The control method of exoskeleton system for lower limb rehabilitation is complex, and there are many internal and external factors to be considered, which makes the system model exist uncertainties. Through the analysis of the lower extremity exoskeleton, the corresponding equation is obtained, and then linearized. Considering the internal and external factors of the connecting rod, the uncertain parts are introduced and the optimal control method is adopted to control the system. In this paper, LQR control and robust optimal control are used to not only optimize the state of the lower extremity exoskeleton system, but also make the system robustness and stability. Finally, the physical parameters of the lower extremity exoskeleton are given, and simulation analysis is carried out to verify the effectiveness of the control method.