Data Mining for Energy Analysis

Source de subvention

Professeur(e)s impliqués

• Mohamed Bouguessa (UQAM)

Résumé

The advent of information technology and electronic have facilitated the recording process of energyconsumption of costumers using smart meters. The recorded information can be explored to figure out relevantinsights. However, digging into this data pose a serious problem in the sense we have tremendous amount ofcostumers, which present various behaviour at different time point. Motivated by the need of effective softwarefor analyzing energy-related big data, SAP Canada Inc. (« SAP »), is developing SAP Energy Analysis, a newgeneration of real time business intelligence software for energy-related issues. SAP Energy Analysis offers aset of tools to run the data collected from smart meters and any other monitoring systems into actionableinsights. Specifically, SAP Energy Analysis provides tools to manage the cost effectiveness of energyproduction and consumption in the form of interactive dashboard, personalized report, real time businessactivity monitoring, prescriptive analysis, and more. By effectively analysing customers’ consumption datausing SAP Energy Analysis, it becomes possible to monitor the energy production and consumption and, thus,allow better load balance, load matching and load scheduling through improving energy efficiencies andmanagement practices. To this end, the company wants to increase SAP Energy Analysis software capabilitiesto fully harness the potential of the collected data. To achieve this goal, the proposed project aims to investigateadvanced data mining algorithms in order to extract, from the collected data, useful patterns of actions thatoffer insight to track customers’ consumption over different time stamps and develop an automatic tool toinstantly aggregate and analyze customers’ energy consumption profiles by any dimension or level ofgranularity as well as predicting future consumption trends which in turn permits efficient energy use.