Analysis and Suppression of the Pump Light Intensity Error in the K–Rb–²¹Ne Atomic Spin Comagnetometer

As a novel quantum sensor utilizing atomic spin coupling mechanisms, spin-exchange relaxation-free (SERF) comagnetometers demonstrate promising applications in precision metrology, such as magnetic field and rotation sensing. However, the ultrahigh-precision measurement capability exhibits vulnerability to multiple disturbance factors, with pump light intensity fluctuations emerging as a critical instability source that compromises operational reliability during prolonged measurements. In this article, a model-driven paradigm for pump light intensity error suppression in SERF comagnetometers is proposed. The nonlinear dynamics are rigorously linearized via Taylor expansion within their valid range to construct a state-space model, ultimately deriving the pump light intensity frequency response model. The proposed methodology undergoes experimental validation on a K–Rb–²¹Ne comagnetometry platform, confirming the predictive capability for stability degradation mechanisms. Through theoretical analysis and experimental verification, this study reveals that magneto-optical nonorthogonality, which introduces a differential component in the light intensity transfer function, creates an approximate proportional relationship between the output signal and pump light intensity within the operating bandwidth. Light intensity noise is decomposed, and light intensity errors are quantitatively evaluated via the frequency response model. To address this critical issue affecting inertial measurement accuracy, a closed-loop control-based suppression method with optimized feedback gain is proposed. By dynamically adjusting the closed-loop feedback gain through real-time intensity monitoring, the intensity instability associated with optical pumping is suppressed from 0.56% to 0.14%, concomitant with a 2.8-fold improvement in the long-term stability of the SERF comagnetometer.