Abstract: The finite time stabilization was discussed for an incomplete symmetry unmanned platform under a model perturbation and environmental disturbance condition. Due to the non-diagonal inertia and damping matrices existing in the original model, the investigated model was converted into two cascade subsystems with the global diffeomorphism transformation just for once to simplify the process of calculations. For these subsystems, a finite time convergence controller was developed based on a terminal sliding mode and back-stepping technique to improve the robustness and convergence speed. Introducing a first order low-pass filter, the composedness controller was optimized to prevent the saturation of the actuator. A new nonlinear disturbance observer was designed to estimate and compensate the perturbations and disturbances for the improvement of the controller's robustness without the definite boundaries of the estimation errors. Finally, the system was proved to be mean-square stable based on the Lyapunov theorem. The simulation results validate the effectiveness and robustness of the proposed controller.