In order to improve the computational efficiency and convergence of the algorithm, this paper derives the pseudo-spectral discrete Lagrange equation based on Lagrange mechanics, discrete mechanics and optimal control (DMOC) calculation method, combining with the advantages of high accuracy of the pseudo-spectral method. In conjunction with the successive convex optimization technique, a symplectic pseudo-spectral successive convex optimization-based trajectory optimization method is suggested. By significantly reducing the dimension of the discrete system’s state variables, the symplectic pseudo-spectral sequential convex optimization method can considerably increase convergence and computational efficiency while maintaining the structural features of the original continuous system. In contrast to the classical pseudo-spectral successive convex optimization method, the simulation results demonstrate that the symplectic pseudo-spectral successive convex optimization method has a good adoptability to the initial disturbance and can significantly increase computational efficiency without sacrificing accuracy.