Abstract: To tackle the low efficiency problem of data collection in UAV-assisted Wireless Sensor Networks (where UAV is the abbreviation for Unmanned Aerial Vehicle), an energy efficient data collection scheme named DPKM-PN ("Density Peak K-Means" combined with "Pointer Networks") is proposed. Firstly, a new node-clustering algorithm is established based on the Density-Peak-Clustering method and the K-Means-Clustering method; Then, the access sequence and hover positions of UAV to all the clusters of Wireless Sensor Networks are optimized. The experimental results show that, compared with the Ptr-A^* scheme proposed recently, the DPKM-PN scheme can reduce the energy consumption of the sensor nodes by about 7.9%, and the total energy consumption of the whole system by about 6.3%, effectively improving the data collection efficiency of the system.