Abstract: This paper proposes an adaptive wideband disturbance control for beam jitter in optical stabilization systems. The control strategy transforms jitter suppression into real-time optimization of multiple parallel peak filters, which are achieved based on Youla parameterization observation. We identify the dynamic center frequency of the interference signal using the closed-loop system error. This identification employs a normalized unbiased gradient algorithm, and the bandwidth parameters are adaptively adjusted by combining the recursive least squares method to achieve precise tracking of multi-frequency interferences. Additionally, a cascade filter is innovatively designed to offset the system delay, which improves the system stability. Theoretical analysis and experimental verification show that this method has a significant suppression effect on multi-frequency broadband interferences, significantly enhancing the anti-disturbance capability of the system.