In January 2014, the east coast (PJM) electricity grid experienced an increase in the price of electricity from $31/MWh to $2,680/MWh in a matter of 10 minutes. This extreme price volatility has become the new norm in our electric grids. Building additional peak generation capacity is not environmentally or economically sustainable. Furthermore, the traditional view of Energy Efficiency does not address this need for Energy Flexibility. The solution lies with Demand Response (DR) from the customer side - curtailing demand during peak capacity for financial incentives. However, this is a very hard problem for commercial, industrial and institutional plants, the largest electricity consumers: they cannot model each building as they are all unique, cannot decide which of the 100,000's of control knobs to turn as it is too complex, must rely on rule-based curtailment approaches which are ad hoc, inefficient and do not provide any guarantees for energy reduction.
To this end, we have developed data-driven approaches to synthesize control-oriented models that bridge machine learning and controls for volatile energy markets. Our algorithms generate predictive models using Regression Trees, Ensemble Learning and Gaussian Processes for finite-time receding horizon control - where we can not only predict the state of the building but also generate control strategies using only historical weather, schedule, set-points and electricity consumption data. We call this approach Data Predictive Control (DPC). We have shown that, for a realistic building model, control strategies generated by DPC are remarkably similar to Model Predictive Control (MPC), while being scalable at the same time, unlike MPC.