Inspired by biological balance and stability control, increasingly diverse tail designs have been applied in robot balance control. However, there are no precedents for using tails in balance and stability control for wheeled-legged robots. Drawing inspiration from a cat’s ability to reorient itself mid-air, we propose a method using a tail to enable four-wheeled-legged robots to self-adjust during falls, thereby minimizing fall-induced damage. We simplify the robot as a rectangular prism and a system consisting of a single rigid body and a point mass tail in a zero-gravity environment. Using the MSDDP algorithm, we find a successful self-adjustment trajectory during the fall. We then compare this trajectory with the strategy found using the DDP algorithm. The results demonstrate that our proposed method offers higher convergence speed and stability, lower sensitivity to initial values, and greater computational efficiency.