Engineers in Prof. Ram Rajagopal’s laboratory have developed scalable algorithms to identify energy-usage patterns for efficiently targeting energy efficiency (EE) and demand response (DR) programs. This technology utilizes high resolution temporal energy consumption data from smart meters to build customer profiles and analyze the potential for successfully engaging them in EE or DR programs (see usage profile examples below). This data-driven approach takes costs and systems constraints into account to help develop programs tailored for specified lifestyles or to target existing programs to customers for whom they will be most effective. Applications: Energy demand management - data analytics of high resolution temporal energy consumption for: segmenting customers by lifestyle profile to tailor EE and DR programs for them target DR programs to selected customers that are most likely to adequately balance demand
1) Data-driven models rely on high resolution usage data (not demographic variables) to make predictions 2) Accounts for system constraints - optimization formulation includes both costs and network system constraints 3) Scalable - approximate algorithm designed to cope with computation issues coming from large data sets