In order to solve the problem of emergency material allocation in flood disaster area, this paper constructs the research framework combined with “Internet +” and proposes a mixed optimization method. Firstly, real-time demand point estimation is provided for the material needs of the disaster area. Then, unsupervised clustering is applied to the correlation properties of the flood disaster area to obtain the division of the region. Subsequently, the urgency priority of the material distribution area is determined. According to the weight, the remaining relief supplies of emergency inventory are allocated to the disaster areas, and the disaster area material demand is updated immediately, until the demand in the disaster area reaches the preset level of satisfaction. Finally, the results show that the hybrid optimization method can fully optimize the system in terms of demand satisfaction ratio and time allocation cost, and make effective use of the golden rescue period to complete the humanitarian rescue.