Building management system with machine learning for detecting anomalies in vibration data sets
A building management system including building equipment operable to affect a variable state or condition of a building. The building management system includes a controller including a processing circuit. The processing circuit is configured to obtain a vibration data set related to vibrations of the building equipment. The processing circuit is configured to analyze the vibration data set by one or more machine learning models to generate a set of probabilities. The set of probabilities is related to a probability that the vibration data set is abnormal. The processing circuit is configured to identify the vibration data set as normal or abnormal based on the set of probabilities. The processing circuit is configured to initiate a corrective action responsive to identifying the vibration data set as abnormal.
1. A building management system comprising:
building equipment operable to affect a variable state or condition of a building; a controller comprising a processing circuit configured to:
obtain a vibration data set related to vibrations of the building equipment;
analyze the vibration data set by one or more machine learning models to generate a set of probabilities, the set of probabilities related to a probability that the vibration data set is abnormal;
identify the vibration data set as normal or abnormal based on the set of probabilities; and initiate a corrective action responsive to identifying the vibration data set as abnormal;
wherein identifying the vibration data set as normal or abnormal comprises at least one of: determining whether a maximum probability of the set of probabilities exceeds a threshold probability; or providing the set of probabilities to a model configured to label the vibration data set as normal or abnormal based on the set of probabilities.
2. The building management system of claim 1 , wherein the processing circuit is configured to:
perform one or more fast Fourier transforms on the vibration data set to generate a fast Fourier transform (FFT) spectra;
wherein analyzing the vibration data set comprises providing the FFT spectra to the one or more machine learning models to generate the set of probabilities.
3. The building management system of claim 2 , wherein the one or more machine learning models comprise one or more convolutional neural networks configured to analyze one or more frequency ranges of the FFT spectra for abnormalities.
4. The building management system of claim 1 , wherein the processing circuit is configured to:
provide the vibration data set to an analyst responsive to identifying the vibration data set as abnormal; and
obtain analyst feedback from the analyst indicating whether the analyst believes the vibration data set is abnormal;
wherein the corrective action is initiated based on the analyst indicating the vibration data set is abnormal.
5. The building management system of claim 1 , wherein the processing circuit is configured to:
generate a report responsive to identifying the vibration data set as normal or abnormal; and
provide the report to a user device.
6. The building management system of claim 1 , wherein the corrective action comprises at least one of:
scheduling maintenance or replacement for the building equipment;
generating an abnormal report describing abnormality in the vibration data set; or
disabling the building equipment.
7. A method for analyzing vibration data sets of equipment, the method comprising:
obtaining a vibration data set related to vibrations of the equipment;
analyzing the vibration data set by one or more machine learning models to generate a set of probabilities, the set of probabilities related to a probability that the vibration data set is abnormal;
identifying the vibration data set as normal or abnormal based on the set of probabilities; and initiating a corrective action responsive to identifying the vibration data set as abnormal;
wherein identifying the vibration data set as normal or abnormal comprises at least one of: determining whether a maximum probability of the set of probabilities exceeds a threshold probability; or providing the set of probabilities to a model configured to label the vibration data set as normal or abnormal based on the set of probabilities.
8. The method of claim 7 , further comprising:
performing one or more fast Fourier transforms on the vibration data set to generate a fast Fourier transform (FFT) spectra;
wherein analyzing the vibration data set comprises providing the FFT spectra to the one or more machine learning models to generate the set of probabilities.
9. The method of claim 8 , wherein the one or more machine learning models comprise one or more convolutional neural networks configured to analyze one or more frequency ranges of the FFT spectra for abnormalities.
10. The method of claim 7 , further comprising:
providing the vibration data set to an analyst responsive to identifying the vibration data set as abnormal; and
obtaining analyst feedback from the analyst indicating whether the analyst believes the vibration data set is abnormal;
wherein the corrective action is initiated based on the analyst indicating the vibration data set is abnormal.
11. The method of claim 7 , further comprising:
generating a report responsive to identifying the vibration data set as normal or abnormal; and
providing the report to a user device.
12. The method of claim 7 , wherein the corrective action comprises at least one of:
scheduling maintenance or replacement for the equipment;
generating an abnormal report describing abnormality in the vibration data set; or
disabling the equipment.
13. A controller for analyzing vibration data sets of equipment, the controller comprising:
one or more processors; and one or more non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
obtaining a vibration data set related to vibrations of the equipment;
analyzing the vibration data set by one or more machine learning models to generate a set of probabilities, the set of probabilities related to a probability that the vibration data set is abnormal;
identifying the vibration data set as normal or abnormal based on the set of probabilities; and initiating a corrective action responsive to identifying the vibration data set as abnormal;
wherein identifying the vibration data set as normal or abnormal comprises at least one of: determining whether a maximum probability of the set of probabilities exceeds a threshold probability; or providing the set of probabilities to a model configured to label the vibration data set as normal or abnormal based on the set of probabilities.
14. The controller of claim 13 , the operations further comprising:
performing one or more fast Fourier transforms on the vibration data set to generate a fast Fourier transform (FFT) spectra;
wherein analyzing the vibration data set comprises providing the FFT spectra to the one or more machine learning models to generate the set of probabilities.
15. The controller of claim 14 , wherein the one or more machine learning models comprise one or more convolutional neural networks configured to analyze one or more frequency ranges of the FFT spectra for abnormalities.
16. The controller of claim 13 , the operations further comprising:
providing the vibration data set to an analyst responsive to identifying the vibration data set as abnormal; and
obtaining analyst feedback from the analyst indicating whether the analyst believes the vibration data set is abnormal;
wherein the corrective action is initiated based on the analyst indicating the vibration data set is abnormal.
17. The controller of claim 13 , the operations further comprising:
generating a report responsive to identifying the vibration data set as normal or abnormal; and
providing the report to a user device.
18. The controller of claim 13 , wherein the corrective action comprises at least one of:
scheduling maintenance or replacement for the equipment;
generating an abnormal report describing abnormality in the vibration data set; or
disabling the equipment.