Unsupervised multivariate anomaly detection through variational auto-encoding in hvac machinery

Abstract

A system for HVAC anomaly detection includes a sensor configured to capture temperature, pressure data, flow data, and/or current draw, a processor, and a memory. The memory includes instructions stored thereon, which, when executed cause the system to access the captured sensor data, provide the sensor data as an input to a machine learning network, and predicting one or more anomalies using the machine learning network,

What is claimed is: 
     
         1 . A system for heating, ventilation, and air conditioning (HVAC) anomaly detection comprising:
 a sensor configured to capture temperature, pressure data, flow data, and/or current draw;   a processor; and   a memory, including instructions stored thereon, which, when executed by the processor, cause the system to:
 access the captured sensor data; 
 provide the sensor data as an input to a machine learning network; and 
 predict one or more anomalies using the machine learning network. 
   
     
     
         2 . The system of  claim 1 , wherein the machine learning network includes variational auto-encoding, a transformer, other RNN based models (RNN, LSTM), Decision Trees (Isolation Forest), a Support Vector Machine (SVM), sequence to sequence model, K-means clustering, and/or an ensemble model. 
     
     
         3 . The system of  claim 1 , wherein the instructions, when executed by the processor, further cause the system to generate a report indicating the predicted one or more anomalies. 
     
     
         4 . The system of  claim 1 , wherein the instructions, when executed by the processor, further cause the system to transmit an indication to a user device about the predicted one or more anomalies. 
     
     
         5 . The system of  claim 1 , wherein the instructions, when executed by the processor, further cause the system to disable one or more components of the HVAC system based on the predicted one or more anomalies. 
     
     
         6 . A processor-implemented method for HVAC anomaly detection comprising:
 accessing captured sensor data from a sensor configured to capture temperature, pressure data, flow data, and/or current draw;   providing the sensor data as an input to a machine learning network; and   predicting one or more anomalies using the machine learning network.   
     
     
         7 . The method of  claim 6 , wherein the machine learning network includes variational auto-encoding, a transformer, other RNN based models (RNN, LSTM), Decision Trees (Isolation Forest), a Support Vector Machine (SVM), sequence-to-sequence model, K-means clustering, and/or an ensemble model. 
     
     
         8 . The method of  claim 6 , further comprising generating a report indicating the predicted one or more anomalies. 
     
     
         9 . The method of  claim 6 , further comprising transmitting an indication to a user device about the predicted one or more anomalies. 
     
     
         10 . The method of  claim 6 , further comprising disabling one or more components of the HVAC system based on the predicted one or more anomalies. 
     
     
         11 . A non-transitory computer-readable medium, storing instructions, which when executed by a processor, cause performance of a processor-implemented method for HVAC anomaly detection, the method comprising:
 accessing captured sensor data from a sensor configured to capture temperature, pressure data, flow data, and/or current draw;   providing the sensor data as an input to a machine learning network; and   predicting one or more anomalies using the machine learning network.   
     
     
         12 . The non-transitory computer-readable medium of  claim 11 , wherein the machine learning network includes variational auto-encoding, a transformer, other RNN based models (RNN, LSTM), Decision Trees (Isolation Forest), a Support Vector Machine (SVM), sequence to sequence model, K-means clustering, and/or an ensemble model. 
     
     
         13 . The non-transitory computer-readable medium of  claim 11 , wherein the instructions, when executed by the processor, further cause the performance of a processor-implemented method for HVAC anomaly detection, the method further comprising generating a report indicating the predicted one or more anomalies. 
     
     
         14 . The method of  claim 11 , wherein the instructions, when executed by the processor, further cause the performance of a processor-implemented method for HVAC anomaly detection, the method further comprises transmitting an indication to a user device about the predicted one or more anomalies. 
     
     
         15 . The method of  claim 11 , wherein the instructions, when executed by the processor, further cause the performance of a processor-implemented method for HVAC anomaly detection, the method further comprising disabling one or more components of an HVAC system based on the predicted one or more anomalies.