Automatic Power Load Balancing using a Multi-Agent System
| Authors | |
|---|---|
| Year of publication | 2010 |
| Type | Article in Proceedings |
| Conference | Proceedings of EPE 2010 |
| MU Faculty or unit | |
| Citation | |
| Field | Informatics |
| Keywords | Multi-Agent System; Power Load Balancing |
| Description | In this paper, we present an experimental implementation of the regulation on the consumer side of the network, which is based on controlling power consumption of a group of computers located in a small local power network. The regulation is driven by the Rice multi-agent system whose primary purpose is a simulation of power flow in power distribution networks. Intelligent reactions of network elements to different external events influencing the local network and vice versa are simulated by connecting specific agents in the system to physical devices in a power network. The experiment described in the paper deals both with simulated accident events (black-out) and planned events (planned consumption reduction). It shows the way how to intelligently reduce the power consumption in case of an emergency while still allowing to use the computers for full-time computations. The presented technique thus seems to be a basis for intelligent power load balancing in case of cluster or cloud computing networks. |
| Related projects: |