Discrete-Event Simulation (eBook)

Concepts and Production in Arena

Abdessalem Jerbi (Autor)

eBook Download: EPUB

2024
426 Seiten
Wiley-Iste (Verlag)
978-1-394-33222-9 (ISBN)

Infos

The use of discrete-event simulation in various fields, such as in industry, logistics and public health, has really taken off over the last few decades. The implementation of discrete-event simulation does however require an understanding, and perhaps even a mastery, of precise theoretical and methodological principles.
Discrete-Event Simulation presents the key concepts involved in any discrete-event simulation project, covering the most frequently used techniques for analysing data and results, the methodological and practical aspects of implementing discrete-event simulation, along with an introduction to the use of the 'Arena' discrete-event simulation tool. This book combines the elements presented with applied examples, as well as numerous examples of simulation projects in various fields.

Abdessalem Jerbi is a doctoral engineer in mechanical engineering and has taught flow simulation at the University of Sfax in Tunisia. His research is interdisciplinary, with a particular interest in simulation-based optimization

Textauszug

1
Simulation

1.1. Introduction

Simulation aims to imitate the behavior of real systems on a computer equipped with appropriate software. It is generally used to analyze systems and make operational or resource policy decisions. Nowadays, simulation is a tool that is becoming increasingly popular, as computers and software become increasingly efficient (Chung 2003; Fleury et al. 2006; Altiok and Melamed 2007; Kelton et al. 2015; Polenghi et al. 2018).

The study of a system leads in most cases to modeling its functioning through the establishment of mathematical or logical relationships. It is possible to use mathematical methods if these relationships are simple enough. This solution is then called an analytical solution. However, systems in reality are most often too complex to be able to apply such an evaluation. We then resort to simulation in order to estimate the desired characteristics of the model.

Let us take the example of an industrial company which seeks to expand without being sure of the potential gain generated by this expansion. Indeed, it would not be profitable for the company to invest money for an extension, then later remove this extension, if the latter is deemed unprofitable. A simulation study could enlighten decision-makers on the consequences of this extension by simulating the operation of the factory before and after extension. We then achieve results without making physical changes in the factory.

Simulation is a process which therefore consists of designing a model of a real system, carrying out experiments on this model, interpreting the observations provided by the execution of the simulation of the model and formulating decisions relating to the system. The goal of this process can be to compare different configurations of the system studied or to evaluate different strategies for its control in order to optimize its performances.

The fields of application of simulation are numerous and varied. A non-exhaustive list of problems for which simulation has proven to be a useful and powerful tool includes:

• production flow systems (Addi Ait 2000):
  • machining operations: simulations of machining operations may include processes involving manually or computer-controlled factory equipment for machining, turning, bending, cutting, and welding,
  • assembly operations: assembly operations simulation can cover any type of assembly line or manufacturing operation that requires the assembly of multiple components into a single part,
• logistics flows and transport systems:
  • material handling equipment: material handling simulations include analysis of cranes, forklifts and automated guided vehicles,
  • warehousing: warehousing simulations may involve manual or automated storage and retrieval of raw materials or finished products;
• the production of services:
  • hospitals and medical clinics: models of hospitals and medical clinics can be simulated to determine the number of rooms, nurses and doctors for a particular location,
  • retail stores: retail stores may need to know how many checkouts to use,
  • food or entertainment facilities: entertainment facilities, such as multi-theater movie theater complexes, may be interested in the number of ticket sellers, ticket checkers or concession stand attendants to employ,
  • information technology: information technology models generally concern the number and type of networks or support resources to be made available,
  • customer ordering systems: customer ordering systems may need to know how many customer order representatives are needed on duty.

1.2. Advantages of simulation

1.2.1. Compressed time experimentation

As the model is simulated on a computer, experimental simulations can be carried out in compressed time. This is a major advantage because some processes can take months or even years to complete. Long system processing times can make robust analysis difficult to achieve. With a computer model, the operation of long processes can be simulated in seconds. Additionally, multiple repetitions of each simulation can easily be performed to increase the statistical reliability of the analysis (Kelton et al. 2015).

1.2.2. Reduced analytical requirements

Before the existence of digital simulation, we were forced to use other tools that were more analytically demanding. Even then, only simple systems involving probabilistic elements could be analyzed. More complex systems were strictly the domain of the mathematical researcher or operations research analyst. Furthermore, systems could only be analyzed with a static approach at a given point in time. On the other hand, the advent of simulation methodologies has made it possible to study systems dynamically in real time. Additionally, the development of simulation-specific software has eliminated many of the complicated basic calculations and programming requirements that might otherwise have been necessary. These reduced analytical requirements have made it possible to analyze many more different types of systems than before with a greater variety of experiments (Kelton et al. 2015; Polenghi et al. 2018).

1.2.3. Easy to demonstrate models

Most simulation software has the ability to dynamically animate the operation of the model. Animation is useful both for debugging the model and for verifying and demonstrating its operation. Animation-based debugging makes it easy to observe flaws in model logic. Using animation during a presentation can help establish the credibility of the model through the dynamic demonstration of how the system model handles different situations. Without the capability of animation, we would be limited to less effective textual and digital presentations (Kelton et al. 2015).

1.3. Disadvantages of simulation

Although simulation has many advantages, there are also some disadvantages. These drawbacks are not really associated directly with the modeling and analysis of a system, but rather with the expectations associated with simulation projects (Chung 2003; Fleury et al. 2006; Melamed and Rutgers 2007). These disadvantages are as follows:

• Under no circumstances can simulation give accurate results when the input data are inaccurate: no matter how good a model is, if its input data are not accurate, we cannot reasonably expect to obtain accurate output data. Unfortunately, the data collection phase is considered the most difficult part of the simulation process.

Nevertheless, it is common that little time is allocated to this phase. In many cases of simulations, historical data of questionable quality have been accepted in order to save input data collection time. Too often, the exact nature or conditions under which these data are collected are unknown.

• Simulation alone cannot solve problems: some managers may believe that carrying out a model simulation project is enough to solve the problem they are facing. However, simulation by itself cannot actually solve the problem. It can provide direction for potential solutions to resolve the problem and it is up to the manager to implement the proposed changes.

1.4. Concepts relating to simulation

1.4.1. System concept

Different definitions have been given to the word “system” in the field of simulation (Kelton et al. 2015). Here are some examples:

• a system is a set of components linked together;
• a system is an organized set of functional elements;
• a system is a combination of parts that coordinate to achieve a result so as to form a set;
• a system is a combination of people, machines, materials and information intended to satisfy a given objective.

All these definitions have in common a few key words which characterize the notion of system. Indeed, a system is characterized by ordered parts which compose it. Each of these parts has its own laws and a certain independence. These parts also have links or relationships between them. Furthermore, the whole thing, or the system, changes over time and is influenced by the environment in which it exists and which reacts on it. Finally, this set is most often subject to constraints and only exists to achieve a goal.

In summary, a system can be defined by the knowledge of its parts or its components, the laws specific to each component and the interactions which determine its purpose.

Example: in a production plant, the parts that make up the system could be the workforce and departments of purchasing and supply, inventory management, manufacturing and production scheduling, sales and administrative. Each of these services has its own operating laws and is partially independent of the others. But also, they interact with each other. Knowledge of these interactions and their application to the system will significantly influence the goal to be achieved by the system such as profitability, investment, profit, etc.

1.4.2. Model and modeling

A model is a representation of a system whose goal is to explain and/or predict certain aspects of its behavior. This representation is more or less faithful, because, on the one hand, the model must be sufficiently complete in order to be able to answer the various questions that can be asked about the system it represents and, on the other hand, it must not be too complex to...

Produkt-Details

• Erscheint lt. Verlag: 30.10.2024
• Reihe/Serie: ISTE Consignment
• Sprache: englisch
• Themenwelt: Technik ► Bauwesen
• Schlagworte: Arena • Data Analysis • Discrete-Event Simulation • logistics • Public Health
• ISBN-10: 1-394-33222-X / 139433222X
• ISBN-13: 978-1-394-33222-9 / 9781394332229