Abstract: The high percentage of photovoltaic(PV) access to the distribution network brings reverse heavy overload risk to the distribution network, which affects the safe and stable operation of the distribution network. In order to accurately evaluate the reverse heavy overload risk of each voltage level in the distribution network with high proportion of PV access and to analyze the sources of reverse heavy overload risk, firstly, a "station-feeder-substation" reverse heavy overload risk assessment index system is established for the distribution network with high proportion of PV access. Secondly, a combined weighting method based on entropy weighting and hierarchical analysis is used to assign weights to each index. Finally, the effectiveness of the proposed method is verified by simulating a distribution network case based on an actual network topology improvement in Hunan. In addition, the proposed method can be adopted to effectively analyze the reverse heavy overload risk situation of distribution network under different distributed PV access locations, which has reference significance for power system risk assessment and management.