Aiming at the practical demand of high-precision servo control of electric cylinder lifting mechanism, a compound control strategy based on a limited memory interval cyclic least(CLS) square method and a variable proportional integral(VPI) controller based on iterative controller parameters and observer parameters combined with proportional multiple integral observer based on the Kalman filter(VPMISAKF) to identify the servo loop and differential excitation signals is proposed. The experimental results show that compared with the least square method, the inertia and damping root mean square errors of the CLS identification method are reduced by 93.61% and 82.39% respectively. Compared with the PI controller, the composite control strategy of the VPI and VPMISAKF improves the step response fit of the system to 0.994 and 0.991 under no-load and loaded conditions respectively. Compared with the VPI controller, the combined control strategy of the VPI and VPMISAKF reduces the sinusoidal response error of the system by 88.89% and 86.45% under no-load and loaded conditions respectively. The CLS identification method and the compound control strategy of the VPI combined with the VPMISAKF have certain reference significance for high-precision servo control of electric cylinder lifting mechanism.