Abstract:In order to achieve balanced allocation of production capacity among multiple factories in a group enterprise and reduce operating costs,this study investigates an integrated scheduling problem for distributed factory collaborative production and transportation.A multi-agent simulation optimization model with the objectives of minimizing total cost and the makespan is established considering practical production characteristics.An improved NSGA-II algorithm is proposed for solution.The algorithm is based on order splitting,allocation,and processing,where a three-stage encoding structure is designed.Furthermore,simulation-based decoding is embedded to guide the action execution of each agent.A semi-random initialization mechanism is developed based on the characteristics of the problem to improve the quality of the initial solution,including three strategies:the number of split sub-orders,the scale of sub-orders,and factory allocation.In addition,a unique equal-probability insertion and splitting strategy is introduced.A multi-point crossover operator with an adjustment strategy is designed,while two specialized mutation operators are proposed considering the impact of sub-order scales on objectives.Finally,experiments are conducted through a real-word production scheduling case from a yeast group,and the results show that the proposed improved algorithm is effective and superior in solving the integrated scheduling problem.

Abstract:During the reentrant multi-purpose batch process with multiple equipment interactions,blockage,preemption and deadlocks occur frequently,leading to significant challenges in collaborative production and high cost of changing production types.Relying solely on intelligent optimization algorithms often results in severe difficulties at both the encoding and decoding stages and fails to provide detailed and feasible scheduling plans.To address these issues,an intelligent optimization framework integrating heuristic rules and high-fidelity simulation is proposed with the objective of minimizing the makespan.Specifically,a simulation execution framework for production systems is established based on production logic,incorporating three rules including push-pull combination,heavy phase blending,and reentrant priority,to ensure efficient and continuous production.Furthermore,equipment allocation rules under various order combinations are explored using the genetic algorithm to guarantee high-quality decoding.Finally,a genetic algorithm incorporating hybrid initialization and fully reachable mutation is designed to optimize batch allocation and production sequencing.Experimental results on 1 000 sets of orders across 28 product types demonstrate that the proposed optimization framework surpasses high-fidelity simulation-based optimization methods based on industrial simulation platforms in terms of stability,effectiveness,and computational efficiency.

Abstract:To address the collaborative dilemmas in smart logistics,such as information silos,lack of trust,and goal conflicts during deep collaboration,a novel solution is proposed from the perspective of large language models(LLMs).The bottlenecks of collaborative management in smart logistics operations are systematically analyzed.The mechanisms through which five key technologies of LLMs—natural language understanding,multimodal perception fusion,agent-based interactive reasoning,simulation and prediction,and resource scheduling optimization—empower collaborative management are elaborated.A four-layer collaborative management framework is constructed,comprising a data layer,a support layer,an application layer,and a guarantee layer.An information collaboration platform is proposed to achieve automated alignment and interoperability of logistics data.A data-driven collaborative negotiation mechanism is designed to provide transparent quantitative analysis of benefits and risks.Furthermore,a global resource optimization system is built to coordinate multiple constraints and reconcile conflicting goals.The framework introduces novel mechanisms for cross-entity semantic interoperability,trust establishment,and global resource optimization,providing systematic theoretical support and a feasible implementation pathway for the evolution of smart logistics from decentralized operations to integrated intelligent collaboration.

Abstract:The efficiency of emergency supply distribution is a core factor influencing the effectiveness of post-disaster relief.However,road damage caused by sudden disasters severely constrains the efficiency of traditional ground transportation.Consequently,coordinated delivery employing Unmanned Aerial Vehicles(UAVs)and trucks has been widely adopted in emergency scenarios such as post-disaster relief and medical supply distribution.To optimize supply distribution efficiency in disaster scenarios,this paper focuses on a truck-UAV coordinated delivery mode.According to the varying accessibility of villages,a coordinated distribution model is established,where trucks serve as mobile depots and UAVs perform multi-sortie operations.With the objective of minimizing the total task load,including truck travel time,UAV flight time,and mutual waiting time,a mixed-integer programming model is formulated,incorporating constraints such as vehicle capacity,UAV payload,and task time windows.This study relaxes the traditional constraint that UAV can only take off and land at fixed nodes,allowing UAVs to perform takeoff,landing,and replenishment at any point along the truck route,thereby enabling in-route replenishment.To solve the model,a two-stage hybrid heuristic algorithm is designed.In the first stage,a genetic algorithm is employed to optimize the basic travel routes of both trucks and UAVs as well as their rendezvous points.In the second stage,dynamic programming is applied to determine the optimal assignment of delivery tasks between trucks and UAVs for villages.Numerical experiment results demonstrate that the proposed"en-route replenishment"strategy significantly reduces both UAV flight time and mutual waiting time by at least 8.3%and up to 31.72%.Compared with the"village-node-only replenishment"strategy,this study provides a scheduling solution that balances timeliness and feasibility,thereby extending the theory of coordinated optimization in emergency management.

Abstract:For the job shop scheduling problem(JSP),a bidirectional coupling scheduling decoding method and a hybrid genetic-tabu algorithm with multi-dimensional enhanced search are proposed with the objective of minimizing the makspan.For the same coded individual,forward and backward active scheduling decoding are performed respectively,and then bidirectional coupling is carried out combining the head-tail lengths of machines and jobs.The proposed method integrate the advantages of left-shift and right-shift operations,enabling more effective utilization of machine idle time and improving decoding quality.This decoding method is integrated into the hybrid algorithm of genetic and tabu search algorithms to solve the JSP.In the process of local search,multiple decoding methods are used to decode a single individual to generate multiple individuals with potentially improved makespan.These individuals are then further optimized through tabu search,achieving multidimensional enhanced search for single individuals.The effectiveness of the algorithm is verified by testing benchmark examples of JSP.

Abstract:A comparative analysis is conducted to evaluate the performance of several classical priority rules for the resource-constrained project scheduling problem with flexible network structures,stochastic activity durations,and random rework.Flexible project scheduling involves two interrelated subproblems,namely activity selection and activity sequencing,which may be addressed using either identical or different priority rules.To more accurately capture the impact of random rework on activity prioritization,a priority evaluation method based on aggregated remaining processing time estimation is proposed.A comprehensive set of test instances covering diverse problem characteristics is constructed,and extensive simulation experiments are conducted to compare the performance of single priority rules and paired priority rules under different scheduling environments.Results show that paired rules significantly outperform single rules,and the best-performing rules differ from those reported in the literature for deterministic problem settings.In addition,project flexibility and resource tightness are found to have significant influence on the performance of priority rules,whereas the effects of other factors are relatively limited.Overall,the TTSL-MSLK paired rule performs best when resources are relatively abundant,while the TTSL-LFT paired rule exhibits superior performance in other settings.These findings provide valuable insights for selecting appropriate scheduling rules in practical engineering applications.

Abstract:The application of automated cluster tools in monolithic wafer processing has effectively improved wafer production efficiency.In order to further improve the utilization of cluster tools,this paper investigates feasible scheduling for dual-arm cluster tools considering concurrent wafer processing and mask change in chambers,while satisfying wafer residency constraints.First,a novel terminal transient scheduling strategy based on a virtual wafer method is proposed.A Petri net model with the new strategy is established and transition triggering conditions are defined to avoid deadlocks for control system operation.Then,according to the process requirements and temporal characteristics,different scheduling conditions are considered,while the waiting time during transient states is redistributed.An algorithm is developed to generate the mask changing time in transient states and the activity sequences of the transfer robot.Finally,the feasibility of the strategy is verified by two examples.Experimental results show that,compared with the traditional approach of mask changing after processing,the proposed scheduling strategy can effectively reduce the batch switching time of single-product wafers without affecting the completion time of terminal transient states,thereby minimizing the total mask changing time.

Abstract:As manufacturing and service systems evolve towards high levels of digitalization,networking,and intelligence,the focus of industrial engineering(IE)is shifting from relatively predictable production systems to complex systems characterized by cross-hierarchy interactions,strong coupling,and continuous evolution.The expansion of system scale,increased demand uncertainty,integration of multi-source heterogeneous data,deepening of human-machine collaboration,and multi-objective conflicts introduced by sustainable development goals pose significant challenges to traditional core methods centered on modeling,prediction,and optimization in terms of complexity representation and decision support.Generative artificial intelligence(GenAI),with its capabilities in cross-modal understanding,conditional generation,and strategic distribution learning,offers a new technological pathway for IE to address planning and operational problems of complex systems.This paper systematically reviews the developmental trajectory of AI-empowered IE,analyzes the characteristics of complexity in future manufacturing and service systems,and focuses on discussing the enabling mechanisms and application value of GenAI in scenarios such as manufacturing system planning,production management,quality control,human-machine collaboration,logistics optimization,and major public health events,supported by typical industrial cases.Results indicate that GenAI contributes to enhancing the flexibility and resilience of complex system decision-making,providing a new direction for the evolution of the IE methodological framework.

Abstract:Against the backdrop of global technological and industrial transformation,the low-altitude economy has emerged as a strategic frontier for reshaping regional competitiveness.However,existing studies offer no thorough investigation into the operational management issues associated with the allocation and pricing of electric vertical take-off and landing(eVTOL)vertiports.To address this gap,this paper designs an allocation and pricing mechanism that enhances social welfare,incentivizes truthful bidding,and mitigates market monopolization.An eVTOL vertiport allocation model based on Vickrey-Clarke-Groves(VCG)auction theory is developed,with Shenzhen as a case study.Through experimental simulations,the performance of the one-shot VCG auction(O-VCG)and the partitioned sequential VCG auction(S-VCG)is compared across various market environments,with focus on the impact of vertiport quantity,operator bidding behavior,cross-regional package preferences,and regional partition granularity on allocation efficiency.Experimental results demonstrate that the optimal auction mechanism depends on the market environment and agent characteristics.(1)Auction mechanisms,especially the S-VCG auction,outperform fixed pricing mechanisms in terms of social welfare.Although fixed pricing may increase short-term platform revenue,it reduces operator profits and market participation.(2)Partitioned auctions show greater robustness across most scenarios where S-VCG comprehensively outperforms O-VCG in platform revenue and achieves higher social welfare when resources are sufficient or bidding activity is high.(3)An optimal granularity exists for regional partition.Increasing regional quantity initially promotes but eventually suppresses social welfare,indicating the need to balance competition incentives with matching efficiency.This study provides low-altitude economy regulators with a scientific resource allocation tool and decision basis based on the auction theory,and confirms the substantial potential of mechanism design to address complex airspace resource management problems.

Abstract:Urban emergency logistics systems face multiple challenges under sudden disaster scenarios,including dynamic demand evolution,real-time fluctuations in road network capacity,and limited inventory resources.Traditional static optimization methods struggle to effectively address environmental uncertainties.To address these issues,this paper proposes a digital twin-driven dynamic optimization approach for emergency distribution and constructs an integrated decision-making framework combining digital twin,optimization modeling,and reinforcement learning.The proposed method comprises three core components.First,a digital twin architecture for urban emergency logistics is established for real-time perception and virtual mapping of physical system states through multi-source data fusion.Second,a multi-period emergency distribution scheduling model considering dynamic inventory constraints is developed.Inventory balance equations are introduced to establish cross-period coupling relationships between distribution decisions and warehousing states,with the objective of minimizing the weighted response time.Third,an adaptive decision-making algorithm based on Deep Q-Network(DQN)is designed.The computational complexity is effectively reduced through state feature selection and action discretization strategies,enabling real-time plan adjustment in dynamic environments.An empirical study is conducted based on the"7·20"extreme rainstorm disaster in Zhengzhou,China.Results demonstrate that,compared with sequential decision-making methods,the proposed dynamic optimization model reduces weighted response time by 21.9％.The digital twin-based real-time decision algorithm maintains a solution feasibility rate of 92.3％in dynamic environments.Moreover,the DQN-based algorithm achieves a 23.2-fold improvement in computational efficiency compared to exact solution methods,with the acceleration ratio reaching up to 45.1 times as the problem scale increases.The research findings provide theoretical support and methodological reference for intelligent management of urban emergency logistics.

Abstract:This paper aims to address the problem of suboptimal pipeline transfer rates in short-term crude oil scheduling using a decomposition approach that converts discrete pipeline transfer rates into a continuous range.A novel decision generation method is introduced to avoid direct search in the continuous transfer rate domain,thereby maintaining algorithm performance.A crude oil scheduling method based on the Soft Actor-Critic(SAC)algorithm is proposed,by reasonably designing state features,action space,and reward function.Five objectives are considered including pipeline mixing costs,tank bottom mixing costs,distiller switching tank costs,charging tank usage costs,and energy consumption costs.Case analysis shows that the SAC-based scheduling method improves single-objective optimization by 1.2%to 77.8%compared to existing methods.

Abstract:To meet the practical demand of cost reduction and efficiency improvement in workshops under the digital manufacturing backdrop,random disturbance events occurring during the assembly process may pose significant risks and losses to factories.In response to disturbances such as urgent order insertion and machine failures encountered in the production of a digital workshop,a rolling time window strategy is employed to select scheduling strategies based on varying production states.During rescheduling,multiple objectives are simultaneously considered,including minimization of the makespan deviation from the initial schedule,minimization of penalty costs,and minimization of the average equipment load.Furthermore,the stability and robustness of rescheduling plans are evaluated from three aspects:equipment deviation,process deviation,and minimum delay time.Based on these considerations,a multi-objective rescheduling model under uncertain disturbances is established.Taking a digital workshop for automobile seat assembly as an example,this study addresses random disturbances in workshop production by integrating a hybrid tabu search genetic algorithm to validate the effectiveness of the proposed scheduling framework and model.Results demonstrate that under time window constraints,the newly generated scheduling plan exhibits excellent stability and robustness in handling uncertain disturbances,thereby expanding new approaches for digital workshops to respond to uncertainty in production.

Abstract:To address issues of operations management and performance evaluation in low-altitude logistics transportation,this paper designs a digital twin-based multi-modal low-altitude logistics delivery system.The proposed system adopts a three-layer architecture consisting of a physical system layer,a simulation engine layer,and an information system layer,enabling visual analysis and operational cost estimation of the designed logistics system.The simulation engine layer adopts Unity3D to build a 3D virtual environment with the same scale as the physical world,achieving visualization of transportation processes and reproduction of logistics behaviors.The information system layer integrates a vehicle-drone collaborative delivery scheduling model.With the objective of minimizing total delivery cost,a hybrid particle swarm optimization algorithm incorporating large neighborhood search is developed to solve the scheduling problem.An improved artificial potential field method is also integrated for drone route planning.The simulation engine layer simulates the scheduling plans and drone trajectories.Based on the simulation results,the operational processes of the logistics system are analyzed,and feasible solutions are subsequently applied to the physical system layer.A digital-twin implementation is conducted for a logistics scenario in an office campus.Results demonstrate that the proposed system can provide efficient vehicle-drone coordination,with safe and feasible drone trajectory plans,verifying the effectiveness of the proposed system.

Abstract:To address the issue of food delivery loss risk,this study proposes a cost-sharing model between a food delivery platform and couriers to improve user satisfaction and courier delivery incentives.The game theory is adopted to analyze the multi-agent interactions in the food delivery market and to clarify the risk transmission mechanism between platforms and couriers.The backward induce method is then utilized to derive the equilibrium results,including the optimal cost-sharing ratio,courier utility,platform profit,and total system utility.Results show that the platform can efficiently stimulate courier incentives,so as to improve overall service efficiency and total system utility by appropriately sharing part of cost related to delivery loss risk.Further analysis shows that the performance of the cost-sharing model is affected by factors such as the weight of courier utility,external risk level,risk sensitivity parameters,delivery time,and order attributes.In high-risk or high-uncertainty scenarios,the platform should assume a larger share of risk cost to avoid reducing courier incentives.In contrast,in medium-and low-risk scenarios,the platform can moderately transfer risk cost to couriers to establish a win-win risk-sharing mechanism.

Abstract:An assembly line model incorporating part shortages is developed to address the issue of reduced production efficiency caused by the shortage of critical parts.The impact of these shortages on production performance is systematically analyzed,and strategies for improvement are proposed.The Markov process is employed for precise analysis of two-station scenarios,while a decomposition-based method is adopted to efficiently approximate the performance of complex multi-station systems.Simulation results demonstrate that the decomposition method significantly enhances computational efficiency while maintaining estimation accuracy,with an average error of less than 0.91%and solution time of under 1 second per iteration.Extensive numerical experiments are conducted to analyze the effects of part shortage probability and part arrival rate on system performance.Results show that,for balanced assembly lines,the optimal location of adjusting part arrival rates and shortage probabilities is mainly concentrated in the central region of the line and its vicinity,with greater improvements observed downstream of the center compared to symmetric locations.It is recommended that the part arrival rate must be adjusted to exceed the minimum processing rate to minimize shortage probabilities and maximize production efficiency.Notably,as the shortage probability decreases,further reductions have a progressively greater impact on throughput improvement.The findings provide theoretical insights and methodological support for the modeling and optimization of assembly lines under part shortage conditions.

Abstract:Financial technology(FinTech),represented by blockchain,helps to overcome information asymmetry in traditional supply chain finance,so as to alleviate the financing difficulties faced by small and medium-sized enterprises(SMEs).However,enterprises often engage in free-riding behavior in FinTech investment,and how government subsidies guide such investment decisions remains insufficiently explored.To this end,an evolutionary game model between banks and financial companies under government subsidies is constructed.The model systematically analyzes additional ecological benefits from unit output effectiveness,investment cost,and technology investment,as well as the impact of free-riding behavior on enterprises' FinTech investment strategies.Results show that when the unit output effectiveness of FinTech investment is at a moderate level,the evolutionary equilibrium converges to two different stable states(i.e.,only the bank or only the financial company participates in investment).The additional ecological benefits of technology investment and the gains from free-riding behavior determine the specific investment decisions of enterprises.Moreover,both the levels and the asymmetry of government subsidies significantly influence enterprises' willingness to invest in FinTech:1)only when subsidies for both banks and financial companies reach relatively high levels will both parties jointly invest in FinTech;2)when both subsidy levels are low,the equilibrium still converges to unilateral investment;3)when subsidies are asymmetric,the party receiving lower subsidies tends not to invest,leading to a subsidy bias that suppresses investment by the disadvantaged party.Finally,free-riding profits substantially weaken investment incentives;effective stimulation of proactive investment can only be achieved when government subsidies offset the gap between investment cost and free-riding gains,or when technological protection reduces free-riding benefits.

Abstract:In the context of increasingly fierce manufacturing competition and growing complex production tasks,accurate production capacity evaluation is essential for enterprises to maintain competitiveness and enhance production efficiency.However,traditional man-hour-based estimation methods demonstrate significant limitations in addressing complex manufacturing environments,particularly in handling dynamic disturbances during production processes.This study proposes a task-driven capacity evaluation simulation system for complex manufacturing environments.The system integrates functional modules including production line definition,order management,capacity analysis,simulation execution,and capacity evaluation.It calculates theoretical effective capacity through analytical methods and estimates actual capacity via the order-driven discrete-time simulation,thereby achieving production simulation and capacity evaluation under non-ideal conditions.A case study conducted in a machining workshop of a manufacturing enterprise demonstrates that the proposed method and system can more accurately evaluate production capacity and calculate equipment utilization in non-ideal environments,providing greater flexibility and reliability compared with traditional approaches.

Abstract:The development of launch vehicle technologies reflects a key indicator of national technological strength.This study proposes a comprehensive evaluation system for the technologies from an"input-output"perspective across four dimensions:technological resources,capabilities,benefits,as well as social and environmental impacts.By integrating the Analytic Hierarchy Process(AHP),expert weighting method,and entropy weighting method,the weight of each indicator is systematically determined,while Grey Relational Analysis(GRA)is applied for comprehensive evaluation.Finally,the effectiveness of the proposed evaluation system and methodology is verified through a case study on sheet metal related technologies at Company T.The established indicator system balances considerations of technology resource allocation,capability development,and final benefits,providing a systematic methodological support for assessing and selecting future rocket manufacturing technologies.Moreover,it offers valuable theoretical and practical references for the technological evaluation of Chinese aerospace industry.

Abstract:To address the practical issues of divergent objectives and weak interconnections between digital-intelligent transformation and servitization in equipment manufacturing enterprises during their parallel advancement,a multi-objective matching model between digital-intelligent technology functions and service types is constructed from the perspective that effective matching of these two elements can promote their integration.The objectives of the model are maximizing the alignment between them and the value creation through digital-intelligent technologies.Corresponding solution methods are provided,while the effectiveness of the proposed theoretical framework and methods are verified through a case study.Results indicate that matching digital-intelligent technology functions and service types is a key factor for the integration of digital-intelligent transformation and servitization in equipment manufacturing enterprises;the bilateral matching model can provide an effective matching;the generated result is presented in the form of digitally empowered servitization.Compared with existing studies,the introduction of bilateral matching theory eliminates potential conflicts in the deep integration of digital-intelligent transformation and servitization,providing a practical theoretical model and decision-making basis for equipment manufacturing enterprises to achieve coordinated transformation.

Abstract:Smart and connected products(SCPs)integrate Internet of Things(IoT)and artificial intelligence technologies,relying on embedded sensors and network connections to achieve real-time interaction,and have become a core driving force for the development of smart homes,Industry 4.0 and smart cities.As an important data source that reflects consumer needs and satisfaction,user reviews provide essential support for product design optimization and user experience improvement.To address the limitations of existing sentiment analysis methods in complex semantic modeling,fuzzy linguistic quantification and interpretability,this paper proposes a comprehensive sentiment analysis framework and satisfaction evaluation method.Firstly,a bidirectional long short-term memory(BiLSTM)network is used to accurately capture semantic dependencies of review texts.Secondly,the probabilistic linguistic term set(PLTS)is utilized to quantify the uncertainty of fuzzy expressions.Furthermore,the contribution of features to sentiment prediction is revealed through Shapley additive explanations(SHAP).Finally,SHAP is combined with importance-performance analysis(IPA)to generate optimized priority recommendations.Taking smart speakers as an example,the effectiveness of the proposed sentiment analysis framework is verified by experiments.It achieves an F1 score of 0.967 and an AUC of 0.979 on the test set,significantly outperforming traditional methods.Based on the results,this paper not only helps to improve the sentiment analysis of SCPs,but also provides scientific basis and practical guidance for product function optimization and personalized recommendations.