Intelligent agents for electricity demand management

Smart agents: an intelligent way to manage and control energy

CSIRO researchers are developing a range of intelligent, self learning devices known as agents to improve the efficiency of electricity networks.

4 June 2007 | Updated 14 October 2011

Objectives
Background
Application and benefits
Partners
More information

Objectives

The objective of the project – an initiative of the Energy Transformed Flagship – is to optimally manage energy consumption within electricity networks while taking into account consumer-sensitive variables such as:

comfort – for example, heating and cooling
energy costs – ensuring that energy services are commercially viable and affordable.

The CSIRO’s management and control system consists of a range of intelligent, self learning devices or agents that are installed in the target application. Agents are able to:

The inter-connectivity of the intelligent agents provides global appliance supervision.

sense conditions such as temperature
switch appliances and generators based on customer preferences
receive and communicate information such as consumer preferences, price signals or weather conditions between neighbouring sites and remote utilities
process inputs to determine immediate and future optimum actions.

Background

The Australian electricity industry needs to encourage summer peak demand reduction by consumers – domestic and commercial – during heatwaves.

The industry has identified that alternative tariff structures and appliance roll-outs are not the complete answer.

In terms of consumer behaviour, CSIRO studies have shown that direct-control programs that tell consumers what they should do generally do not succeed, as they fail to address the varied needs of individual consumers.

However, a distributed intelligent agent management and control system allows the level of demand-response to be tailored to consumer preferences.

Businesses can reduce their exposure to high wholesale electricity prices.

Application and benefits

The distributed intelligent agent system gives individual consumers real-time information about their power demand and energy pricing during peaks; utilities are provided with a tool to arrange coordinated action by thousands of users.

The inter-connectivity of the intelligent agents provides global appliance supervision. The aggregated response of the agents can address higher order goals like matching consumer demand to intermittent renewable energy supply systems.

As a result, local generation and load control are made more attractive and beneficial as investment options for power generation companies.

Because power retailers are able to manage demand by aggregating consumer-level supply and demand, businesses can reduce their exposure to high wholesale electricity prices.

The project’s scientists see the typical application of this technology being in a number of commercial buildings and domestic residences where there are numerous energy loads and some local generation.