With the access of wind power, the nonlinear characteristics of the power grid system are intensified. The traditional damping control unit is no longer suitable in this case, and there may even be failure risk in extreme cases, which brings hidden dangers to the safety of the power system. We put forward a low frequency oscillation suppression strategy for power systems including wind power access: A multi machine PSS collaborative sparse adaptive optimization, which adaptively updates parameters in the PSS that are strongly related to system stability based on real-time state changes in the system, ensuring full performance of the controller and improving system stability. Finally, the scientificity of this method was verified through numerical simulation.