Enhancing dynamic performance of vapor cell temperature control system in SERF magnetometer

The vapor cell serves as the core sensing element in a Spin-Exchange Relaxation-Free (SERF) atomic magnetometer (AM). A platinum resistance is attached to the outer wall of the vapor cell, opposite the heater, to monitor temperature. However, the measurement delay adversely affects the dynamic performance of the temperature control system. To address this issue, this paper integrates smith predictor and linear active disturbance rejection control (Smith-LADRC) to enhance the dynamic performance of vapor cell temperature control system in SERF AMs. First, the parameters of the vapor cell heating model are identified using experimental and least squares methods, with an accompanying analysis of the inter-channel variations and their origins within the 8-channel AM. Then, a temperature control system based on Smith-LADRC is designed and implemented. Compared to conventional PID, the proposed method exhibits a 2.21 % reduction in the overshoot of the vapor cell’s temperature rise curve and shortens the time required to stabilize within 10 mK to 214.6 s. Moreover, the maximum temperature deviation after disturbance is reduced by 57.14 %, and the recovery time is shortened by 66.93 %. Finally, the influence of vapor cell temperature stability on the low frequency noise and magnetic measurement stability of the AM is analyzed. Experimental results demonstrate that the best measurement performance is achieved when the vapor cell temperature is stabilized within ±10 mK of the target value.