With the increasing installation capacity of distributed photovoltaic（PV）systems in distribution networks, grid operators frequently issue curtailment instructions. Traditional group control strategies mainly adopt complete curtailment approaches when receiving these instructions, which reduces PV consumption rates and compromises user fairness. A distributionally robust optimization-based group control strategy for distributed PV systems that considers both demand response uncertainty and PV output uncertainty is proposed. Recognizing that demand response loads can partially fulfill curtailment requirements, models for reducible and interruptible loads are developed. Then, firstly a distributionally robust optimization model for PV group control that aims to minimize operational costs while employing a combination paradigm to characterize PV output uncertainty is constructed. This approach reduces the impact of prediction errors and output fluctuations on distribution network stability. The proposed model is solved using a column and constraint generation algorithm.The effectiveness of our proposed strategy is validated using data from a distribution network in Henan Province. Results demonstrate that the strategy can develop reasonable group control strategies when curtailment instructions are issued, thereby improving the consumption capacity of distributed PV systems.