| Popis: |
An anomaly is an observation that highly deviates from other observations. These anomalies create abnormal time series that are different from a collection of other time series. Data collected using wireless sensors, such as temperature and humidity, can provide insight into a building's heating, ventilation, and air conditioning (HVAC) system. When using sensors that are properly designed, installed and calibrated the indoor environmental quality of a building can be measured. This will allow anomalies to be identified through sensor measurements, which point to areas with poor design or insufficient maintenance. Identifying these can improve both thermal comfort and energy efficiency and improve building performance. In this study, we applied the Dynamic Time Warping (DTW) based anomaly detection method to identify anomalies and introduced a scoring method to identify abnormal sensors. The number of anomalies, vertical distance to an anomaly point, and DTW distance was considered to identify abnormal sensors. Then we used high-resolution temperature measurements from two school buildings using wireless sensors to evaluate the performance of the developed scoring method. Based on the results, visually we could observe that the method accurately detects the abnormal sensors. |
