International Journal of Industrial Engineering: Theory, Applications and Practice
Not a member yet
943 research outputs found
Sort by
Multi-Objective Optimization of The Buffer Allocation Problem Considering The Overall Equipment Effectiveness Indicator
Get PDFThis paper presents a formulation of the buffer allocation problem of a parallel serial production line of a footwear stitching process. This was analyzed under a multi-objective optimization approach that aims to maximize the value of the Overall Equipment Effectiveness indicator used in Lean Manufacturing, as well as to minimize the total cost of buffer allocation, this being a new proposal in the buffer allocation problem solution. The case study involves unreliable operating conditions. Process times, times between failures, and repair times are considered Normal distribution functions. The evaluation method used in the study involves the use of a simulation meta-model built from a design of experiments and simulations of the production line; also, the Evolutionary Solver algorithm is used to provide a solution to the mathematical model. The results report the allocation of buffers and their impact on the objectives, as well as a comparison between the optimization criteria
Optimizing Assembly Line Balancing with Carbon Footprint and Spatial Constraints: A Customized Q-Learning Algorithm
No full textAgainst the backdrop of the accelerating transformation of global manufacturing towards intelligence and sustainability, assembly lines have become key to improving production efficiency and flexibility. However, they are often constrained by increasingly stringent carbon footprint regulations. This study focuses on the balancing problem of assembly lines with spatial constraints, aiming to minimize both cycle time and carbon footprint. Given the NP-hard nature of the problem, a mixed-integer programming model was developed, and a Q-Learning algorithm with a restart mechanism was used to solve medium to large-scale problems for efficient resource allocation and carbon reduction. The experimental results demonstrate that the Q-learning algorithm with a restart mechanism exhibits significant performance advantages compared to other heuristics, such as random search and standard Q-learning. Specifically, it achieves a 100% superiority over the compared algorithms in two key metrics: the ratio of non-dominated solutions (Rp) and the convergence degree of non-dominated solutions (Cp). Additionally, it shows a 70% advantage in the spread metric (Sp)
A Deep Learning-based Data-driven Approach for Modeling and Optimization of Laser Transmission Welding of Polypropylene
Get PDFIn this study, a novel multi-stage framework is explored for laser transmission welding of polypropylene by integrating the design of experiments (DoE), artificial neural networks (ANN), non-dominated sorting genetic algorithm-II (NSGA-II), and multi-objective optimization by ratio analysis (MOORA). The framework enables comprehensive experimental investigation, process modeling, and multi-objective optimization. The response surface method (RSM) based DoE is used to develop correlations between welding parameters and responses, which form the foundation for experimental investigations. ANN models, incorporating additional fractional factorial DoE data, are employed for precise non-linear mapping of process parameters and responses, with predictive accuracy surpassing that of RSM models. The 3-6-1 ANN architecture is demonstrated to predict weld strength with high precision, while the 3-7-2-1 model is found to predict weld width accurately. These ANN models are used as objective functions for simultaneous optimization via NSGA-II, generating Pareto-optimal sets. These sets are further prioritized by MOORA, with an optimal parameter set of 220 W laser power, 81.29 mm/s scanning speed, and 63.97 mm defocus distance, yielding a weld strength of 63.86 N/mm and a weld width of 3.24 mm. The proposed synergistic DoE-ANN-NSGA-II-MOORA framework not only confirms its efficacy in this particular case but is also adaptable for other materials and processing applications
Intelligent Collaborative Sustainable Supply Chain Optimization: An Evolutionary Transfer Learning Framework
Get PDFA multi-objective, multi-product, and cross-network sustainable supply chain network design problem is considered. We propose a mixed-integer linear programming model for cross-network collaboration, considering carbon trading, and design an evolutionary transfer learning algorithm to solve the model. The proposed model incorporates economic and environmental objectives, integrating carbon pricing, partner sharing, and logistics collaboration to account for carbon emissions within economic costs, thus achieving multi-objective optimization. The evolutionary transfer learning algorithm incorporates evolutionary procedures and a Markov decision-making process to solve the model efficiently. Extensive experiments based on real-world data are constructed, and the results demonstrate that the proposed method enhances problem-solving efficiency and accuracy across various scenarios while enhancing its stability and robustness. Additionally, case studies of different scales are demonstrated to verify the strong transferability of the proposed method
Model-Driven Transformation of Digital Business Processes
Get PDFDigitalization through technologies such as sensors, information, and communication technologies creates enormous volumes of data, promotes increased internal and external interactions, and opens new opportunities, particularly in maintenance. One of the major impediments to enterprise transformation is the complexity of business processes. This complexity, in turn, manifests itself through stochastic dynamics to degrade performance. Therefore, we propose a framework for business process transformation. The key model for this framework is based on the business process complexity model with a stochastic dynamic (BPCSD) using Markov chains, which estimates the expectation and variance of the processing steps and flow time. We also propose mapping key elements of the widely used Business Process Modeling Notation (BPMN) to BPCSD and validate these in a maintenance operation. In the case study, we present a method to integrate these models and derive options for business process transformation. We identify tacit processes from a communication log and find that including the tacit processes in the BPSCD model improves its predicted expectation and variance of flow times by 2% and 28%, respectively. We also show the application of the BPCSD as a tool for business process transformation. Using the BPCSD, we identify the efficient options that can reduce the flow time and the standard deviation by 20~ 29%. The business processes presented in the case study are pervasive, and the associated digital infrastructure is widely used in a variety of enterprises, which makes the study relevant to maintenance management in many digital enterprises
Implementing Reverse Logistics Optimization in Hazardous Waste Frameworks: Bridging Practice and Strategy
No full textCruise construction entails a protracted logistical cycle, intricate processes, and a vast assortment and quantity of raw materials and intermediate products. Inevitably, reverse materials are generated. To mitigate the potential for stock occupation, production impact, and occupational safety risks stemming from reverse materials, a novel optimization model for reverse logistics networks is proposed, which integrates cost, efficiency and risk factors from a health, safety and environment (HSE) management perspective, adopts a fuzzy operating time window to reflect safety satisfaction, and proposes a three-phase Levy mutation-based Discrete Crow Search Algorithm (DCSA) to derive green, efficient and safe scheduling schemes. The model is able to optimize the allocation of transport vehicles, path planning and risk assessment to improve operational efficiency and safety. To enhance the realism of the model, this paper considers factors such as vehicle load, transport cost, loading time and risk weight, which ultimately achieves better transport efficiency and safety. As well as comparing it with a cost-only scenario, the results show that although cost-oriented scenarios may provide economic benefits in the short term, they tend to ignore the risks that may lead to long-term consequences. This underscores the importance of incorporating safety considerations into operational planning and validates the effectiveness of the model and methodology
An Adaptive Multi-Swarm Gray Wolf Optimizer for Flexible Job Shop Scheduling Problem with Assembly/Disassembly, and Lot Streaming
Get PDFIn the manufacturing/remanufacturing systems of complex products, assembly/disassembly constraints introduce challenges of synchronization and precedence constraints. And large-scale orders have rendered mass production models obsolete. Although existing research on the flexible job shop scheduling problem (FJSP) and its integration with lot streaming (LS) has been extensive, it is still unable to solve these problems simultaneously. To bridge this gap, this paper proposes a novel FJSP with assembly/disassembly and LS (FJSP-AD-LS) to minimize makespan. A MILP model is developed to describe the problem, and an adaptive multi-swarm grey wolf optimizer (AMGWO) is proposed to solve the NP-hard problem. The AMGWO employs a multi-swarm framework with dynamic regrouping to balance exploration and exploitation, and incorporates multiple encoding-decoding methods, evolutionary operators, and knowledge-based neighborhood search methods. A total of 49 new instances are conducted by extending the classical benchmarks. Computational results demonstrate that AMGWO achieves over 10% improvement in makespan compared to the basic algorithm through ablation studies. While commercial MILP solvers could only obtain feasible solutions for 4 small-scale instances within time limits, the proposed algorithm consistently generates high-quality solutions for all test instances. Comparative studies against state-of-the-art methods confirm the superior performance of AMGWO across all 49 instances
Grey Three-Way Decision Approach with The Change of Decision Objects
Get PDFIn practical decision problems, some decision objects may enter or exit the decision system, which will affect the decision results. To deal with dynamic decision problems with uncertain information, we construct a dynamic three-way decision method by exploiting three-way decisions, grey numbers, and grey targets. In this paper, firstly, by considering the similarity of positive and negative bullseye distances, we exploit Topsis and grey target to construct a conditional probability of three decisions. Then we propose a dynamic update rule based on the changing of the decision objects and determine the object's evaluation function and the threshold's calculation method. Finally, a case is used to verify the effectiveness and feasibility of the proposed model
A Novel Multimodal Transport Model for School Commuting
No full textDuring peak hours, the roads surrounding schools often become a source of concern due to severe traffic congestion. This not only leads to a substantial increase in travel time but also poses heightened safety risks. This study proposes a multimodal transport model that integrates private cars, shared parking spaces, and school buses (CPB) to address these challenges. The model aims to improve traffic efficiency and reduce safety hazards. The approach involves two key phases: identifying optimal locations for private car parking and optimizing school bus routes. Results show an 18.53% reduction in private car costs and an 8.13% decrease in traffic delays within the road network. The advantages of this model become particularly significant when student commuting demand exceeds 70% of peak transportation demand. This study provides a robust scientific foundation for developing traffic management strategies around schools
Jointly Optimizing Parallel Batch Processing Scheduling in A Semiconductor Manufacturing Environment
No full textThe rapid growth of the semiconductor industry has led to high water and energy consumption and substantial greenhouse gas emissions. Achieving sustainability in the semiconductor industry has become an exceedingly important issue. This paper investigates a complex batch processing scheduling problem in the final testing phase of semiconductor manufacturing, where chambers and chips are modeled as batch processing machines and jobs. Machines can process multiple jobs simultaneously, with each job defined by its processing time, release time, and size. A mixed-integer linear programming model is presented, along with a constructive-based metaheuristic, the ACS-PBPMs algorithm, to optimize batch formation and scheduling decisions jointly. The algorithm uses an effective candidate list strategy to address constraints and incorporates a local search phase based on solution characteristics. Experimental results on diverse problem instances show that the ACS-PBPMs algorithm outperforms CPLEX and competitive algorithms in computational efficiency and solution quality