Abstract Emergent Structures in Supply Chains— A Study Inte

Supply chain management is a demanding and complicated task due to its broad scope and the strong connectedness of its objects and issues. In order to make theoretical investigations of supply chains feasible and to support decision-making in real world su

Emergent Structures in Supply Chains—

A Study Integrating Agent-Based and System Dynamics Modeling

Nadine Schieritz and Andreas Größler

Industrieseminar, Mannheim University

D-68131 Mannheim

nadines@is.bwl.uni-mannheim.de

agroe@is.bwl.uni-mannheim.de

Abstract [9]; in some of these studies supply chain models were used within an experimental context, in particular the Supply chain management is a demanding and ‘beer distribution game’ [24] [25].

complicated task due to its broad scope and the strong In order to make theoretical investigations of supply connectedness of its objects and issues. In order to make chains feasible and to support decision-making in real theoretical investigations of supply chains feasible and world supply chains, simulation models are used. to support decision-making in real world supply chains, Simulations allow for systematic testing of supply chain simulation models are used. An integration of system strategies. Such experiments can be conducted without dynamics and discrete agent-based modeling is a being confronted with real world consequences. They promising combination of methods for reducing the a make investigations possible and useful, when in the real priori complexity of the model. The paper discusses the world situation such experimentation would be too costly strengths and weaknesses of system dynamics and or—for ethical reasons—not feasible, or where the discrete agent-based modeling. An approach for decisions and their consequences are too broadly integration of the two modeling methodologies is separated in space or in time. Other reasons for the use of presented. Issues concerning the practical coupling of simulations are the possibility to replicate the initial software environments and a simple, prototypical supply situation, and the opportunity to investigate extreme chain model are discussed. Experiments for which the conditions without risk [19]. The use of simulation as a integrated simulation solution is applied are described. tool for analyzing and evaluating supply chain strategies Insights in emergent structures in supply chains are gained growing attention in recent years [27]. According derived from these simulation analyses. to Parunak et al. many computer-based models developed in the field of supply chain management use system dynamics [16], an approach for modeling and simulating

1. Introduction systems with the help of ordinary differential equations.

Recently, the emerging field of complexity science gained interest in the modeling and simulation of supply

Supply chain management is a demanding and chains leading to a number of agent-based supply chain complicated task due to its broad scope and the strong models.

connectedness of its objects and issues [13]. The In terms of Angerhofer and Angelides, this paper complexity of supply chains does not allow for pure focuses on the presentation of ‘research work on analytical descriptions of the system’s behavior. Even improving the modeling approach’ of supply chains [5]. relatively simple supply chain structures lead inpiduals The outline of the paper is as follows. In the beginning to systematically make sub-optimal decisions, due to the the strengths and weaknesses of system dynamics and chain’s inherent feedback loops (for example, between discrete agent-based modeling are discussed. Special orders and incoming goods) and delays (for instance, emphasis is put on their usability for modeling supply order processing times). The (negative) effects of chains and past applications in this field. In the next part, feedback loops and delays on decision makers’ an approach for integration of the two modeling performance were demonstrated in various studies [7] methodologies (system dynamics and agent-based

Supply chain management is a demanding and complicated task due to its broad scope and the strong connectedness of its objects and issues. In order to make theoretical investigations of supply chains feasible and to support decision-making in real world su

modeling) is presented. A simple, prototypical Agent-based modeling, a simulation methodology implementation of agents in a supply chain and their coming from the field of complexity science, models internal structure is described in the following section. systems comprised of multiple idiosyncratic agents: “One Issues concerning the practical coupling of software of the basic premises of complexity theory is that much of environments (Vensim® and eM-Plant®) are discussed. the apparently complex aggregate behavior in any system The paper ends with describing experiments for which arises from the relatively simple and localized activities the integrated simulation solution is applied. Insights in of its agents.” [17] Therefore the basic building block of emergent structures in supply chains are derived from a system is the inpidual agent—in the supply chain these simulation analyses. case, a company. In contrast to system dynamics, agent- based modeling is a bottom-up approach [6]. The

2. System dynamics and agent-based dynamics of the system arises from the interactions of modeling as modeling paradigms agents whereby the behavior of an agent is determined by its cognitive structure, its schema. “Different agents may System dynamics is a continuous simulation or may not have different schemata…and schemata may methodology that uses concepts from engineering or may not evolve over time. Often agents’ schemata are feedback control theory to model and analyze dynamic modeled as a set of rules, but schemata may be …… 此处隐藏：30450字，全部文档内容请下载后查看。喜欢就下载吧 ……