Abstarct: Due to the importance of fault detection in maintaining the performance and immunity of control process, In this thesis, In addition to the fault detection problem, The issue of system robustness to uncertain and unknown inputs is important. In this regard it should be considered a trade-off between fault sensitivity of the system and its robustness. The Η_2⁄Η_∞ method uses a fault detection problem for a dynamic system with faults modeling and unknown inputs. The proposed method is baxsed on transforming robust fault detection problem into a standard model matching Η_2⁄Η_∞ problem. Initially, we select an appropriate reference model for wind turbine system in terms of robust fault detection. Then, a fault detection filter will be designed baxsed on minimization Η_2⁄Η_∞ of deference between reference model and available residual generation using LMI tools.