs an indirect latent alignment strategy, using Gaussian prior distributions in the latent feature space to indirectly align the feature distributions of the source and target domains, achieving more precise feature alignment. In addition, an adaptive factor is introduced to dynamically assess the method's transfer and discriminative capabilities. Experimental data from two bearing systems validate that the method can effectively transfer source domain knowledge to the target domain, confirming its effectiveness.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"5 1","pages":"16-33"},"PeriodicalIF":0.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10934127","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wei-Li Liu;Yixin Chen;Xiang Li;Jinghui Zhong;Rongjun Chen;Hu Jin
Guardrails commonly play a significant role in guiding pedestrians and managing crowd flow to prevent congestion in public places. However, existing methods of the guardrail layout mainly rely on manual design or mathematical models, which are not flexible or effective enough for crowd control in large public places. To address this limitation, this paper introduces a novel automated optimization framework for guidance guardrails based on a multi-objective evolutionary algorithm. The paper incorporates guidance signs into the guardrails and designs a coding-decoding scheme based on Gray code to enhance the flexibility of the guardrail layout. In addition to optimizing pedestrian passage efficiency and safety, the paper also considers the situation of pedestrian counterflow, making the guardrail layout more practical. Experimental results have demonstrated the effectiveness of the proposed method in alleviating safety hazards caused by potential congestion, as well as its significant improvements in passage efficiency and prevention of pedestrian counterflow.
{"title":"Automatic Optimization of Guidance Guardrail Layout Based on Multi-Objective Evolutionary Algorithm","authors":"Wei-Li Liu;Yixin Chen;Xiang Li;Jinghui Zhong;Rongjun Chen;Hu Jin","doi":"10.23919/CSMS.2024.0020","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0020","url":null,"abstract":"Guardrails commonly play a significant role in guiding pedestrians and managing crowd flow to prevent congestion in public places. However, existing methods of the guardrail layout mainly rely on manual design or mathematical models, which are not flexible or effective enough for crowd control in large public places. To address this limitation, this paper introduces a novel automated optimization framework for guidance guardrails based on a multi-objective evolutionary algorithm. The paper incorporates guidance signs into the guardrails and designs a coding-decoding scheme based on Gray code to enhance the flexibility of the guardrail layout. In addition to optimizing pedestrian passage efficiency and safety, the paper also considers the situation of pedestrian counterflow, making the guardrail layout more practical. Experimental results have demonstrated the effectiveness of the proposed method in alleviating safety hazards caused by potential congestion, as well as its significant improvements in passage efficiency and prevention of pedestrian counterflow.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"4 4","pages":"353-367"},"PeriodicalIF":0.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10820945","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142918104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiao Jing;Xin Pei;Pengpeng Xu;Yun Yue;Chunyang Han
Freeway logistics plays a pivotal role in economic development. Although the rapid development in big data and artificial intelligence motivates long-haul freeway logistics towards informatization and intellectualization, the transportation of bulk commodities still faces serious challenges arisen from dispersed freight demands and the lack of co-ordination among different operators. The present study thereby proposed intelligent algorithms for truck dispatching for freeway logistics. Specifically, our contributions include the establishment of mathematical models for full-truckload (FTL) and less-than-truckload (LTL) transportation modes, respectively, and the introduction of reinforcement learning with deep Q-networks tailored for each transportation mode to improve the decision-making in order acceptance and truck repositioning. Simulation experiments based on the real-world freeway logistics data collected in Guiyang, China show that our algorithms improved operational profitability substantially with a 76% and 30% revenue increase for FTL and LTL modes, respectively, compared with single-stage optimization. These results demonstrate the potential of reinforcement learning in revolutionizing freeway logistics and should lay a foundation for future research in intelligent logistics systems.
{"title":"Reinforcement Learning-Driven Intelligent Truck Dispatching Algorithms for Freeway Logistics","authors":"Xiao Jing;Xin Pei;Pengpeng Xu;Yun Yue;Chunyang Han","doi":"10.23919/CSMS.2024.0016","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0016","url":null,"abstract":"Freeway logistics plays a pivotal role in economic development. Although the rapid development in big data and artificial intelligence motivates long-haul freeway logistics towards informatization and intellectualization, the transportation of bulk commodities still faces serious challenges arisen from dispersed freight demands and the lack of co-ordination among different operators. The present study thereby proposed intelligent algorithms for truck dispatching for freeway logistics. Specifically, our contributions include the establishment of mathematical models for full-truckload (FTL) and less-than-truckload (LTL) transportation modes, respectively, and the introduction of reinforcement learning with deep Q-networks tailored for each transportation mode to improve the decision-making in order acceptance and truck repositioning. Simulation experiments based on the real-world freeway logistics data collected in Guiyang, China show that our algorithms improved operational profitability substantially with a 76% and 30% revenue increase for FTL and LTL modes, respectively, compared with single-stage optimization. These results demonstrate the potential of reinforcement learning in revolutionizing freeway logistics and should lay a foundation for future research in intelligent logistics systems.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"4 4","pages":"368-386"},"PeriodicalIF":0.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10820940","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142918106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Since the increasing demand for surgeries in hospitals, the surgery scheduling problems have attracted extensive attention. This study focuses on solving a surgery scheduling problem with setup time. First, a mathematical model is created to minimize the maximum completion time (makespan) of all surgeries and patient waiting time, simultaneously. The time by the fatigue effect is included in the surgery time, which is caused by doctors' long working time. Second, four mate-heuristics are optimized to address the relevant problems. Three novel strategies are designed to improve the quality of the initial solutions. To improve the convergence of the algorithms, seven local search operators are proposed based on the characteristics of the surgery scheduling problems. Third, Q-learning is used to dynamically choose the optimal local search operator for the current state in each iteration. Finally, by comparing the experimental results of 30 instances, the Q-learning based local search strategy's effectiveness is verified. Among all the compared algorithms, the improved artificial bee colony (ABC) with Q-learning based local search has the best competitiveness.
{"title":"Q-Learning Based Meta-Heuristics for Scheduling Bi-Objective Surgery Problems with Setup Time","authors":"Ruixue Zhang;Hui Yu;Adam Slowik;Kaizhou Gao","doi":"10.23919/CSMS.2024.0021","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0021","url":null,"abstract":"Since the increasing demand for surgeries in hospitals, the surgery scheduling problems have attracted extensive attention. This study focuses on solving a surgery scheduling problem with setup time. First, a mathematical model is created to minimize the maximum completion time (makespan) of all surgeries and patient waiting time, simultaneously. The time by the fatigue effect is included in the surgery time, which is caused by doctors' long working time. Second, four mate-heuristics are optimized to address the relevant problems. Three novel strategies are designed to improve the quality of the initial solutions. To improve the convergence of the algorithms, seven local search operators are proposed based on the characteristics of the surgery scheduling problems. Third, Q-learning is used to dynamically choose the optimal local search operator for the current state in each iteration. Finally, by comparing the experimental results of 30 instances, the Q-learning based local search strategy's effectiveness is verified. Among all the compared algorithms, the improved artificial bee colony (ABC) with Q-learning based local search has the best competitiveness.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"4 4","pages":"321-338"},"PeriodicalIF":0.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10820941","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142918240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Shuttle tankers scheduling is an important task in offshore oil and gas transportation process, which involves operating time window fulfillment, optimal transportation planning, and proper inventory management. However, conventional approaches like Mixed Integer Linear Programming (MILP) or meta heuristic algorithms often fail in long running time. In this paper, a Graph Pointer Network (GPN) based Hierarchical Curriculum Reinforcement Learning (HCRL) method is proposed to solve Shuttle Tankers Scheduling Problem (STSP). The model is trained to divide STSP into voyage and operation stages and generate routing and inventory management decisions sequentially. An asynchronous training strategy is developed to address the coupling between stages. Comparison experiments demonstrate that the proposed HCRL method achieves 12% shorter tour lengths on average compared to heuristic algorithms. Additional experiments validate its generalizability to unseen instances and scalability to larger instances.
{"title":"Graph Pointer Network Based Hierarchical Curriculum Reinforcement Learning Method Solving Shuttle Tankers Scheduling Problem","authors":"Xiaoyong Gao;Yixu Yang;Diao Peng;Shanghe Li;Chaodong Tan;Feifei Li;Tao Chen","doi":"10.23919/CSMS.2024.0017","DOIUrl":"https://doi.org/10.23919/CSMS.2024.0017","url":null,"abstract":"Shuttle tankers scheduling is an important task in offshore oil and gas transportation process, which involves operating time window fulfillment, optimal transportation planning, and proper inventory management. However, conventional approaches like Mixed Integer Linear Programming (MILP) or meta heuristic algorithms often fail in long running time. In this paper, a Graph Pointer Network (GPN) based Hierarchical Curriculum Reinforcement Learning (HCRL) method is proposed to solve Shuttle Tankers Scheduling Problem (STSP). The model is trained to divide STSP into voyage and operation stages and generate routing and inventory management decisions sequentially. An asynchronous training strategy is developed to address the coupling between stages. Comparison experiments demonstrate that the proposed HCRL method achieves 12% shorter tour lengths on average compared to heuristic algorithms. Additional experiments validate its generalizability to unseen instances and scalability to larger instances.","PeriodicalId":65786,"journal":{"name":"复杂系统建模与仿真(英文)","volume":"4 4","pages":"339-352"},"PeriodicalIF":0.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10820942","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142918239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：