Quasi Real-Time Individual Customer Based Forecasting of Energy Load Demand Using In Memory Computing

The goal of the project was to design and implement quasi-real time load forecasting for a big number of individual customers profiles using in-memory computing. We formulated a hypothesis: forecasting accuracy can be substantially increased by individual forecasting on the level of each customer done in quasi real-time using SAP HANA parallelism. The additional goal of the project was also to evaluate efficiency and performance of other computational strategies. Also, the possibility for adding different exogenous variables to the models and being able to quickly analyze their significance opens up new applications of energy load forecasting.

The project was awarded in April 2013 by the HPI with a title: Featured Project of the Month.

Project duration: Fall 2012