The problem considered in the paper is motivated by production planning in a foundry equipped with the furnace and casting line, which provides a variety of castings in various grades of cast iron/steel for a large number of customers. The quantity of molten metal does not exceed the capacity of the furnace, the load is a particular type of metal from which the products are made in the automatic casting lines. The goal is to create the order of the melted metal loads to prevent delays in delivery of goods to customers. This problem is generally considered as a lot-sizing and scheduling problem. The paper describes two computational intelligence algorithms for simultaneous grouping and scheduling tasks and presents the results achieved by these algorithms for example test problems.
Mathematical programming, constraint programming and computational intelligence techniques, presented in the literature in the field of operations research and production management, are generally inadequate for planning real-life production process. These methods are in fact dedicated to solving the standard problems such as shop floor scheduling or lot-sizing, or their simple combinations such as scheduling with batching. Whereas many real-world production planning problems require the simultaneous solution of several problems (in addition to task scheduling and lot-sizing, the problems such as cutting, workforce scheduling, packing and transport issues), including the problems that are difficult to structure. The article presents examples and classification of production planning and scheduling systems in the foundry industry described in the literature, and also outlines the possible development directions of models and algorithms used in such systems.
The paper presents a novel Iterated Local Search (ILS) algorithm to solve multi-item multi-family capacitated lot-sizing problem with setup costs independent of the family sequence. The model has a direct application to real production planning in foundry industry, where the goal is to create the batches of manufactured castings and the sequence of the melted metal loads to prevent delays in delivery of goods to clients. We extended existing models by introducing minimal utilization of furnace capacity during preparing melted alloy. We developed simple and fast ILS algorithm with problem-specific operators that are responsible for the local search procedure. The computational experiments on ten instances of the problem showed that the presence of minimum furnace utilization constraint has great impact on economic and technological conditions of castings production. For all test instances the proposed heuristic is able to provide the results that are comparable to state-of-the art commercial solver.
Systemic approach to design of factories requires that engineering, organisational and economic aspects should be considered concurrently. That prompts the need to develop a solution, based on the state-of-the-art IT technologies, to enable us to solve the problems associated with foundry production planning. The paper outlines a methodology of creating the simulation model of a virtual foundry, as a tool for foundry design. An integrative approach is suggested for development of a complete foundry model, enabling the design of more efficient production systems. The underlying principles of such models are discussed, the basic stages involved in the methodology are outlined and the range of its applicability is defined.Keywords: Virtual foundry, Modeling, SimulationSystemowe podejście do projektowania bądź rekonstrukcji systemów wytwarzania wymaga równoległego rozpatrywania zagadnień technologicznych, technicznych, organizacyjnych i ekonomicznych. Stwarza to potrzebę opracowania rozwiązania, opartego na najnowszych osiągnięciach technologii informatycznych, pozwalającego na kompleksowe rozwiązywanie problemu projektowania systemów produkcji odlewniczej. W pracy przedstawiono metodykę tworzenia (budowy) modelu symulacyjnego tzw. wirtualnej odlewni jako narzędzia projektowania zakładu odlewniczego. Zaproponowano iteracyjne podejście tworzenia kompleksowego modelu odlewni, dające w efekcie możliwość projektowania bardziej wydajnych systemów wytwórczych. Przedstawiono zasady konstrukcji takiego modelu, opisano podstawowe etapy metodyki oraz określono możliwości jej zastosowania.
In the paper we studied a production planning problem in a mid-size foundry that provides tailor-made cast products in small lots for a large number of clients. Assuming that a production bottleneck is the furnace, a mixed-integer programming (MIP) model is proposed to determine the lot size of the items and the required alloys to be produced during each period of the finite planning horizon that is subdivided into smaller periods. As using an advanced commercial MIP solvers may be impractical for more complex and large problem instances, we proposed and compared a few computational intelligence heuristics i.e. tabu search, genetic algorithm and differential evolution. The examination showed that heuristic approaches can provide a good compromise between speed and quality of solutions and can be used in real-world production planning.Keywords: Application of information technology to the foundry industry, Production planning, Lot-sizing problem W pracy przedstawiono problem planowania produkcji w odlewni średniej wielkości, która dostarcza odlewy na zamówie-nie dla dużej liczby klientów. W takim problemie konieczne jest określenie wielkości partii produkcyjnej oraz ilości i gatunku metalu w każdym okresie skończonego horyzontu planowania, który jest podzielony na mniejsze podokresy. Przy założeniu, że wąskim gardłem jest piec do topienia metalu, zaproponowano programowanie całkowitoliczbowe mieszane (Mixed-Integer Programming -MIP) jako model planowania i harmonogramowania produkcji w odlewni. Jako że użycie zaawansowanych komercyjnych solverów może być niepraktyczne dla złożonych problemów, zaproponowano i porównano trzy heurystyki inteligencji obliczeniowej tj. tabu search, algorytm genetyczny i ewolucja różnicowa. Eksperymenty obliczeniowe wykazały, że algorytmy heurystyczne zapewniają zadowalającą szybkość i jakość rozwiązań.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.