Taking the Zambezi River Basin as a typical case, this paper studied the location problem of dam group. Based on the topographical and elevation maps of the Zambezi River Basin, we evaluated each region by five indicators (Water head difference elevation, Geological environment, Climatic environment, Population distribution and Biodiversity), and selected the 22 candidate dam sites. Meanwhile, the relative feasibility index of dam construction is calculated by the entropy weight-grey correlation analysis. On this basis, combined with the water management capacity of the dam, a set coverage model of dam selection is established, and the neighborhood adaptive particle swarm optimization algorithm (NAPSO) is used to solve the 12 most suitable dam sites. Comparing with the water management capacity of the original Kariba Dam, the new dams’ water storage and flood control capacity, hydroelectric power generation capacity, domestic water supply capacity and other water supply capacity have been increased by 235.92%, 250.62%, 189.66% and 223.61% respectively. Our study can provide some guidance for the site selection project of river dam group.