LCL-type converters are widely used in grid-connected systems due to their small size and good filtering performance. However, the resonance suppression problem brought by its own application is also not negligible. The active damping method of capacitive current feedback is a commonly used resonance suppression method. In practical applications, the effect of the grid impedance on the resonance of LCL filter cannot be ignored. On this basis, an adaptive resonance suppression method based on genetic algorithm is proposed to optimize the RBF neural network. The initial parameters of the RBF neural network are optimized according to the genetic algorithm. The parameters of the PI controller are used to identify the parameters of the RBF neural network. Identification, real-time correction of PI control parameters and active damping coefficient, thus achieving the LCL-type converter to maintain system stability when the grid impedance changes. The effectiveness of the proposed method is verified by simulation experiments.