Dynamic deployment of UAVs in edge data acquisition scenario is important to make use of UAVs’ resources. First, within the edge data acquisition system, the Optimization Model of UAV Cluster Dynamic Deployment (OMUCDP) is established. OMUCDP takes into account the temporality of tasks, dynamically optimizing the deployment and movement decision of UAVs, which is more in line with the realistic characteristic. Then, the deployment and movement decision of UAVs is optimized separately using Differential Evolution and Hungarian algorithm. Finally, fixed and variable time slice mechanism and three different optimization schemes are proposed to cope with dynamic deployment. Experimental results indicate that the energy consumption with variable time slice can achieve the best case of fixed time slice. And Scheme 2 is more practical and achieves the tradeoff between IoT nodes’ transmission energy consumption and UAVs’ movement and hovering energy consumption.