Multi-objective Optimization for SERF Atomic Magnetometer using Zero-field Parameter Modulation

The spin-exchange relaxation-free (SERF) atomic magnetometers exhibit ultrahigh sensitivity for the magnetic field measurement, and their performance is further enhanced through zero-field parameter modulation. However, this modulation introduces additional parameter coupling that limits the performance of magnetometers in multi-axis measurement. In this paper, we develop a modified dual-axis output equation based on the nonlinear interaction between the optical pumping rate and modulation field amplitude, mediated by the spin-exchange relaxation and polarization. We employ a multi-objective optimization algorithm incorporating the implicit constraints among these parameters to enable efficient, simultaneous optimization of both the pump rate and modulation amplitude. Experimental results show that under optimal parameters obtained by the algorithm, the dual-axis signal responses improve by approximately 20% and 30% over conventional optimization methods, achieving 2.5 lT/Hz¹/² magnetic measurement sensitivities for the j-axis and 4.0 lT/Hz¹/² for the A-axis, respectively.