Data-driven Intelligent Energy Management System for a Micro Grid

Completed

Renewable Energy Sources(Wind Energy)
Other Power and Storage Technologies(Electricity transmission and distribution)
Other Power and Storage Technologies(Energy storage)
Renewable Energy Sources(Solar Energy)

Basic and strategic applied research

PHYSICAL SCIENCES AND MATHEMATICS (Applied Mathematics)
PHYSICAL SCIENCES AND MATHEMATICS (Computer Science and Informatics)
ENGINEERING AND TECHNOLOGY (Electrical and Electronic Engineering)

Not Cross-cutting

Dr X Zhao
School of Engineering
University of Warwick

Standard

EPSRC

28 June 2018

27 June 2022

48 months

£609,660

Energy

West Midlands

ISCF - Skills

Dr X Zhao, School of Engineering, University of Warwick

Project Contact, Gazprom Marketing & Trading
Project Contact, Carlton Power Limited
Project Contact, WH Power System Consultant
Project Contact, National Grid plc
Project Contact, WATT3 S.A., Switzerland
Project Contact, FTI Consulting, USA
Project Contact, Tsinghua University
Project Contact, Aalborg University

With the fast development of network technology and computing power, a huge amount of data has been generated in almost every aspect of our lives. The International Data Corporation reported that 90 ZB of data will be created each year by 2020, indicating that a big data era is upon us. A typical example is in the energy sector where a large amount of data is generated every day due to smart meter and other digitized changes. These are in turn changing the operation of the energy industry as big data analytics can provide efficient and effective decision support processes. The effect of decentralised generation in the future electricity landscape has and will continue to significantly increase the population of microgrids comprising renewable generation (wind and PV) and battery energy storage supplying local demand, with the excess being exported to the grid. The traditional control design for the energy management system of microgrids is based on a highly simplified model, whose results are highly suboptimal for such a complicated distributed system. Data-driven control could largely improve performance as there is enough data and computing power available today. In addition, energy management systems and market trading optimization packages provided by the big companies are generally designed for large utility and power generation companies and not tailored for smaller prosumers. Given the rapid growth of small prosumers, the PI will develop packages which are tailored to the micro level and meet their individual needs. The PI aims to develop a data-driven intelligent energy management system for a micro grid (connected to a main grid) consisting of wind and solar generation, batteries, and local load in order to provide an integrated, local, smart source of energy. It will use available information (e.g. wind data, weather forecast, energy pricing profile, balancing services pricing etc) to manage the energy generation/utilization and export on site to maximise the financial return to the stakeholder of the microgrid site, and provide balancing services to the System Operator (e.g. my project partner National Grid in the UK). Eventually this will benefit the environment and lead to cheaper energy to the end users due to the improved usage efficiency of renewable energy and the reduced system operation cost

14/09/18