Non-Intrusive Load Monitoring Applied to AC Railways

Non-intrusive load monitoring takes place in residential and industrial contexts to disaggregate identify loads connected a distribution grid. This work studies the applicability effectiveness for AC railways, considering highly dynamic behavior of rolling stock as an electric load, immersed varying moving loads. Both voltage–current diagrams harmonic spectra were considered identification extraction features relevant classification clustering. Principal components extracted, approaching problem using principal component analysis (PCA) partial least square regression (PLSR). Clustering methods then discussed, verifying separability performance railway context, checking by means balanced accuracy index. Based on more than one hundred measured spectra, PLSR has been confirmed with superior lower complexity. Independent verification based dispersion correlation used spot spectrum use clustering confirm outcome.