Automated demand response energy management system

摘要

The power flexibility of energy loads are maximized using a value function for each load and outputting optimal control parameters per load. These loads are aggregated into a virtual load by maximizing a global value function that includes the value function for each individual load. The solution yields a dispatch function providing: a percentage of energy to be assigned to each individual load, a possible time-varying power level within a time interval for each load, and control parameters and values. An economic term of the global value function represents the value of the power flexibility to different energy players. A user interface includes for each time interval upper and lower bounds representing respectively the maximum power that may be reduced to the virtual load and the maximum power that may be consumed by the virtual load. An energy trader modifies an energy level in a time interval relative to the reference curve for the virtual load. Automatically, energy compensation for the other intervals and recalculation of upper and lower boundaries occurs and is displayed. The energy schedule for the virtual load is distributed to the actual loads using the dispatch function.

主权项

1. A method of rescheduling energy for a virtual energy load, said method comprising:
computing an energy schedule for said virtual load, said energy schedule indicating an energy level to be consumed by said virtual load for each of a plurality of time intervals within a time period, said virtual load representing a plurality of actual energy loads; distributing an individual control schedule to each of said actual loads such that said actual loads in total consume energy according to said energy schedule; receiving power logging data from at least one of said actual loads during a current one of said time intervals, said power logging data indicating an amount of energy consumed by said at least one of said actual loads in said current time interval; estimating a total energy consumption of said virtual load in said current time interval; comparing said total energy consumption with said energy level from said energy schedule for said current time interval; recomputing said individual control schedules using an updated value for said energy level for said current time interval; distributing said recomputed individual control schedules among said actual loads, whereby each actual load receives its recomputed individual control schedule and whereby said actual loads in total consume energy according to said energy schedule; recalibrating a load model for each of said actual loads using logged state variables for each of said actual loads received during said current time interval, wherein said recomputing includes using an inverse model for each of said recalibrated load models in order to output said recomputed individual control schedules; and controlling said actual loads during said time period using said recomputed individual control schedules.