China
Africa
Explore chapters and articles related to this topic
Solutions to the Decisions: Review and Suggestions
Published in Hassan Rashidi, Edward P. K. Tsang, Port Automation and Vehicle Scheduling, 2023
Hassan Rashidi, Edward P. K. Tsang
This type of simulation is a relatively new approach to modeling systems into which agents interact. Agents are individual, autonomous, and their behaviors affect others. A typical agent-based model has three elements: (a) Agents, their attributes and behaviors; (b) Agent relationships and methods of interaction; and (c) Agents' environment. For making a model for container terminal, we have four different types of agents in the system: (a) the Quay Crane Agents (QCAs); (b) Straddle Carrier Agents (SCAs); (c) the Traffic Agents (TAs), and (d) the Area Manager Agent (AMA). Each QC is controlled by a QCA, each straddle carrier is controlled by an SCA, and each cell of the yard highway that contains more than one entry point, such as a crossing, is governed by a TA. Desktop computing for ABS application development are Spreadsheets, Dedicated Agent-based Prototyping Environments, and General Computational Mathematics Systems such as MATLAB and Mathematica. For ABS, large-scale agent development environments are Repast, Swarm, MASON, and AnyLogic.
Agent-Based Models for Water-Related Disaster Risk Management
Published in Neiler de Jesús Medina Peña, Adaptive Disaster Risk Assessment Combining Multi-Hazards With Socioeconomic Vulnerability and Dynamic Exposure, 2021
In terms of software use, we cannot conclude which one is better to use to implement ABMs for WR-DRM, since that depends on modellers’ preferences, abilities and also the scope of the model itself. Nevertheless, we encourage modellers to use Netlogo for basic and proof of concepts because it is easy to learn and use, and the ability to run prototype models. We recommend the use of Repast Symphony for more robust implementations because it can handle high computational demands, a larger number of agents and interactions and better representation of a more complex environment (space and time). In addition, with the sophistication of models and the increase in the scale of simulation, it is expected as well a surge in computational demand. In such cases, ABM tools that support parallel computing would be required, which Repast HPC, Swarm and MATSIM are the most promising tools (Abar et al., 2017).
Reflections and outlook
Published in Yared Abayneh Abebe, Modelling Human-Flood Interactions, 2021
On ABM software implementation After developing conceptual models, converting such models to an ABM software is a daunting task. The main reason is the “nature” of the ABM modelling paradigm. In hydrodynamic modelling, 2D surface water flow in any case study can be modelled using the shallow water equations—a mass equation and two momentum equations in the x and y directions. If a modeller knows the initial conditions, the boundary conditions and the model parameter values of any study area, an off-the-shelf hydrodynamic modelling software such as MIKE21 solves the equations numerically and provide outputs such as water level and discharge at each computational cell. Unfortunately, there is no universal way of describing human behaviour in an ABM, especially considering heterogeneous agents and their interactions. In fact, there is no such ABM software. There are only ABM development environments such as NetLogo and Repast Simphony that provide the platform to write lines of codes that describe the conceptual model. Thus, developing ABM software requires a “certain” level of programming skills. In relation to that, using different ABM development platform requires knowledge of different programming languages. For example, NetLogo uses a simplified logo language while Repast Simphony uses the Java programming language. Besides, as every case is different, the modeller needs to develop the ABM software for every case. Researchers/modellers that will build ABMs should consider this fact while designing their research plan.
The effect of replenishment policy on bullwhip effect considering the higher-level shipments
Published in Journal of Industrial and Production Engineering, 2018
Dayong Wang, Guozhu Jia, Chunting Liu, Hengshan Zong, Wei He
In view of the above problems, this study has established a multi-level and multi-node supply chain simulation model, proposed a comparative analysis of the effect of different replenishment strategies on bullwhip effect in supply chain with or without the consideration of the superior shipments. This study not only analyzes the bullwhip effect of every level of the supply chain, but also the bullwhip effect of each node enterprise. The simulation model is a multi-agent model built on the Repast platform. Repast is a set of simulation kits that allow to build agent simulation models more easily and quickly. Repast is widely used in many fields of supply chain study [17].