Predictive modeling and control system for building equipment with multi-device predictive model generation
A system includes a plurality of devices of building equipment, an additional device of building equipment, and a computing system. The computing system is configured to process data from the plurality of devices to extract common features of the plurality of devices, train a global model based on the common features, obtain additional data from the additional device, adapt the global model for the additional device based on the additional data to obtain an adapted model for the additional device, predict a status of the additional device using the adapted model, and affect an operation of the additional device based on the status.
1. A system, comprising:
a first device of building equipment;
a plurality of additional devices of building equipment; and
a computing system programmed to:
assess whether a data set from the first device is sufficient to train a fault prediction model for the first device;
in response to a determination that the data set from the first device is sufficient to train the fault prediction model for the first device, train the fault prediction model for the first device using the data set from the first device;
in response to a determination that the data set from the first device is insufficient to train the fault prediction model for the first device:
generate a ranking of the plurality of additional devices based on similarities between the first device and the plurality of additional devices;
augment the data set with supplemental data from one or more of the plurality of additional devices to obtain an augmented data set, the supplemental data obtained from each subsequent device of the plurality of additional devices in an order based on the ranking until the augmented data set is sufficient to train the fault prediction model; and
train the fault prediction model for the first device using the augmented data set; and
influence operations of the first device using the fault prediction model.
2. The system of claim 1 , wherein the computing system is programmed to augment the data set with the supplemental data from the one or more of the plurality of additional devices by increasing an amount of the supplemental data until the augmented data set is sufficient to train the fault prediction model.
3. The system of claim 1 , wherein the computing system is programmed to augment the data set with the supplemental data from the one or more of the plurality of additional devices by increasing a count of the one or more of the plurality of additional devices from which the supplemental data is used until the augmented data set is sufficient to train the fault prediction model.
4. The system of claim 1 , wherein the supplemental data is specific to a particular type of fault and the fault prediction model predicts the particular type of fault.
5. The system of claim 1 , wherein the fault prediction model is per device, wherein the computing system is further programmed to train a plurality of fault prediction models including the fault prediction model, each of the plurality of fault prediction models configured to predict a different type of fault.
6. The system of claim 1 , wherein the computing system is programmed to augment the data set with the supplemental data from the one or more of the plurality of additional devices by:
clustering the plurality of additional devices in a plurality of clusters based on characteristics of the plurality of additional devices;
associating the first device with a first cluster of the plurality of clusters; and
extracting the supplemental data from the first cluster.
7. The system of claim 1 , wherein influencing the operations of the first device using the fault prediction model comprises identifying, on an interface, that the first device is in need of maintenance and causing maintenance to be performed on the first device.
8. A method, comprising:
assessing whether a data set from a first device of building equipment is sufficient to train a fault prediction model for the first device;
in response to a determination that the data set from the first device is sufficient to train the fault prediction model for the first device, training the fault prediction model for the first device using the data set from the first device;
in response to a determination that the data set from the first device is insufficient to train the fault prediction model for the first device, augmenting the data set with supplemental data from one or more of a plurality of additional devices of building equipment to obtain an augmented data set and training the fault prediction model for the first device using the augmented data set;
ranking the plurality of additional devices based on similarities between the first device and the plurality of additional devices, wherein the supplemental data is obtained from each subsequent device of the plurality of additional devices in an order based on the ranking until the augmented data set is sufficient to train the fault prediction model; and
influencing operations of the first device using the fault prediction model.
9. The method of claim 8 , comprising augmenting the data set with the supplemental data from the one or more of the plurality of additional devices by increasing an amount of the supplemental data until the augmented data set is sufficient to train the fault prediction model.
10. The method of claim 8 , comprising augmenting the data set with the supplemental data from the one or more of the plurality of additional devices by increasing a count of the one or more of the plurality of additional devices from which the supplemental data is used until the augmented data set is sufficient to train the fault prediction model.
11. The method of claim 8 , wherein the supplemental data is specific to a particular type of fault and the fault prediction model predicts the particular type of fault.
12. The method of claim 8 , wherein the fault prediction model is per device, the method further comprising training a plurality of fault prediction models including the fault prediction model, each of the plurality of fault prediction models configured to predict a different type of fault.
13. The method of claim 8 , wherein augmenting the data set with the supplemental data from the one or more of the plurality of additional devices comprises:
clustering the plurality of additional devices in a plurality of clusters based on characteristics of the plurality of additional devices;
associating the first device with a first cluster of the plurality of clusters; and
extracting the supplemental data from the first cluster.
14. The method of claim 8 , wherein influencing the operations of the first device using the fault prediction model comprises identifying, on an interface, that the first device is in need of maintenance and causing maintenance to be performed on the first device.
15. One or more non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
assessing whether a data set from a first device of building equipment is sufficient to train a fault prediction model for the first device;
in response to a determination that the data set from the first device is sufficient to train the fault prediction model for the first device, training the fault prediction model for the first device using the data set from the first device;
in response to a determination that the data set from the first device is insufficient to train the fault prediction model for the first device;
ranking a plurality of additional devices based on similarities between the first device and the plurality of additional devices;
augmenting the data set with supplemental data from one or more of the plurality of additional devices of building equipment to obtain an augmented data set, the supplemental data obtained from each subsequent device of the plurality of additional devices in an order based on the ranking until the augmented data set is sufficient to train the fault prediction model; and
training the fault prediction model for the first device using the augmented data set; and
influencing operations of the first device using the fault prediction model.
16. The one or more non-transitory computer-readable media of claim 15 , the operations comprising augmenting the data set with the supplemental data from the one or more of the plurality of additional devices by increasing an amount of the supplemental data until the augmented data set is sufficient to train the fault prediction model.
17. The one or more non-transitory computer-readable media of claim 15 , the operations comprising augmenting the data set with the supplemental data from the one or more of the plurality of additional devices by increasing a count of the one or more of the plurality of additional devices from which the supplemental data is used until the augmented data set is sufficient to train the fault prediction model.
18. The one or more non-transitory computer-readable media of claim 15 , wherein the supplemental data is specific to a particular type of fault and the fault prediction model predicts the particular type of fault.
19. The one or more non-transitory computer-readable media of claim 15 , wherein the fault prediction model is per device, the operations further comprising training a plurality of fault prediction models including the fault prediction model, each of the plurality of fault prediction models configured to predict a different type of fault.
20. The one or more non-transitory computer-readable media of claim 15 , wherein augmenting the data set with the supplemental data from the one or more of the plurality of additional devices comprises:
clustering the plurality of additional devices in a plurality of clusters based on characteristics of the plurality of additional devices;
associating the first device with a first cluster of the plurality of clusters; and
extracting the supplemental data from the first cluster.