Clustering commercial and industrial load patterns for long-term energy planning

In future smart energy systems, consumers are expected to change their load patterns as they become a significant source of flexibility. To ensure reliable profile forecasts for long-term planning, conventional classification approaches will not hold and more advanced solutions required. this article, we propose an automatic, data-driven clustering methodology that accounts heterogeneity in electricity consumers’ profiles using unsupervised learning. We consider hourly measurements from 9412 smart-meters the commercial industrial sector Denmark. A wavelet transform is applied min-max scaled data, extracted coefficients used input K-means algorithm. Through cluster validation, eight clearly distinct identified compared industry constituents. Finally, flexibility potential traced each cluster.