Reconfigurable intelligence surface and unmanned aerial vehicle (RIS-UAV) assisted wireless communications is potential for new generation of Internet of Things (IoT) networks. However, the network performance is severely affected by the channel correlation in practical systems. To address this issue, we investigate the energy efficiency optimization in RIS-UAV assisted networks with channel correlation in this paper. Specifically, we propose an energy efficiency optimization scheme, which first performs user grouping to deal with the channel correlation issue, and then optimizes the system energy efficiency through UAV trajectory optimization and RIS passive beamforming based on double deep Q-Network (DDQN). The numerical outcomes have conclusively validated the efficacy of the proposed scheme.