Electricity Tariff Design Based on Clustered Load Profiles Classified by Exploiting Questionnaires

摘要

Since the deregulation of electricity sector, load profiles of electricity customers are even more important because the knowledge gathered from load profiling can be used for various additional purposes. Special use is, for example, tariff design under market environment. In this paper new framework is developed for tariff design. This framework uses k-means algorithm for clustering samples of measured load profiles and makes classification by using decision trees for data from questionnaires which were filled by household customers. This way it is possible to maximize combined use of historic load measurements and simple information about various household customers.