US 11,870,256 B2
AI-based software service to shave electricity consumption peak
Yi Wang, Singapore (SG); Elissa Yanting Lim, Singapore (SG); Rully Adrian Santosa, Singapore (SG); Soon Siang Tan, Singapore (SG); Yiqun Hu, Singapore (SG); and Yanhui Chen, Singapore (SG)
Assigned to SPACE PTE. LTD., Singapore (SG)
Filed by SPACE PTE. LTD., Singapore (SG)
Filed on Sep. 28, 2021, as Appl. No. 17/487,096.
Claims priority of application No. 10202009796Y (SG), filed on Oct. 1, 2020.
Prior Publication US 2022/0109298 A1, Apr. 7, 2022
Int. Cl. H02J 3/00 (2006.01); H02J 3/24 (2006.01); H02J 3/28 (2006.01)
CPC H02J 3/003 (2020.01) [H02J 3/004 (2020.01); H02J 3/24 (2013.01); H02J 3/28 (2013.01)] 16 Claims
OG exemplary drawing
 
1. A method for optimizing usage of electricity from a grid and energy storage comprising:
forecasting a predetermined number of hours of load at a predetermined interval to determine a forecasted load profile;
subtracting a peak load from the forecasted load profile by a maximum amount of electricity that can be charged into or discharged from the energy storage to determine a subtracted peak;
smoothening peaks within a 30-minute period that are more than the subtracted peak to flatten out a consumption profile around a forecasted peak so that a flattened period covers the peak load to obtain a final forecast result;
determining optimized battery setpoint actions based on the final forecast result and energy storage operating constraints comprising:
applying the energy storage operating constraints; and
minimizing an objective function that can be expressed in the following expression:

OG Complex Work Unit Math
where
t∈T refers to a set of time steps in optimization horizon, Pgrid,t refers to electricity drawn from the power grid at time step t, kWh, Pbat,t refers to electricity charged into/discharged from battery at time step t, kWh, Pgrid,t∈tpeak refers to peak electricity drawn from the power grid during optimization horizon, kWh, μt refers to electricity price at time step t, Currency unit/kWh, λ refers to battery lifespan coefficient, Currency unit/kWh, wherein the steps of forecasting, subtracting, smoothening and determining are performed using at least one processor; and
executing, by an energy management system, the optimized battery setpoint actions to control usage of the grid and energy storage.