Enabling a More Reliable Smart Grid with Big Data
The Royal Society
Smart grids are expected to play a central role in the transition to a low-carbon energy future. This would help meet the UK target of an 80% reduction in CO2 emissions by 2050, the European target of a 20% share of renewable energy sources by 2020. However, the integration of renewable energy resources into the power grid has brought a significant challenge to the reliability of the existing power infrastructure due to their highly variable nature. This project aims to develop advanced predictive data analytics empowered with high performance computing techniques such as cloud computing to enable a more reliable and efficient power grid.
Electricity power grid operators have traditionally had to contend with large amounts of data, from sources such as Supervisory Control And Data Acquisition (SCADA) systems to deduce useful information about the condition of the grid.
Future smart grids will, however, take the volume of data up by orders of magnitude both in terms of size and frequency of measurement as high resolution data sources are integrated into the system. Some of the key sources of this future massive data include synchrophasor data from the increased deployment of Phasor Measurement Units (PMUs), energy user data from the emerging smart meters, high-rate digital data from increasingly networked Digital Fault Recorders (DFR), fine-grained weather data and energy market pricing signals. This high resolution data on grid, environmental and market conditions has the potential to tremendously improve the reliability of the power grid enabling much closer-to-optimal planning and operation of the grid. However, the volume, velocity, variety, and the complications of value and veracity of this set of data not only qualify it as big data, but as bigger data which necessitates big data analytics.
This project aims to develop advanced predictive big data analytics empowered with high performance computing techniques such as cloud computing to enable a more reliable and efficient power grid. Specifically, the objectives of the project are:
1. To devise fast dimension reduction approaches to reducing the computation complexity in processing data of high dimensions. For this purpose, parallel computing techniques will be employed to decompose a large problem into a set of small problems which can be executed in parallel.
2. To develop hierarchical data clustering techniques to deal with data of highly noisy.
3. To investigate a scalable computing platform which supports the processing of ever increasing data sets.