Water Meter Replacement Recommendation for Municipal Water Distribution Networks using Ensemble Outlier Detection Methods

Due to their structure and usage conditions, water meters face degradation, breaking, freezing, and leakage problems. There are various studies intended to determine the appropriate time to replace the degraded water meters. Earlier studies have used several features such as meteorological parameters, usage conditions, water network pressure, and structure of meters in order to detect the failed water meters. This article proposes a recommendation framework that uses the registered water consumption values as the input data and provides the meter replacement recommendations. This framework takes the time series of registered consumption values and pre-processes them in two rounds in order to extract the effective features. Then the multiple un-/semi-supervised outlier detection methods are applied to the processed data and assigns outlier/normal labels to them. At the final stage, a hypergraph-based ensemble method receives the labels and combines them to discover the suitable label. Due to the unavailability of the ground truth labeled data for meter replacement, we compare our method with respect to its FPR and two internal metrics: Dunn index and Davies-Bouldin index. The results of our comparative experiments show that the proposed framework detects more compact clusters with a smaller variance.