To address the resonance problem of wind farms with grid-following (GFL ) static var generator (SVG )，a resonance detection model based on the fuzzy cognitive map (FCM ) algorithm was proposed.Granular computing combined with Fourier transform was used to preprocess the operation data of the constructed wind farm.Based on the processed data，the model was trained，and the detection accuracy test was completed.By using the overall sequence impedance model of the constructed wind farm combined with the impedance analysis method，the influence of SVG using GFL and grid-forming (GFM ) control on the overall stability of the wind farm system was analyzed，and a wind farm resonance suppression method based on SVG control mode switching was proposed.The electromagnetic simulation model of the wind farm system with SVG was established by using StarSim-HIL，and the simulation results prove that the proposed resonance detection and suppression methods can quickly and accurately realize the detection and suppression of resonance.