Resilient evaluation algorithm of caesium fountain clock based on robust adaptive observation model

Abstract The caesium fountain clock (CSF) is a primary frequency standard used to define the second, which can greatly improve the autonomous timekeeping ability of laboratories. However, CSF are susceptible to issues such as data interruptions, outliers, and significant fluctuations during operation, ultimately hindering their application in timekeeping. To address these issues, a resilient evaluation algorithm based on a robust adaptive observation model has been proposed. Robust and adaptive factors were introduced into the observation model of the Kalman filter to mitigate abnormal CSF operations. The proposed algorithm was tested through simulations and experiments, which demonstrated significant improvements in the accuracy and stability of the evaluation. As a result, this algorithm effectively reduces the impact of abnormal CSF operations and expands the application scope of CSF in the field of timekeeping.