Energy consumption modeling
Methods, devices, and systems for energy consumption modeling are described herein. One device includes a memory, and a processor configured to execute executable instructions stored in the memory to assign energy consumption data to an on state and an off state of a device, generate a model configured to predict a future state of the device and a duration of the future state based on a duration of a current state of the device, a duration of a previous state of the device, and operating conditions of the device, and predict a future energy consumption of the device using the assigned energy consumption data and the generated model.
1. A controller for an energy consumption modeling system, comprising:
a memory; and
a processor configured to execute executable instructions stored in the memory to:
generate a model to predict a future state of a device and a duration of the future state of the device by predicting a transition between a current state of the device and the future state of the device using a probability distribution for a duration of the current state of the device and an amount of time already spent by the device in the current state;
predict future energy consumption data of the device using energy consumption data of the device and the generated model;
detect a fault of the device using the energy consumption data and the generated model in response to the energy consumption data being out of step from the predicted future energy consumption data based on the duration of the current state of the device being longer or shorter than the duration of the predicted future state of the device; and
generate a notification in response to the fault of the device being detected.
2. The controller of claim 1 , wherein the processor is configured to execute the instructions to detect the fault by comparing the energy consumption data to the predicted future energy consumption data.
3. The controller of claim 1 , wherein the processor is configured to execute the instructions to detect the fault of the device in response to the energy consumption data being different from the predicted future energy consumption data.
4. The controller of claim 3 , wherein the energy consumption data being different from the predicted future energy consumption data indicates degradation of the device.
5. The controller of claim 1 , wherein the processor is configured to execute the instructions to determine no fault of the device exists in response to the energy consumption data being the same as the predicted future energy consumption data.
6. The controller of claim 1 , wherein the controller is at least one of:
a part of a residential thermostat;
a thermostatically controlled device; and
a part of an energy meter.
7. The controller of claim 1 , wherein the energy consumption data includes:
energy consumption data of an on state of the device; and
energy consumption data of an off state of the device.
8. The controller of claim 1 , wherein the probability distribution is a function of a duration of a previous state of the device and operating conditions of the device.
9. A non-transitory computer readable medium having computer readable instructions stored thereon that are executable by a processor to:
generate a model to predict a future state of a device and a duration of the future state of the device by predicting a transition between a current state of the device and the future state of the device using a probability distribution for a duration of the current state of the device and an amount of time already spent by the device in the current state;
predict future energy consumption data of the device using energy consumption data of the device and the generated model;
compare the energy consumption data to the predicted future energy consumption data;
determine a fault of the device in response to the energy consumption data being out of step from the predicted future energy consumption data based on the duration of the current state of the device being longer or shorter than the duration of the predicted future state of the device; and
generate a notification in response to the fault of the device being detected.
10. The computer readable medium of claim 9 , wherein the computer readable instructions are executable by the processor to determine the fault in response to at least one of:
the energy consumption data being higher than the predicted future energy consumption data; and
the energy consumption data being lower than the predicted future energy consumption data.
11. The computer readable medium of claim 9 , wherein the computer readable instructions are executable by the processor to predict the future state of the device and the duration of the future state of the device based on:
the duration of the current state of the device;
a duration of a plurality of previous states of the device; and
operating conditions of the device.
12. The computer readable medium of claim 9 , wherein:
the energy consumption data is associated with the current state of the device; and
the predicted future energy consumption data is associated with the future state of the device.
13. A method for operating an energy consumption modeling system, comprising:
generating, by a controller, a model to predict a future state of a device and a duration of the future state of the device by predicting a transition between a current state of the device and a future state of the device using a probability distribution for a duration of the current state of the device and an amount of time already spent by the device in the current state;
predicting, by the controller, future energy consumption data of the device using energy consumption data of the device and the generated model;
comparing, by the controller, the energy consumption data to the predicted future energy consumption data;
detecting, by the controller based on the comparison:
a fault of the device in response to the energy consumption data being out of step from the predicted future energy consumption data based on the duration of the current state of the device being longer or shorter than the duration of the predicted future state of the device; and
that no fault of the device exists in response to the energy consumption data being the same as the predicted future energy consumption data; and
generating, by the controller, a notification in response to detecting the fault of the device.
14. The method of claim 13 , wherein the method includes:
determining, by the controller, outlier data included in the energy consumption data; and
generating, by the controller, the probability distribution while ignoring the outlier data.
15. The method of claim 13 , wherein the method includes receiving the energy consumption data from an energy meter.
16. The method of claim 13 , wherein the method includes updating the model based on additional energy consumption data received from an energy meter.
17. The method of claim 13 , wherein the method includes generating a demand response capacity based on the predicted future energy consumption data.
18. The method of claim 13 , wherein the method includes generating the probability distribution for the duration of the current state of the device as a function of a duration of a previous state of the device and operating conditions of the device.