Agent-Based Modeling for Carpooling

2015 ◽  
pp. 662-688
Author(s):  
Luk Knapen ◽  
Ansar-Ul-Haque Yasar ◽  
Sungjin Cho ◽  
Tom Bellemans
Keyword(s):  
Network Structure ◽  
Large Scale ◽  
Agent Based Modeling ◽  
Transportation Infrastructure ◽  
Problem Size ◽  
Agent Based ◽  
Traffic Demand ◽  
Software Structure ◽  
Large Scale Simulations ◽  
Mutually Independent

Modeling activities and travel for individuals in order to estimate traffic demand leads to large scale simulations. Most current models simulate individuals acting in a mutually independent way except for the use of the shared transportation infrastructure. As soon as cooperation between autonomous individuals is accounted for, the individuals are linked to each other in a network structure and interact with their neighbours in the network while trying to achieve their own goals. In concrete traffic-related problems, those networks can grow very large. Optimization over such networks typically leads to combinatorially explosive problems. In this chapter, the case of providing optimal advice to combine carpooling candidates is considered. First, the advisor software structure is explained; then, the characteristics for the carpooling candidates network derived for Flanders (Belgium) are calculated in order to estimate the problem size.

Agent-Based Modeling for Carpooling

2014 ◽  
pp. 232-258
Author(s):  
Luk Knapen ◽  
Ansar-Ul-Haque Yasar ◽  
Sungjin Cho ◽  
Tom Bellemans
Keyword(s):  
Network Structure ◽  
Large Scale ◽  
Agent Based Modeling ◽  
Transportation Infrastructure ◽  
Problem Size ◽  
Agent Based ◽  
Traffic Demand ◽  
Software Structure ◽  
Large Scale Simulations ◽  
Mutually Independent

Modeling activities and travel for individuals in order to estimate traffic demand leads to large scale simulations. Most current models simulate individuals acting in a mutually independent way except for the use of the shared transportation infrastructure. As soon as cooperation between autonomous individuals is accounted for, the individuals are linked to each other in a network structure and interact with their neighbours in the network while trying to achieve their own goals. In concrete traffic-related problems, those networks can grow very large. Optimization over such networks typically leads to combinatorially explosive problems. In this chapter, the case of providing optimal advice to combine carpooling candidates is considered. First, the advisor software structure is explained; then, the characteristics for the carpooling candidates network derived for Flanders (Belgium) are calculated in order to estimate the problem size.

COMPUTATIONAL FEATURES OF AGENT-BASED MODELS

2005 ◽  
Vol 02 (01) ◽  
pp. 33-48 ◽  
Author(s):  
MASSIMO BERNASCHI ◽  
FILIPPO CASTIGLIONE
Keyword(s):  
Data Structures ◽  
Large Scale ◽  
Agent Based Modeling ◽  
System Response ◽  
Agent Based Models ◽  
Design Data ◽  
Agent Based ◽  
Data Structures And Algorithms ◽  
Immune System Response ◽  
Large Scale Simulations

Agent-based modeling allows the description of very complex systems. To run large scale simulations of agent-based models in a reasonable time, it is crucial to carefully design data structures and algorithms. We describe the main computational features of agent-based models and report about the solutions we adopted in two applications: The simulation of the immune system response and the simulation of the stock market dynamics.

Emergent Behavior in a Population of Thermostatically Controlled Loads With Peer-to-Peer Communication

2019 ◽  
Author(s):  
Ryan Schwartz ◽  
John F. Gardner
Keyword(s):  
Demand Response ◽  
Air Conditioning ◽  
Large Scale ◽  
Large Population ◽  
Agent Based Modeling ◽  
Emergent Behavior ◽  
Peer Communication ◽  
Agent Based ◽  
Thermostatically Controlled Loads ◽  
Residential Air Conditioning

Abstract Thermostatically controlled loads (TCLs) are often considered as a possible resource for demand response (DR) events. However, it is well understood that coordinated control of a large population of previously un-coordinated TCLs may result in load synchronization that results in higher peaks and large uncontrolled swings in aggregate load. In this paper we use agent based modeling to simulate a number of residential air conditioning loads and allow each to communicate a limited amount of information with their nearest neighbors. As a result, we document emergent behavior of this large scale, distributed and nonlinear system. Using the techniques described here, the population of TCLs experienced up to a 30% reduction in peak demand following the DR event. This behavior is shown to be beneficial to the goals of balancing the grid and integrating increasing penetration of variable generators.

scholarly journals A hierarchical multi-resolution agent-based modeling and simulation framework for household electricity demand profile

SIMULATION ◽  
2020 ◽  
Vol 96 (8) ◽  
pp. 655-678 ◽  
Author(s):  
Imran Mahmood ◽  
Quair-tul-ain ◽  
Hasan Arshad Nasir ◽  
Fahad Javed ◽  
José A Aguado
Keyword(s):  
Modeling And Simulation ◽  
Energy Demand ◽  
Large Scale ◽  
Electricity Consumption ◽  
Agent Based Modeling ◽  
Energy Planning ◽  
Modeling Framework ◽  
Agent Based ◽  
The Individual ◽  
Domestic Electricity

Analyzing demand behavior of end consumers is pivotal in long term energy planning. Various models exist for simulating household load profiles to cater different purposes. A macroscopic viewpoint necessitates modeling of a large-scale population at an aggregate level, whereas a microscopic perspective requires measuring loads at a granular level, pertinent to the individual devices of a household. Both aspects have lucrative benefits, instigating the need to combine them into a modeling framework which allows model scalability and flexibility, and to analyze domestic electricity consumption at different resolutions. In this applied research, we propose a multi-resolution agent-based modeling and simulation (ABMS) framework for estimating domestic electricity consumption. Our proposed framework simulates per minute electricity consumption by combining large neighborhoods, the behavior of household individuals, their interactions with the electrical appliances, their sociological habits and the effects of exogenous conditions such as weather and seasons. In comparison with the existing energy models, our framework uniquely provides a hierarchical, multi-scale, multi-resolution implementation using a multi-layer architecture. This allows the modelers flexibility in order to model large-scale neighborhoods at one end, without any loss of expressiveness in modeling microscopic details of individuals’ activities at house level, and energy consumption at the appliance level, at the other end. The validity of our framework is demonstrated using a case study of 264 houses. A validated ABMS framework will support: (a) Effective energy planning; (b) Estimation of the future energy demand; (c) and the analysis of the complex dynamic behavior of the consumers.

Large-Scale Social Simulation, Dealing with Complexity Challenges in High Performance Environments

2014 ◽  
pp. 106-123 ◽  
Author(s):  
C. Montañola-Sales ◽  
X. Rubio-Campillo ◽  
J. Casanovas-Garcia ◽  
J. M. Cela-Espín ◽  
A. Kaplan-Marcusán
Keyword(s):  
Information Technology ◽  
Computer Simulations ◽  
Real World ◽  
High Performance ◽  
Large Scale ◽  
Social Dynamics ◽  
Agent Based Modeling ◽  
Agent Based ◽  
Important Amount ◽  
Distributed Implementation

Advances on information technology in the past decades have provided new tools to assist scientists in the study of social and natural phenomena. Agent-based modeling techniques have flourished recently, encouraging the introduction of computer simulations to examine behavioral patterns in complex human and biological systems. Real-world social dynamics are very complex, containing billions of interacting individuals and an important amount of data (both spatial and social). Dealing with large-scale agent-based models is not an easy task and encounters several challenges. The design of strategies to overcome these challenges represents an opportunity for high performance parallel and distributed implementation. This chapter examines the most relevant aspects to deal with large-scale agent-based simulations in social sciences and revises the developments to confront technological issues.

Sign in / Sign up

Export Citation Format

Share Document