The residential power is the lifeblood of the national economy, and its efficient management is of great significance. Precise prediction of the residential power demand has always been a most important concern of management mechanism which can be carried by a software defined network (SDN) platform. However, existing methods are heavily reliable on data with multiple features and high dimensions, failing to discovering sequential characteristics from simple and sparse data. In this paper, we develop a residual correction-based grey prediction model for residential power management under SDN. In detail, the residual function is used to correct the prediction value of the traditional gray model, so that prediction accuracy can be improved. Besides, a set of computational experiments are carried out on real-world business data to assess precision accuracy of the proposed model. It is concluded through experiments that the proposed model can better predict residential power demand.