Distribution networks, which are already complex large scale systems with diverse loads, are becoming more complicated in the smart grid era due to the increasing connections of distributed generators (e.g., solar and wind) and smart devices (e.g., electric vehicles and storage batteries). Although highly automated and integrated control systems have been proposed, systems with coverage from the zone substation to the low-voltage feeders practically do not exist. Based on the natural voltage decomposition of distribution networks and authors’ previous research on distribution system control, this paper proposes an efficient multi-objective hierarchical optimization strategy that can fully exploit the capabilities of both traditional and emerging devices, expand network load capacity and accommodate more renewable connections. Detailed simulations are performed on a real Australian distribution network over 24 hours to verify the feasibility, effectiveness and superiority of the proposed hierarchical optimization and some future work is also discussed.