Non-Intrusive Load Monitoring Design Using K-Means Clustering Extreme Learning Machine
Dimas Fajar Uman Putra, Ontoseno Penangsang, Adi Soeprijanto · International Review on Modelling and Simulations (IREMOS) · 2018
This paper purposes a new algorithm to determine the on-off status for Non-Intrusive Load Monitoring (NILM) by using K-means Clustering Extreme Learning Machine (ELM). K-means Clustering is used in this research to cluster the current group from the load. By clustering this current, the input for ELM can be grouped by the current magnitude from wavelet process. This clustering process is done to prevent a singularity matrix that usually happens when the inversed H matrix in ELM process is done. The simulation result shows that the proposed method has a good result as Neural Network when identifying the loads status but it has a faster training and testing time. Singularity occurs when ELM is used without applying the clustering method before.