Abrupt Change Detection of Home Electrical Appliance Signals Using Mathematical Morphology Gradient

This paper presents an Abrupt Change Detection (ACD) algorithm based on Mathematical Morphology Gradient (MMG) for Non-Intrusive Load Monitoring (NILM). In the present research of MMG, it is focused on filtering the original signals rather than locating change points. Therefore, this paper considers this approach to detect abrupt changes. For the purpose of evaluating the performance of different algorithms, the True Positive Rate (TPR) and the F-score are employed in this paper. The simulation studies have been carried out on low-frequency signals and high-frequency signals from the Reference Energy Disaggregation Data Set (REDD) respectively. The MMG algorithm, which has the advantages of high accuracy and fast calculation speed, is superior for both two types of signals. In addition, it can provide not only the information of the position of a change but also the direction of a change. The results demonstrate that the MMG algorithm shows better performance than the CUmulative SUM (CUSUM), the Bayesian Information Criterion (BIC) and the Generalized Likelihood Ratio (GLR) algorithms.