Technical gazette, Vol. 25 No. 5, 2018.
Original scientific paper
https://doi.org/10.17559/TV-20161221220842
Robustness Analysis of Holonic Multi-Agent Systems: Application to Traffic Signals Control
Monireh Abdoos
; Faculty of Computer Science and Engineering, Shahid Beheshti University, Velenjak, 19839-63113 Tehran, Iran
Abstract
Multi-Agent Systems (MASs) provide a powerful tool to model distributed systems. Large-scale systems contain many autonomous agents, and therefore, the agents should be able to work in a group and collaborate toward common objectives. Holonic Multi-Agent Systems (HMASs) present a suitable organization, especially in large-scale systems. The idea behind HMASs is a division of a system into smaller sub-systems in a recurrent way. A holon is defined as a self-similar structure that comprises holons as sub-structures. Therefore, a holarchy is a hierarchy of holons that act as autonomous wholes in super-ordination to their parts and as dependent parts in sub-ordination to controls on higher levels. There are two main attributes for a holarchy, the first attribute ensures that holons are in stable forms, which are robust against disturbances. The second one confirms that the holons are in intermediate forms, which provide the proper functionality for the whole. In this paper, we study the robustness of a holarchy for traffic signals control. Robustness is an essential feature for providing reliable solutions, especially in real world applications. We show that holonic MAS can be effectively used for traffic signals control as a robust modeling method.
Keywords
Holonic Multi-Agent Systems; robustness; statistical test; traffic signals control
Hrčak ID:
207427
URI
Publication date:
28.10.2018.
Visits: 1.773 *