A Multi-Time-Scale Coordinated Dispatch Framework for Cascade Hydropower–PV Systems Based on Model Predictive Control

Abstract

The large-scale integration of photovoltaic generation increases the difficulty of maintaining secure and stable power system operation because solar output changes rapidly and is difficult to predict accurately. In cascade hydropower – photovoltaic systems, hydropower units can provide flexible regulation, but a dispatch scheme designed on only one time scale cannot simultaneously satisfy long-term water management requirements and short-term balancing needs. To overcome this limitation, a two-layer dispatch approach is developed in this study. At the upper layer, day-ahead scheduling is used to produce a baseline generation plan and feasible operating boundaries. At the lower layer, an intra-day rolling optimization model based on model predictive control updates hydropower decisions according to refreshed photovoltaic forecasts and current system states. A case study for the Xiaojinchuan River Basin is carried out to examine the effectiveness of the proposed method. Compared with conventional day-ahead dispatch, the proposed strategy decreases the maximum power tracking error from 13.00 MW to 7.07 MW and reduces the average deviation from 4.551 MW to 1.23 MW, while keeping reservoir operation within an acceptable range. The results indicate that combining day-ahead planning with intra-day correction can improve dispatch flexibility and operational robustness in cascade hydropower-PV complementary systems.