<p>Dynamic State Estimation (DSE) is pivotal to the effective operation and maintenance of power systems. It enables the prediction of system state variables and facilitates advanced functionalities such as real-time surveillance, fault diagnosis, and grid planning. Nevertheless, challenges arise from measurement inaccuracies, modeling discrepancies, and imprecise initialization, which can undermine DSE’s performance, culminating in estimation inaccuracies and compromised tracking of system dynamic states. In pursuit of enhanced DSE precision, this study initiates by formulating a comprehensive dynamic model encapsulating both transmission line and generator dynamics, subsequently adopting the Extended Kalman Filter (EKF) as the framework for state estimation. To counteract the detrimental effects of systemic parameter inaccuracies on estimation outcomes, this paper proposes an innovative parameter refinement strategy grounded in the Bacteria Foraging Optimization Algorithm (BFOA). This proposition capitalizes on the outcomes from static state estimation exercises and harnesses the BFOA’s prowess in tackling black-box optimization tasks via iterative improvement. By identifying parameter configurations that maximize estimation accuracy, the proposed methodology corrects erroneous line parameters, thereby enhancing the precision of dynamic state estimation. Simulated validations within the IEEE 39-bus benchmark system affirm the efficacy of our approach. Results emphatically demonstrate that post-correction parameters markedly elevate DSE accuracy, efficaciously mitigating estimation discrepancies, elevating the signal-to-noise ratio, and ensuring closer adherence to actual system behaviors. Consequently, the research outcomes constitute a pivotal advancement in bolstering the operational safety and reliability of power grids, underscoring their profound practical implications for the energy sector.</p>