System and methods for power system forecasting using deep neural networks

1. A method of managing energy, by use of processing logic that comprises a load processor as a cloud service, the method comprising:

receiving, by the load processor, power load information from a data collection system located at a building and using a cloud analysis layer that employs machine-learning and artificial intelligence for optimization control;

analyzing, by the load processor, the received power load information to disaggregate load waveform signals and identify device-based power loads by use of a neural network to perform historical device demand and performance analysis to generate device-based demand forecasting;

generating, by the load processor, demand forecasts for the building to mitigate peak demand based on analysis of a power draw signal and the generated device-based demand forecasting;

determining, by the load processor, whether the generated demand forecast for the building is to peak in a near future, based on threshold values of at least one of generated device-based demand forecasting, power price or cost information, and user behavior analysis; and

detecting, by the load processor, a malfunction in individual devices connected to a central power system of the building based on the received power load information, the generated demand forecasts for the building, and the generated device-based demand forecasting.

2. The method of claim 1 , wherein the receiving power load information comprises receiving power load information based on an electric draw of individual devices connected to the central power system of the building without requiring any changes to the individual devices.

3. The method of claim 1 , wherein the generating demand forecasts for the building further comprises building one or more predictive models of how a building operates and power performance of each device in the building.

4. The method of claim 3 , wherein the one or more predictive models of how the building operates comprises:

analyzing, by the load processor, historical power performance for each device;

correlating, by the load processor, historical power performance to weather information associated with the building.

5. The method of claim 3 , wherein the one or more predictive models of how the building operates comprises:

analyzing, by the load processor, historical power performance for each device;

correlating, by the load processor, historical power performance to geographic information associated with the building.

6. The method of claim 3 , wherein the one or more predictive models of how the building operates comprises:

analyzing, by the load processor, historical power performance for each device;

correlating, by the load processor, historical power performance to operation information associated with operation activities performed within the building.

7. The method of claim 1 , further comprising generating, by the load processor, an adjustment schedule determining device-based set-points to reduce power load of at least one device associated with the determined device-based set-points during a time period based on changes in one or more of weather data and energy cost changes.

8. The method of claim 1 , further comprising detecting, by the load processor, anomalies in building's energy performance based on the received power load information, the generated demand forecasts for the building, and the generated device-based demand forecasting.

9. The method of claim 8 , further comprising scheduling a response strategy based on the received power load information and the generated device-based demand forecasting to drive the received power load information back to the generated demand forecast for the building.

10. A computing device for managing energy, the computing device comprising:

a load processor, and

a data collection system configured to provide power load information, the data collection system located at a building and using a cloud analysis layer that employs machine-learning and artificial intelligence for optimization control;

wherein the load processor is configured to:

analyze received power load information to disaggregate load waveform signals and identify device-based power loads by use of a neural network to perform historical device demand and performance analysis to generate device-based demand forecasting;

generate demand forecasts for the building to mitigate peak demand based on analysis of a power draw signal and the generated device-based demand forecasting;

determine whether the generated demand forecast for the building is to peak in a near future, based on threshold values of at least one of generated device-based demand forecasting, power price or cost information, and user behavior analysis; and

detect a malfunction in individual devices connected to a central power system of the building based on the received power load information, the generated demand forecasts for the building, and the generated device-based demand forecasting.

11. The computing device of claim 10 , wherein the data collection system is configured to power load information based on an electric draw of individual devices connected to the central power system of the building without requiring any changes to the individual devices.

12. The computing device of claim 10 , wherein the load processor is configured to generate the demand forecasts for the building by building one or more predictive models of how a building operates and the power performance of each device in the building.

13. The computing device of claim 12 , wherein the load processor is configured to build the one or more predictive models of how the building operates by:

analyzing historical power performance for each device;

correlating historical power performance to weather information associated with the building.

14. The computing device of claim 12 , wherein the load processor is configured to build the one or more predictive models of how the building operates by:

analyzing historical power performance for each device;

correlating historical power performance to geographic information associated with the building.

15. The computing device of claim 12 , wherein the load processor is configured to build the one or more predictive models of how the building operates by:

analyzing historical power performance for each device;

correlating historical power performance to operation information associated with operation activities performed within the building.

16. The computing device of claim 10 , wherein the load processor generates an adjustment schedule determining device-based set-points to reduce power load of a least one device associated with the determined device-based set-points during a time period based on changes in one or more of weather data and energy cost changes.

17. The computing device of claim 10 , wherein the load processor detects anomalies in building's energy performance based on the received power load information, the generated demand forecasts for the building, and generated device-based demand forecasting.

18. The computing device of claim 17 , wherein the load processor schedules a response strategy based on the received power load information and the generated device-based demand forecasting to drive the received power load information back to the generated demand forecast for the building.

19. A non-transitory computer readable medium having stored therein a program for making a computer execute a method of managing energy, the method comprising:

receiving power load information from a data collection system located at a building and using a cloud analysis layer that employs machine-learning and artificial intelligence for optimization control;

analyzing the received power load information to disaggregate load waveform signals and identify device-based power loads by use of a neural network to perform historical device demand and performance analysis to generate device-based demand forecasting;

generating demand forecasts for the building to mitigate peak demand based on analysis of a power draw signal and the generated device-based demand forecasting;

determining whether the generated demand forecast for the building is to peak in a near future, based on threshold values of at least one of generated device-based demand forecasting, power price or cost information, and user behavior analysis; and

detecting a malfunction in individual devices connected to a central power system of the building based on the received power load information, the generated demand forecasts for the building, and the generated device-based demand forecasting.

20. The non-transitory computer readable medium of claim 19 , wherein the generating demand forecasts for the building further comprises building one or more predictive models of how a building operates and the power performance of each device in the building; and

wherein the one or more predictive models of how the building operates comprises:

analyzing historical power performance for each device;

correlating historical power performance to:

weather information associated with the building,

geographic information associated with the building, and

operation information associated with operation activities performed within the building.