The fusion of radar sensing and wireless communication has sparked much interest in the research of integrated sensing and communication (ISAC) systems, among which using MIMO-OFDM systems to sense moving targets is a key task. However, the existing FFT-based moving target parameter estimation methods are severely limited by grid resolution limitations, and the sensing scheme based on a single base station can only obtain the radial velocity of the target. This paper transforms the problem of estimating moving target parameters into a generalized array signal form. A moving target estimation method based on the ESPRIT algorithm is proposed, achieving super-resolution and low-complexity estimation of target velocity and distance. Moreover, we offer a target actual velocity estimation scheme based on double BSs, in which the two-dimensional actual velocity vector can be estimated based on the radial velocity and geometric relationship. Simulation results have proved the effectiveness of the proposed scheme.