Development of a Predictive Optimal Controller for Thermal Energy Storage Systems

Predictive Optimal Control of Active and Passive Building Thermal Storage Inventory

“…In the presence of complex rate structures, the optimal controller was found to have a vast performance benefit (saving 40%) over conventional controls while requiring only simple predictors. 11 Interestingly, these general findings did not change when there was considerable model mismatch, i.e., uncertainty with respect to the behavior of the actual cooling plant. Recent investigations regarding forecasting uncertainty determined that this controller is robust and does not require high accuracy in predicting loads and utility rates.…”

“…24 Optimal building control proved most effective in dry climates with large diurnal temperature swings, in the presence of utility rates strongly encouraging load-shifting, and when cool storage systems allow more effective load-shifting than building pre-cooling alone. These results in conjunction with personal experience in optimal control applied to building systems, [10][11][12][13][25][26][27] encouraged me to develop this idea into a predictive supervisory controller suitable for implementation in commercial buildings with dynamic utility rates.…”

“…Recent investigations regarding forecasting uncertainty determined that this controller is robust and does not require high accuracy in predicting loads and utility rates. 13 In summary, dynamic rate structures including RTP could make thermal energy storage systems more economically attractive while TES systems may increase the applicability of dynamic rates. We employ both the active and passive thermal storage capacity of a commercial building under optimal supervisory control using dynamic utility rates.…”

“…Presence of Uncertainty -During a subsequent investigation phase, we determined the extent of the merits of optimal control that can be retained when the optimal controller is subject to uncertainty in the external variables driving the physical process, such as future weather variables and cooling loads. 11 The predictive optimal strategy is based on closed-loop optimization, i.e., an optimal storage charging and discharging strategy is developed at every time step over a marching planning horizon utilizing updated forecasts. The prediction of climate conditions, utility rates and cooling loads is updated at the beginning of each time step over the optimization period, and a new optimal strategy is computed.…”

A Wireless Platform for Energy Efficient Building Control Retrofits

“…Using the insights gained after analyzing the 2011 demonstration experiment data summarized above, the MPC algorithm was revised and improved, including refined dynamical models for the mixing, hot deck and cold deck temperatures with the goal of improving the poor peak power usage performance for the MPC mode. Another set of demonstration experiments were proposed to CERL staff and, following their approval, were conducted during the week of Feb. [13][14][15][16][17]2012. The results were processed following the same procedure as explained for the data generated in Despite best efforts to ensure consistency in indoor and ambient conditions, there is variability in the energy savings estimated.…”