Joint operation of large-scale renewable energy sources (e.g., hydro and solar) has become a trend in modem power systems, and more operators of existing hydropower reservoir systems are adopting it. This study aimed to improve guidance for the daily generation scheduling of a large hydro-photovoltaic (PV) power plant. First, a robust optimization model that accounts for uncertain PV power generation was formulated, in which the delivered power output of the hybrid system and hydro unit status are robust decision variables. To deal with the complexity of the model solution, a three-layer nested framework was proposed to solve the generation scheduling problem in a hierarchical structure. In the outer layer, a direct search algorithm optimizes the delivered power output of the system, aiming to maximize energy production while satisfying specified load characteristics. In the middle and inner layers, a cuckoo search algorithm and dynamic programming technique optimize the hydro unit status and load dispatch strategies, respectively, with the objective of minimizing water consumption. Finally, a decision interval on delivered power output of the system was derived by relaxing the load characteristic constraint. Results for a case study using China's Longyangxia hydro-PV power plant indicated that the robust optimization model and the three-layer nested approach could provide effective power generation plans for the hybrid system within a reasonable time. Compared with actual operation, the power generation plans could increase energy production of the hybrid system by 1.9% while decreasing total online time of the hydro units by 9.7%. Therefore, the proposed method could improve guidance for the hydro-PV power plant's generation scheduling. In practice, conjunctive use of the power generation plan and the decision interval could also inform flexible decision-making for plant operations management.