In this edition:

High IQ: Smart Agents are getting smarter

“Intelligent energy” research has seen a significant advance in Smart Agent energy management technology over the past year, linking the control of energy generation with the management of demand.

This project contributes to the Energy Transformed National Research Flagship’s Distributed Energy research, which aims to develop and demonstrate energy generation systems that are located close to the consumer and are cleaner and more efficient than conventional energy sources.

Image of circuit boards for smart agent technology

The Energy Transformed Flagship is developing more sophisticated energy management systems through the smart agent project.

Dr Glenn Platt and his team have successfully developed their distributed intelligent agent software to integrate energy generation control with the management of local electricity loads; making Smart Agents even smarter.

“We've been working on the introduction of distributed generation control into our agent-based energy management system. Here, an agent is associated with a distributed generator, such as a microturbine or a small solar photovoltaic system, monitoring conditions such as fuel price, or sunlight predictions, or local electricity needs, and controlling the generator with these local issues in mind,” Dr Platt explains.

“With the introduction of generation into the agent system, the whole system of electricity loads and generators has great flexibility in working to address a global goal.

“For example, given a goal such as ‘Keep the total electricity purchased from the grid by the Newcastle Energy Centre below 150kW’, the agents may decide to turn an electricity load at this site off, or, conversely, turn a local electricity generator on.

“In deciding which path to follow, the agents will consider things such as the energy available to the generator, and what this will cost, what the site's energy load is likely to be later in the day, and so on.”

Through linking intelligent control of electricity loads and generation, this system of Smart Agent management allows for optimisation of energy consumption and increases the options available for management goals.

Continuing work within the Flagship is now focussed on introducing ‘prediction’ agents into the management system.

“These agents try to predict the power that will be available some time in the future from renewable energy sources such as solar. We can then optimise how an entire network utilises energy, to bring maximum benefit from the solar system,” Dr Platt said.

Through the further ‘education’ of Smart Agents, a more sophisticated energy management system is emerging which confers many benefits to the consumer and achieves the major Flagship goal of reducing greenhouse gas emissions as a consequence of energy consumption.