This paper investigates the collaborative optimization performance route on the reconfigurable intelligent surface (RIS)-aided integrated satellite unmanned aerial vehicle-terrestrial network (IS-UAV-TN). It mainly has two primary challenges: long transmission paths with severe signal fading and the presence of obstacles along the wireless environment. To efficiently address above issues, it is assumed that RIS is deployed on the UAV to enable precise control over signal transmission, and non-orthogonal multiple access (NOMA) transmission strategies are employed at the transmitter. Based on the different transmission phases, a multi-vector optimization problem is proposed with the aim of maximizing the system achievable rate by optimizing the UAV path design, transmit beam and RIS passive phase shift. Especially the deep reinforcement learning (DRL)-based multi-vector deep deterministic policy gradient (MV-DDPG) algorithm, is utilized to achieve adaptive decision-making in non-deterministic wireless environments. Experimental results show that the proposed method has the ability to efficiently consider multiple optimization objectives designed compared to traditional comparison schemes.