شبیه سازی ابزار پشتیبانی تصمیم مبتنی بر بهینه سازی برای تولید فولاد
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|5730||2013||8 صفحه PDF||سفارش دهید||6040 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Production Economics, Volume 141, Issue 1, January 2013, Pages 269–276
To buffer against increasing global competition and variability in the price of raw materials, steel manufacturers continuously strive to improve operations and lower costs. In this research, we employ a simulation optimization approach to develop a decision support tool to aid in strategic and operational decision-making. Specifically, we investigate work-in-process inventory levels and potential manufacturing process modifications to reduce utilization costs. A simulation model captures the complex nature of the system while an optimizer searches the solution space and sends trial solutions to the simulation for evaluation. Experimentation suggests that significant daily cost savings are possible by modifying current inventory practices and production process capabilities. Overall, the work demonstrates the ability of the solution approach to analyze complex industrial systems and identify potential improvements in a short time frame.
Fueled by changes in government policies, labor relations, and industry structure, the steel industry has emerged with renewed vitality in the new millennium. Although growth has slowed due to the economic conditions of recent years, Bekaert et al. (2009) suggest strength of global steel intensity (the amount of steel needed per dollar of global GDP) will likely fuel growing demand for years to come. To remain competitive in the industry, which is subject to large amounts of variability in raw material prices and increasing global competition, manufacturers must continuously improve operations and lower costs. In such efforts, many firms implement lean practices and minimize work-in-process (WIP) inventories. While this approach may be beneficial with respect to reducing inventory holding costs, it can lead to production downtime if sufficient “buffer” stock is not available to shield against upstream variability in the manufacturing process. Thus, it is critical that manufacturing firms do not ignore the cost trade-offs which exist between holding inventory and process downtime. Our research focuses on the development of a simulation optimization-based support tool to aid in operational decision-making in complex industrial environments. To illustrate the utility of the solution approach, we work with a steel manufacturing firm to investigate production strategies aimed at cutting costs. While we perform experimentation based on the facility under consideration, this work demonstrates that the general solution approach is likely applicable to other complex industrial environments; assisting firms to make strategic and operational decisions in reasonable time while accurately accounting for inherent system uncertainties. Many researchers agree that simulation-based solution approaches are superior to analytic models for investigating complex stochastic systems (Wang and Chatwin, 2005). However, a review by Taylor et al. (2009) reveals that only 10% of simulation-based papers published from 2000 to 2005 address real-world applications. Furthermore, a survey by Semini et al. (2006) shows that very few simulation studies related to manufacturing system decision-making focus on primary metal industries. Thus, using the steel manufacturing facility under consideration as a testbed, we address this deficiency by employing a simulation optimization solution approach to investigate the impact of varying inventory policies on system performance as well as the impact of manufacturing process improvements in a steel manufacturing environment. Specifically, we determine key work-in-process inventory levels to buffer against upstream process variability while minimizing costs related to inventory holding and production downtime. In this solution approach, a computer simulation model works in concert with a separate optimization component to search for high quality solutions. The simulation model represents the manufacturing process including structural, material/information flow, and logical considerations. The optimization component searches the solution space and dictates trial solutions for the simulation model to evaluate so as to measure solution quality. Moreover, the solution approach provides high quality solutions in a very short time with respect to the performance period. The remainder of this paper is organized as follows. Section 2 reviews the literature pertaining to the use of simulation in manufacturing environments. Section 3 describes the production process and the objectives of the work. Section 4 details the simulation optimization approach used for the investigation. Section 5 presents experimental results and offers managerial insights. Finally, Section 6 summarizes the work and its utility.
نتیجه گیری انگلیسی
In this research, we develop and employ a simulation optimization solution approach in the analysis of a complex system and demonstrate the ability to glean practical insights via extensive experimentation on the system. Specifically, we develop a decision support tool to investigate potential changes to the design and operation of a steel manufacturing facility. Experimentation suggests possible significant cost savings are possible with adjustments to work-in-process inventory levels as well as changes to capacity and processing rates in the production area. Further examination of the results reveals the incremental benefit of particular changes to the system and/or policy. This valuable information may be used to determine which changes to implement based on the practicality of the implementation as well as the potential cost savings. The performance and ease of implementation of the solution approach suggests its applicability to other complex environments, allowing management to make timely and informed strategic and operational decisions which account for existing system uncertainties.