Skip to Content

Smart Energy Management
Decentralised Collective Demand–Response Management in Smart Grids

Aim

Develop smart energy management software services providing decentralised decision support, optimisation and automation for collective energy usage at the individual household, building, micro-grid and grid levels using intelligent software agents.

Objective

Enable the financial and environmental efficiencies of the smart energy grids through the coordinated management of supply and demand. This includes local energy usage and overall energy demand collectively optimised according to the individual consumers’ preferences, distribution and transmission needs and constraints, and the energy production and greenhouse gas emission goals.

Background

The emergence of smart energy grids together with smart meters, sensors, devices and appliances promises new financial and environmental efficiencies in the energy market and the overall economy. They have been recognised by the governments and industry around the world, as evidenced by the multi-billion dollar investments and commitments directed towards the development and adoption of new hardware and software technologies supporting the realisation of smart energy grids [1, 2, 6, 7]. The first step is the introduction of smart meters, which can track electricity use in real-time, making that information available to the consumer and the power company. For example Smart Meters are planned to be installed across Victoria by 2013 [3, 4] and the leading energy producers and distributors have already commenced their roll-out.

Benefits

The developed solutions will significantly contribute to the realisation of the full potential of smart grids and achieving the envisioned energy efficiencies. They will offer substantial benefits to the consumers, businesses and society, including:

Consumer:

  • Provide end-users with the automation support tools to more efficiently manage their energy consumption and cost
  • Enable flexible choice of the energy type (e.g. green energy)
  • Optimise individual energy use, generation and storage (e.g. electric cars, solar energy supply/sell, time of usage)
  • Enable local energy markets (e.g. neighbourhood energy trading)

Business:

  • Collective demand response management (e.g. towards constant demand, peak demand minimisation)
  • Dynamic pricing optimisation (e.g. relate price to cost, supply-demand, business rules etc)
  • Eliminate the gap between local usage optimisation and overall/system demand optimisation
  • Coordinated management across individual households, neighbourhoods and distribution system, reducing the erratic and peek demand patters
  • Avoid exuberating the load as a result of all following the same demand-response pattern
  • Competitive advantage (e.g. competitive pricing, add-value services to customers)
  • Reduced cost of infrastructure and maintenance

Society:

  • Greenhouse gas emission reduction

The significant growth in demand for energy is a key challenge globally and in Australia. Managing this demand growth, while still fulfilling customer requirements, can provide significant, readily achievable contributions to emission reduction targets. For example, current demand for energy in Victoria is growing by an average of 1.6% per year and it is projected that at this rate demand for energy will increase by 50% by 2029-30 [1, 2]. Apart from energy created from renewable sources such as wind farms and solar, most of the energy contributes to global warming. The production, supply and use of energy accounts for over 65% of Victoria's greenhouse gas emissions [1]. The ability to optimise energy usage at all levels of the supply chain will become an important sustainability issue.

Australia is a leader in driving improvements in energy efficiency and greenhouse gas abatement, through for example the Victorian Energy Efficiency Target (VEET) scheme and the deployment of new technologies [3, 4]. In particular, starting with the roll out of smart meters Victoria leads the way in deploying smart grid technologies that aim at enabling more efficient energy supply, empowering consumers to more efficiently manage their energy use, and cutting costs and greenhouse gas emissions. It is estimated that smart grid technology can cut electricity consumption by 4 per cent, reduce peak load by 15% and enable 20% reduction in energy greenhouse gas emissions [6]. Data from a smart-grid pilot project carried out by IBM and Consert, which uses smart meters and wireless appliance controllers shows that such a system can reduce energy usage with savings of up to 40 percent [5].

Smart energy management systems are key enablers of the envisioned efficiencies both on the demand and supply sides of the smart energy grids. On the demand side they aim at supporting end-users in optimising their individual energy consumption, e,g, through the deployment of smart meters providing real-time usage and cost of the energy and the use of demand-response appliances that can be switched on/off at a given time depending on the user preferences, energy cost and carbon footprint. On the supply side the smart grid management systems aim at optimising the network load and reliability of the energy provision, e.g. thorough active monitoring and prediction of the energy usage patterns, and pro-active control of the reliable energy delivery over the network. It is also envisaged that they will be able to influence the demand through the dynamic adjustments of the energy price in order to influence the end-user behaviour and energy usage patterns. Although a significant effort and investment have been allocated into the development of smart energy management systems, there are still significant research challenges to be addressed before the promised efficiencies can be realised.

Most existing efforts in smart energy grid technology focus on providing the necessary instrumentation and communication infrastructure, such as smart meters, networking and dash-boards for collecting and providing the information about energy use in real-time. This information will help the end-users in making decisions about their energy use, for example deciding when and which appliance to use depending on the current energy cost. However in order to realise the full potential of smart grid and achieve the envisioned energy efficiency, more advanced, dynamic energy management is required for the coordination, control and optimisation of energy use through the smart grid.

Research Scope

This research aims at developing new decision support and automation models, algorithms and software tools for smart energy management at the individual node (e.g. home), building, micro-grid (e.g. neighbourhood, plant) and grid levels (e.g. state, country). The overall objective is to provide smart management capabilities as software services using negotiating intelligent agents to the users to optimise the energy use at different levels so the individual needs and preferences of the users are satisfied, while the overall objectives of the smart grid are met such as minimised peak lead, energy consumption, greenhouse gas footprint. The specific focus of the research is to develop new models, algorithms and software tools for decentralised coordination, control and optimisation of energy use at:

  • Node level
    Dynamic optimisation of energy use at the household level through coordinated control of demand-response appliances and local energy storage (e.g. plug-in electric car) and generation devices (e.g. solar panel) that can automatically be used at a given time depending on the user preferences and the real-time usage and cost of the energy. It includes design and development of coordinated demand response control algorithms and their implementation as software services available to the users and also the extension of these services to buildings.
  • Micro-grid level
    Adaptive optimisation of the energy load through collective control of energy use patterns and micro-generation within and across the micro-grids. It includes modelling collective energy usage patterns formed by (emerging from) demand response behaviour of the participants, and design and development of decentralised coordination and negotiation mechanisms for the efficient management of energy within the micro-grid.
  • Grid level
    Market-based optimisation of energy supply-demand through the dynamic adjustments of the energy price in order to influence the end-user behaviour and increase the overall energy efficiency of smart energy grid. It includes modelling, design and development of the dynamic pricing strategies (e.g. for time-of-use, peak-day, real-time pricing) and computational market-based mechanisms for smart energy grid.

Our research is conducted in relation to real-world applications. Some of these application domains include automotive, utility, defence, banking, government as well as information technology.

Relevant Capabilities

The Centre for Complex Software Systems and Services (CS3) located at the Faculty of Information and Communication Technologies, Swinburne University of Technology, is one of the leading research centres in Australia and has an international reputation in its research areas with internationally leading researchers. In particular it includes significant research capabilities and activities related to Distributed Systems, Decentralised Optimisation, Artificial Intelligence and Intelligent Agent Technologies. The Centre carries out research in close collaboration with industry and research partners nationally and internationally, supported by a wide range of research grants, strategic partnerships and industrial R&D projects. With some 50 research staff and postgraduate research students, and a substantial portfolio of aligned research activities the Centre is one of the largest and most concentrated research centres in Complex Software Systems and Services in Australia.

This research leverages a number of relevant technology solutions developed at CS3, including:

  • New distributed mechanisms and algorithms for market optimisation with strategic learning agents in market-based resource allocation (e.g. grids, clouds)
  • Automated negotiation mechanisms, and associated algorithms, for consensus-based optimisation in distributed systems (e.g. complex multi-party services, service level agreements)
  • Mechanisms for decentralised management of distributed systems based on self-organising collective optimisation techniques (e.g. traffic management and assignment)
  • Decentralised planning and re-planning techniques for distributed systems (e.g. supply chain)

In addition, CS3 offers a range of complementary capabilities and technologies to support smart grid’s IT infrastructure, including advanced software engineering, web and data, workflow and intelligent agent technologies for building and managing cross-organisational enterprise systems. In particular the Centre has strong capabilities in smart cloud and service-oriented computing that can support the realisation of smart grids with a new powerful, flexible and scalable computing environment where the enormous volume of information generated from disparate sources such as smart meters, sensors, devices and appliances can be collected and tracked in real-time, accessed by the users and processed by smart energy management systems. Cloud computing can provide a core computing environment for the smart grids where the developed smart energy management software services can be provided to the users in ubiquitous and cost-effective manner.

Lead Contact

Prof Ryszard Kowalczyk, rkowalczyk@swin.edu.au