Dsp-based atomic magnetometer rotation angle adaptive detection system and method

Abstract

The application discloses a system and method for adaptive detection of optical rotation angle of an atomic magnetometer based on DSP, and belongs to the technical field of quantum precision measurement. The system and method are characterized in that: an ADC circuit is used to collect light intensity signals, a recursive least square (RLS) adaptive filtering algorithm with a forgetting factor is used to demodulate the optical rotation angle signals in real time, a Jones matrix modeling and orthogonal demodulation technology are combined to invert the magnetic field intensity, and the detection precision is improved. The system and method overcome the problems of high cost and insufficient noise suppression capability of a traditional phase-locked amplifier, have the advantages of low cost, high sensitivity (aT level), strong anti-interference capability, flexibility and scalability, and are suitable for extremely weak magnetic field measurement scenes such as biological magnetic signal detection and inertial navigation.

Description

Technical Field

[0001] This invention belongs to the field of quantum precision measurement technology, specifically relating to an adaptive detection system and method for optical rotation angle of an atomic magnetometer based on DSP. Background Technology

[0002] An atomic magnetometer is a highly sensitive magnetic sensor based on the principle of quantum precision measurement. It inverts the magnetic field strength by detecting the change in the optical rotation angle of atomic spins under the influence of a magnetic field, and has important applications in biomedical imaging, geophysical exploration, and inertial navigation. However, the traditional optical rotation angle signal detection technology of atomic magnetometers still faces the following key problems:

[0003] High costs and import dependence mean that existing systems typically use lock-in amplifiers (LIAs) for signal demodulation. The high hardware costs and reliance on imported core technologies result in exorbitant system costs, limiting their widespread application in both civilian and research fields. Furthermore, the fixed bandwidth and static parameters of LIAs make them ill-suited to adapting to signal variations in dynamic environments, resulting in insufficient flexibility.

[0004] Insufficient noise suppression capability makes the optical rotation angle signal susceptible to environmental interference (such as temperature fluctuations, laser power drift, mechanical vibration, etc.), especially in the low-frequency band (1 / f noise) and under photoelastic modulation (PEM) harmonic interference, significantly reducing the signal-to-noise ratio of traditional differential polarization methods or Faraday modulation methods. Existing technologies mostly rely on hardware filtering or fixed-parameter digital filtering, which cannot dynamically track noise characteristics, resulting in limited measurement accuracy.

[0005] The algorithm has poor adaptability. Traditional adaptive filtering algorithms (such as LMS) have slow convergence speed and large steady-state error when dealing with frequency drift and phase jitter of PEM signals, making it difficult to meet real-time requirements. In addition, the nonlinear relationship between the optical rotation angle signal and the magnetic field needs to be accurately modeled, but existing demodulation methods are not good at separating higher harmonics and orthogonal components, which further reduces the magnetic field resolution and sensitivity.

[0006] In recent years, photoelastic modulation technology has become a research hotspot due to its high modulation frequency and thermal stability, but its demodulation process faces challenges in dynamic noise suppression and real-time signal processing. Although some studies have attempted to improve performance using digital phase-locked loops (DPLLs) or Kalman filters, problems such as high computational complexity and difficulty in parameter tuning still exist. Therefore, developing a low-cost, high-precision method for detecting optical rotation angle signals with dynamic noise suppression capabilities has become a pressing technical challenge in this field. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides a DSP-based adaptive detection system and method for optical rotation angle of an atomic magnetometer. Through the recursive least squares (RLS) algorithm and a fully digital demodulation scheme, it achieves real-time high-precision extraction of the optical rotation angle signal, providing an efficient and reliable solution for measuring extremely weak magnetic fields.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] A DSP-based adaptive detection system for the optical rotation angle of an atomic magnetometer, the system comprising a PEM detection optical path unit, a detection optical sampling circuit unit, and a DSP processing unit, which are sequentially and communicatively connected.

[0010] The PEM detection optical path unit is used to modulate the light beam passing through the alkali metal gas cell;

[0011] The detection light sampling circuit unit is used to collect the modulated light beam and perform analog-to-digital conversion to obtain a digital signal;

[0012] The DSP processing unit is used to demodulate the digital signal to obtain the optical rotation angle signal; wherein, the demodulation processing includes adaptive filtering, and the adaptive filtering adopts a recursive least squares algorithm with a forgetting factor λ to dynamically update the filter parameters.

[0013] On the other hand, the present invention provides a DSP-based adaptive detection method for the optical rotation angle of an atomic magnetometer, comprising:

[0014] The PEM detection optical path unit modulates the light beam passing through the alkali metal gas cell;

[0015] The detection optical sampling circuit unit collects the modulated light beam and performs analog-to-digital conversion to obtain a digital signal;

[0016] The DSP processing unit demodulates the digital signal to obtain the optical rotation angle signal; wherein, the demodulation processing includes adaptive filtering, and the adaptive filtering adopts a recursive least squares algorithm with a forgetting factor λ to dynamically update the filter parameters.

[0017] Thirdly, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned DSP-based adaptive detection method for the optical rotation angle of an atomic magnetometer.

[0018] The beneficial effects of this invention are as follows:

[0019] Significantly reducing costs and enhancing self-controllability, this invention replaces the traditional lock-in amplifier (LIA) with a digital signal processor (DSP). Through a fully digital signal demodulation scheme, it avoids dependence on imported equipment, reducing system hardware costs by more than 50%. Furthermore, the self-developed C language-based programming design allows the algorithm to be flexibly adjusted according to actual needs, further enhancing the system's autonomy and adaptability.

[0020] This invention significantly improves detection accuracy and response speed by introducing a recursive least squares (RLS) adaptive filtering algorithm with a forgetting factor. This invention effectively overcomes the problems of slow convergence speed and large steady-state error of the traditional LMS algorithm.

[0021] With strong anti-interference capabilities and dynamic adaptability, the system effectively suppresses low-frequency noise and system errors caused by laser power drift and temperature fluctuations by dynamically adjusting the forgetting factor λ and optimizing filter parameters in real time. Furthermore, the combined use of sliding window mean filtering and bandpass filtering further enhances noise suppression during signal preprocessing, ensuring stable detection in complex environments.

[0022] With its flexible and scalable software design, this invention adopts a modular software architecture, allowing key parameters (such as preprocessing window size and filter order) to be adjusted online, adapting to different application scenarios without hardware modifications. This design not only facilitates algorithm iteration and optimization but also provides a technical foundation for subsequent functional expansion (such as multi-channel synchronous detection), significantly improving the system's practicality and applicability. Attached Figure Description

[0023] Figure 1 This is a block diagram of the adaptive detection system for the optical rotation angle of an atomic magnetometer based on DSP according to the present invention. Detailed Implementation

[0024] like Figure 1 As shown, the present invention provides an adaptive detection system for optical rotation angle of an atomic magnetometer based on DSP. The system includes a PEM detection optical path unit, a detection optical sampling circuit unit, and a DSP processing unit that are connected in sequence. The PEM detection optical path unit is used to modulate the light beam passing through the alkali metal gas cell.

[0025] The detection light sampling circuit unit is used to collect the modulated light beam and perform analog-to-digital conversion to obtain a digital signal;

[0026] The DSP processing unit is used to demodulate the digital signal to obtain the optical rotation angle signal; wherein, the demodulation processing includes adaptive filtering, and the adaptive filtering adopts a recursive least squares algorithm with a forgetting factor λ to dynamically update the filter parameters.

[0027] Specifically, the PEM detection optical path unit includes a polarizer, a quarter-wave plate, a PEM, and an analyzer connected in sequence. The laser beam is converted into linearly polarized light by the polarizer, and then sequentially passes through a gas cell, the quarter-wave plate, the PEM, and the analyzer to form an intensity modulation signal carrying optical rotation angle information. The polarizer and analyzer are in mutually extinct positions, the optical axis of the quarter-wave plate is aligned with the transmission axis of the polarizer, the PEM modulation axis is 45°, and the modulation frequency is [missing information]. .

[0028] The laser beam is converted into linearly polarized light by a polarizer, and its Jones vector... It can be written as:

[0029] ,

[0030] in, Let be the electric field amplitude of the incident laser beam, and let the intensity of the incident laser beam be expressed as: ;

[0031] The air cell is simulated as a half-wave plate, and the light rotation angle is generated after passing through the air cell. Its Jones vector can be described by the following formula:

[0032] ,

[0033] The Jones vector of a quarter-wave plate with a fast axis orientation of 90° is:

[0034] ,

[0035] The PEM is modeled as a time-varying quarter-wave plate, and its Jones vector with a modulation axis at 45° can be represented as:

[0036] ,

[0037] in, This is the delay of the PEM. The time-varying phase delay of the PEM has the following equation form:

[0038] ,

[0039] The modulation frequency of the photoelastic modulator. PEM modulation depth;

[0040] The Jones vector of an analyzer with a transmission axis of 90° can be approximated as:

[0041] ,

[0042] Then, the total electric field Jones vector obtained from the analyzer for:

[0043] .

[0044] The detection light sampling circuit unit includes a photodetector (response bandwidth ≥10MHz), an AD8132 single-ended to differential circuit, and an AD9248 ADC module (14-bit precision, 65MSPS sampling rate).

[0045] The photodetector is used to collect the light intensity modulation signal output by the analyzer, obtain light intensity information based on the total electric field Jones vector, and convert the light intensity information into a digital signal through a differential circuit and an ADC module. Specifically:

[0046] Light intensity collected by photodetector :

[0047]

[0048] ,

[0049] In the formula, for Conjugate;

[0050] Expanding using the first kind of Bessel formula, we get:

[0051] ,

[0052] ,

[0053] In the formula, For the zeroth order of the Bessel function, and Substituting the 2k and 2k-1 orders of the Bessel function, we get:

[0054]

[0055] ,

[0056] This represents the higher harmonic components, where t is time.

[0057] In the formula and <<1, In the above formula, only the first and second order terms of the Bessel function, which have a significant impact on light intensity, are considered. The light intensity obtained by the photodetector is then transformed into:

[0058] ,

[0059] in, For phase error, The light intensity is obtained by the photodetector.

[0060] Light intensity In the single-ended to differential input circuit, the input is converted into two differential signals with equal amplitude, opposite phase, and containing common-mode noise n through the internal circuit structure. and Then, input the two differential signals from the differential input pins of the ADC, and first calculate the difference between the two differential signals in the ADC:

[0061] ,

[0062] This effectively eliminates common-mode noise, and the difference is the light intensity obtained by the photodetector. Then convert it into a digital value for output.

[0063] The DSP processing unit includes an ADC signal acquisition module, a front-end preprocessing module for processing the digital signals acquired by the ADC, and an adaptive filtering module. The ADC signal acquisition module is used to receive the digital signals. The front-end preprocessing module is used to filter out the high-frequency components and DC bias components of the digital signals to obtain a first filtered signal. The adaptive filtering module is used to demodulate and perform secondary filtering on the first filtered signal to obtain the optical rotation angle signal.

[0064] As the central hub for real-time signal processing, the DSP processing unit achieves high-precision ADC acquisition of the photodetector output signal through a high-speed interface, ensuring that the sampling rate matches the photoelastic modulation frequency to avoid signal distortion. As an algorithm execution platform, its dedicated digital signal processing unit (such as a multiplier-accumulator) can efficiently run signal preprocessing and adaptive filtering algorithms. The preprocessing removes DC bias through sliding window mean filtering and suppresses high-frequency noise through bandpass filtering, laying the foundation for subsequent demodulation. As the core of system control, the DSP coordinates sampling timing, interrupt response, and data transmission, ensuring real-time performance throughout the entire process from signal acquisition to optical rotation angle calculation. It also supports flexible adjustment of algorithm parameters through programming to adapt to different detection scenarios.

[0065] The front-end preprocessing module performs a moving average to remove the DC bias component from the digital signal and a low-pass filter to remove high-frequency components. After the high-harmonic components and DC bias components of the digital signal are filtered out, the first filtered signal is obtained. :

[0066] ,

[0067] In the adaptive filtering module, the above signal is first processed. Demodulation is performed, specifically using the reference signal. and phase difference of 90° The above results By multiplying them sequentially and then converting the product to the sum and difference, two demodulated signals can be obtained:

[0068] ,

[0069] ,

[0070] Then, adaptive filtering is applied to these two demodulated signals to remove higher harmonic components. and The I and Q components are obtained respectively:

[0071] ,

[0072] ,

[0073] From these two components, we can derive:

[0074] ,

[0075] Finally, the optical rotation angle signal can be demodulated. :

[0076] .

[0077] The adaptive filtering module of this invention is based on the recursive least squares (RLS) algorithm with a forgetting factor λ, which achieves high-precision demodulation of the rotation angle signal by dynamically adjusting the filter parameters. Its core principle is that the algorithm aims to minimize the error ε(n) and continuously updates the filter parameters w(n) recursively, so that the deviation between the estimated output and the real signal approaches zero.

[0078] In this process, the forgetting factor λ plays a crucial role: when the ambient noise intensity increases, the system automatically decreases the value of λ, reducing the weight of historical data and making the algorithm focus more on current new data, thus accelerating parameter convergence to cope with rapidly changing interference; when the noise intensity is low, the value of λ is increased to fully utilize the statistical characteristics of accumulated historical data, improve filtering accuracy, and reduce the impact of random fluctuations. This mechanism of dynamically adjusting the forgetting factor and filter parameters according to the noise environment gives the algorithm adaptive capabilities, effectively suppressing interference such as high-order harmonics and laser power drift during photoelastic modulation. Compared with traditional fixed-parameter filtering algorithms, it significantly improves the demodulation accuracy of the optical rotation angle signal and the system's anti-interference capability.

[0079] On the other hand, the present invention provides a DSP-based adaptive detection method for the optical rotation angle of an atomic magnetometer, comprising:

[0080] The PEM detection optical path unit modulates the light beam passing through the alkali metal gas cell;

[0081] The detection optical sampling circuit unit collects the modulated light beam and performs analog-to-digital conversion to obtain a digital signal;

[0082] The DSP processing unit demodulates the digital signal to obtain the optical rotation angle signal; wherein, the demodulation processing includes adaptive filtering, and the adaptive filtering adopts a recursive least squares algorithm with a forgetting factor λ to dynamically update the filter parameters.

[0083] Thirdly, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned DSP-based adaptive detection method for the optical rotation angle of an atomic magnetometer.

[0084] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A DSP-based adaptive detection system for the optical rotation angle of an atomic magnetometer, characterized in that, The system includes a PEM detection optical path unit, a detection optical sampling circuit unit, and a DSP processing unit, which are connected in sequence via communication. The PEM detection optical path unit is used to modulate the light beam passing through the alkali metal gas cell; The detection light sampling circuit unit is used to collect the modulated light beam and perform analog-to-digital conversion to obtain a digital signal; The DSP processing unit is used to demodulate the digital signal to obtain the optical rotation angle signal; wherein, the demodulation processing includes adaptive filtering, and the adaptive filtering adopts a recursive least squares algorithm with a forgetting factor λ to dynamically update the filter parameters; The PEM detection optical path unit includes a polarizer, a quarter-wave plate, a PEM, and an analyzer connected in sequence. The laser beam is converted into linearly polarized light by the polarizer, and then passes through the gas cell, the quarter-wave plate, the PEM, and the analyzer in sequence to form an intensity modulation signal carrying optical rotation angle information. The polarizer and the analyzer are in mutual extinction positions, the optical axis of the quarter-wave plate is aligned with the transmission axis of the polarizer, and the modulation axis of the PEM is 45°. The DSP processing unit includes an ADC signal acquisition module, a front-end preprocessing module, and an adaptive filtering module. The ADC signal acquisition module receives the digital signal. The front-end preprocessing module filters out the high-frequency components and DC bias components of the digital signal to obtain a first filtered signal. The adaptive filtering module demodulates and performs secondary filtering on the first filtered signal to obtain the optical rotation angle signal. The adaptive filtering module uses a digital adaptive filter for filtering. This digital adaptive filter is based on a recursive least squares algorithm with a forgetting factor λ, dynamically updating its filter parameters. When the ambient noise intensity increases, the system automatically decreases the forgetting factor λ value, reduces the weight of historical data, focuses on current new data, and accelerates parameter convergence. When the noise intensity is low, the forgetting factor λ value is increased to fully utilize the statistical characteristics of historical data accumulation.

2. The DSP-based adaptive optical rotation angle detection system for an atomic magnetometer according to claim 1, characterized in that, The PEM detection optical path unit predicts the evolution of the polarization state of the detection light using a Jones matrix, and the total electric field Jones vector of the laser beam after passing through the PEM detection optical path unit is... Represented as: ， in, This represents the deflector Jones vector. Represents the Jones vector of the gas chamber. The Jones vector representing a quarter-wave plate with a fast axis orientation of 90°. The Jones vector representing PEM. The Jones vector represents the analyzer with a transmission axis of 90°.

3. The DSP-based adaptive optical rotation angle detection system for an atomic magnetometer according to claim 1, characterized in that, The detection light sampling circuit unit includes a photodetector, a differential circuit, and an ADC module. The photodetector is used to collect the light intensity modulation signal output by the analyzer, obtain light intensity information based on the total electric field Jones vector, and convert the light intensity information into a digital signal through the differential circuit and the ADC module.

4. The DSP-based adaptive optical rotation angle detection system for an atomic magnetometer according to claim 3, characterized in that, The light intensity signal obtained by the photodetector Represented as: ， in, The intensity of the incident laser beam. The optical rotation angle produced after passing through the gas cell. PEM modulation depth The frequency is the PEM modulation frequency, and t represents time. It is a phase error. It is a higher harmonic component obtained based on the expansion of Bessel's formula.

5. The DSP-based adaptive optical rotation angle detection system for an atomic magnetometer according to claim 1, characterized in that, The front-end preprocessing module performs a moving average to remove the DC bias component from the digital signal and performs a low-pass filter to remove the high-frequency component.

6. A DSP-based adaptive detection method for the optical rotation angle of an atomic magnetometer, applied to the system described in any one of claims 1-5, characterized in that, include: The PEM detection optical path unit modulates the light beam passing through the alkali metal gas cell; The detection optical sampling circuit unit collects the modulated light beam and performs analog-to-digital conversion to obtain a digital signal; The DSP processing unit demodulates the digital signal to obtain the optical rotation angle signal; wherein, the demodulation processing includes adaptive filtering, and the adaptive filtering adopts a recursive least squares algorithm with a forgetting factor λ to dynamically update the filter parameters.

7. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When one or more programs are executed by the one or more processors, the one or more processors implement the DSP-based adaptive detection method for optical rotation angle of an atomic magnetometer as described in claim 6.

Images