Abstract: To address the issues of strong historical data dependency and limited estimation accuracy in the adaptive extended Kalman filter (AEKF) algorithm when processing colored noise, a collaborative optimization algorithm based on dung beetle optimization (DBO) and forgetting factor AEKF was proposed, where a forgetting factor was introduced to dynamically adjust the weight allocation of historical data, thereby reducing its interference with current estimations and improving the AEKF update-correction model. Furthermore, utilizing the strong global optimization capability of the DBO algorithm and employing the signal-to-noise ratio (SNR) as the objective function evaluation metric, a dynamic optimization mechanism for the forgetting factor was established. This ultimately leaded to the development of an adaptive FFAEKF-DBO collaborative optimization algorithm. Experiments were conducted using four types of typical colored noise signal datasets and the rolling bearing vibration dataset provided by Case Western Reserve University (CWRU) to evaluate the performance of the proposed algorithm across multiple dimensions. The results demonstrate that, in colored noise tests, the proposed algorithm exhibits significant advantages over traditional EKF, AEKF, and FFAEKF methods. Specifically, it achieves average reductions of 11.50% in mean absolute error (MAE) and 11.36% in root mean square error (RMSE), along with an average SNR improvement of 1.11 dB, while consistently maintaining the best normalized cross-correlation (NCC) values. In the analysis of rolling bearing vibration signals, the proposed algorithm significantly reduces the mean square error while preserving signal stability and consistently outperforms comparative algorithms in NCC.By introducing a dynamically optimized forgetting factor mechanism, the proposed algorithm effectively mitigates the historical data dependency issue in AEKF and significantly enhances colored noise suppression accuracy. This provides a reliable technical solution for industrial signal processing and fault diagnosis applications.