Building control system with intelligent load shedding
A method includes determining a probability that load shedding will be needed to achieve a target energy consumption. The method also includes, in response to the probability exceeding a threshold, generating a plurality of scores for a plurality of units of equipment indicating relative advantages of shedding the plurality of units of the equipment and implementing the load shedding by controlling plurality of unit of equipment. Controlling the plurality of units of equipment includes shedding a first unit of the plurality of units. The first unit is associated with an extremum score of the plurality of scores.
1. A method for controlling building equipment to achieve a target energy consumption, comprising:
determining a probability that load shedding will be needed to achieve the target energy consumption during a future time period;
in response to the probability exceeding a threshold, generating a plurality of scores for a plurality of units of equipment using one or more machine learning models trained on historical data from a historical time period, the plurality of scores indicating relative load shedding priorities associated with the plurality of units of the equipment during the future time period; and
implementing the load shedding by controlling the plurality of units of equipment in accordance with the plurality of scores during the future time period, wherein controlling the plurality of units of equipment comprises shedding a first unit of the plurality of units, the first unit associated with an extremum score of the plurality of scores.
2. The method of claim 1 , further comprising shedding a second unit of the plurality of units in response to determining that shedding the first unit is insufficient to achieve the target energy consumption.
3. The method of claim 1 , wherein shedding the first unit comprises turning off the first unit.
4. The method of claim 1 , wherein shedding the first unit comprises changing a setting for the first unit.
5. The method of claim 1 , wherein generating the plurality of scores comprises using a plurality of machine learning models associated with the plurality of units of the equipment.
6. The method of claim 5 , wherein:
the plurality of machine learning models comprise a first machine learning model for the first unit of the plurality of units;
using the plurality of machine learning models comprises providing, as inputs to the first machine learning model, a first value of a first building condition associated with the first unit and a second value of a second building condition associated with a second unit of the plurality of units; and
the first building condition is correlated with the second building condition.
7. The method of claim 1 , further comprising controlling the equipment to provide load shifting in response to predicting that the load shedding is insufficient to achieve the target energy consumption.
8. The method of claim 1 , wherein the target energy consumption comprises a plurality of energy amounts associated with a plurality of time steps in a time period.
9. The method of claim 1 , further comprising generating the target energy consumption based on a net energy goal and a forecast amount of energy generation.
10. The method of claim 9 , further comprising generating the net energy goal as an output of a control process configured to drive cumulative net energy over a time period to a value of zero or higher, wherein the net energy goal is for a subperiod of the time period.
11. One or more non-transitory computer-readable media storing program instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
determining a probability that load shedding will be needed during a future time period to achieve a target energy consumption during the future time period;
in response to the probability exceeding a threshold, generating a plurality of scores for a plurality of units of equipment using one or more machine learning models trained on historical data from a historical time period, the plurality of scores indicating relative load shedding priorities associated with the plurality of units of the equipment during the future time period; and
implementing the load shedding by controlling the plurality of units of equipment during the future time period in accordance with the plurality of scores, wherein controlling the plurality of units of equipment comprises shedding a first unit of the plurality of units, the first unit associated with an extremum score of the plurality of scores.
12. The one or more non-transitory computer-readable media of claim 11 , the operations further comprising shedding a second unit of the plurality of units in response to determining that shedding the first unit is insufficient to achieve the target energy consumption.
13. The one or more non-transitory computer-readable media of claim 11 , wherein shedding the first unit comprises turning off the first unit or changing a setting for the first unit.
14. The one or more non-transitory computer-readable media of claim 11 , wherein generating the plurality of scores comprises using a plurality of machine learning models associated with the plurality of units of the equipment.
15. The one or more non-transitory computer-readable media of claim 14 , wherein:
the plurality of machine learning models comprise a first machine learning model for the first unit of the plurality of units;
using the plurality of machine learning models comprises providing, as inputs to the first machine learning model, a first value of a first building condition associated with the first unit and a second value of a second building condition associated with a second unit of the plurality of units; and
the first building condition is correlated with the second building condition.
16. The one or more non-transitory computer-readable media of claim 11 , the operations further comprising controlling the equipment to provide load shifting in response to predicting that the load shedding is insufficient to achieve the target energy consumption.
17. The one or more non-transitory computer-readable media of claim 11 , wherein the target energy consumption comprises a plurality of energy amounts associated with a plurality of time steps in a time period.
18. The one or more non-transitory computer-readable media of claim 11 , the operations further comprising generating the target energy consumption based on a net energy goal and a forecast amount of energy generation.
19. The one or more non-transitory computer-readable media of claim 18 , the operations further comprising generating the net energy goal as an output of a control process configured to drive cumulative net energy over a time period to a value of zero or higher, wherein the net energy goal is for a subperiod of the time period.
20. A system, comprising:
a plurality of units of building equipment serving a facility;
a controller programmed to:
determine a probability that load shedding will be needed to achieve a target energy consumption for the facility;
in response to the probability exceeding a threshold, generate a plurality of scores for the plurality of units of equipment indicating relative advantages of load shedding priorities associated with the plurality of units of the equipment; and
implementing the load shedding by shedding a first unit of the plurality of units in response to the first unit being associated with a maximum score of the plurality of scores.