Abstract: Peak load balancing ability is a key characteristic index to measure the performance of a hydropower station. To operate a regional power grid with new energy sources such as wind and solar power, a task of great significance is to conduct studies on the medium- and long-term peak load shaving capacity of the hydropower stations, so as to improve its new energy absorption and long-term stable operation. This paper presents a calculation method of daily peak load－a basis for determining the peak load shaving capacity on each scale－that is based on the historical typical daily load process corresponding to multiple time scales, and extracts the typical daily load proportion on each scale using the clustering by fast search and find of density peak (CFSFDP) algorithm. We develop a scheduling model of the corresponding calculation periods and time scales with the target of maximized peak power load, and examine the optimization scheduling on different scales－annual monthly scale, monthly ten-day scale, and intra-ten day scale. Application in a case study of the Longtan hydropower station with an annual regulation capacity shows that this method helps the decision-making for regional power grids to carry out medium- and long-term power balance, and it is an effective tool to schedule the peak power load of cascade hydropower stations on various time scales.

Key words: peaking capacity, peak shaving scheduling, typical daily load ratio, layer-by-layer optimization, Longtan hydropower station