Abstract: The integration of distributed new energy sources operating at different voltage levels continuously enhances the interaction capability between medium- and low-voltage distribution networks. How to optimize economic scheduling and power supply quality considering the time-scale differences among resources across multiple voltage levels has become an urgent issue requiring resolution in modern distribution networks. In view of this, this study proposes an optimization and control method for distribution networks based on multi-period coordination of heterogeneous photovoltaics and grid-side voltage regulation equipment. Firstly, a multi-scenario stochastic programming method is adopted to handle the source load uncertainty in the day-ahead stage. Subsequently, a day-ahead control model with the goal of minimizing the total operating cost and voltage deviation is constructed, thereby determining the adjustable resource for the medium-voltage distribution system control instructions and OLTC and CB switching schemes. At the same time, a trend unification method is adopted to incorporate previous instructions into the framework of intraday operation optimization, where they serve as a key constraint. The goal is to optimize the voltage level at the root node through precise daily regulation to improve the consumption rate of new energy. On this basis, a rolling optimization model for the intraday stage is established. Finally, the objective function of the intraday regulation model is designed and set to simultaneously optimize system economic cost and minimize node voltage deviation; the low-voltage resource regulation instruction is then obtained by solving this model. The calculation results indicate that the proposed scheme can achieve comprehensive optimization in both the operational efficiency and power supply quality in the distribution network.