To Switch or Not to Switch? Rate Plan Selection in the Electricity Retail Market

Retail electricity market brings a wide range of tariff choices to customers. However, the residential participants may have difficulty in choosing the suitable one that can minimize their costs due to the lack of real time measurements for their usage and expenditure. To this end, we introduce a data-driven recommendation method to match rate plans’ characteristics with users’ consumption habits. We first divide the mainstream rate plans into three common types — fix-rate plan, variable-rate plan and time-of-use plan. Then we cluster the consumers based on the K-means algorithm, calculate the potential costs of selecting different plans for each cluster, and make the initial recommendation. To better adapt the price fluctuations of variable-rate plans, we adopt a simple threshold policy to conduct demand response for the consumers. Simulation studies verify the effectiveness of our proposed recommendation method.