Abstract. This article focuses on the improvement of the control performance of the electro-hydraulic position servo system. A three-order state feedback (3OSF) controller is designed to eliminate static error. For the purpose that overcome the inadequacy due to high order in the 3OSF control system, a weighted switching algorithm in the finite frequency domain is adopted to realize the switching control using the 3OSF-2OSF dual-controller. Experiments show that the control system using the 3OSF-2OSF dual-controller weighted switching in the finite frequency domain has better dynamic performance and static performance than the control system using the 2OSF. Comparing with the threshold switching, the developed switching algorithm has smaller switching impact.Introduction. Electro-hydraulic servo systems are widely used in industrial applications, national defense, aerospace, robots and other fields because of their high power density, high stiffness, fast response, self-cooling, good positioning capabilities, etc [1]. However, electro-hydraulic servo systems have time-varying parameters, uncertainties and nonlinearity, especially the jumping parameters along with the change of working conditions. Output feedback is the commonly used control method for electro-hydraulic servo system, but the control performance optimization is limited due to the incompleteness of system description and the information controller design with. Output feedback can not meet the demands of the increasingly improved control precision. State variables can comprehensively reflect the system internal characteristics, so the state feedback can improve system performance more effectively than traditional output feedback. Due to the phenomenon such as pressure-flow characteristics, hysteresis in flow gain characteristics, oil leakage, oil temperature variations, characteristics of valves near null, and so on. The exact dynamic model of electro-hydraulic servo systems is hard to be built. The model error limits the improvement of state feedback control system dynamic performance. The feedback linearization can eliminate the system extrinsic nonlinear