An automatic frequency stabilization system based on double-core correlation matching intelligent peak seeking

Abstract

The application discloses an automatic frequency stabilization system based on double-core correlation matching intelligent peak seeking, comprising the following steps: S1, saturated absorption spectrum light path and circuit are built and debugged and optimized; S2, double matching templates are constructed; S3, a triangular wave voltage scanning and a clock timing are started, and signals are collected and transmitted; S4, a KCF algorithm is started, a template tracking model is matched, the obtained results are distinguished, a peak point is found, and a corresponding scanning voltage is calculated; S5, the peak point scanning voltage is recorded, the scanning signal is turned off, and a sinusoidal signal modulation is started at the same time, and the peak value corresponding triangular wave scanning voltage is given to the sinusoidal signal as an offset. ‑12 The application can effectively locate the suspected area of the spectral line peak value, the double-branch and correlation filtering tracking method can accurately find the suspected area for a second time, the locking precision is greatly improved, and the long-term stability of 10

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic frequency stabilization system, and particularly relates to an automatic frequency stabilization system based on double-core correlation matching intelligent peak searching. BACKGROUND

[0002] Gravity is a cross-disciplinary subject of gravimetry and seismology, geology, fundamental physics, geodynamics, hydrology, oceanography and astronomy, etc. It is a key subject to serve the strategic needs of national surveying and mapping datum, resource exploration, military security and disaster monitoring, and to study the basic problems of geoscience. The gravity field can effectively characterize the density distribution of the earth's matter, and the precise gravity field information can provide important physical field information for geophysics, resource exploration, basic physics index detection and navigation, and data support for background information construction. At the same time, it plays a core role in the application of target objects such as oil and gas resource exploration, tidal model construction, seafloor gravity field construction, gravity background field construction, and passive navigation. Therefore, the accurate measurement of gravity index is particularly important for the extension and practical application of basic discipline theory. As a new type of high-precision absolute gravity measurement inertial sensor, cold atom gravimeter has significant performance advantages in gravity measurement. Cold atom gravimeter applies coherent light with strong coherence and good monochromaticity as a coherent light source, and constructs a magneto-optical trap to cool alkali atoms together with an external magnetic field. By adjusting the phase of the laser, the beam is split, reversed, and combined by Raman interference. Further adjustment of the laser frequency scans the interference fringes, and the local gravity value is calculated. With the development of time, cold atom interferometer gradually moves from the static laboratory state to the field, and a certain degree of miniaturization and intelligentization of cold atom gravimeter is developed. As the core of cold atom gravimeter, the optical system is also susceptible to environmental factors in the whole system of mobile cold atom gravimeter, resulting in frequency point loss, drift and other problems. Stable laser frequency has a great influence on atomic cooling, Raman interference, fluorescence detection and gravity dynamic detection accuracy, which seriously restricts the accuracy of gravity measurement. In order to meet the needs of mobility, miniaturization is an important strategy for laser system, and compact laser system is a research direction. The birth of laser has promoted the progress of a series of scientific and technological progress. In order to obtain the required laser, i.e. narrow linewidth and high frequency stability, researchers have studied the laser closed-loop frequency stabilization system. The typical laser closed-loop frequency stabilization system consists of reference frequency, frequency discriminator and feedback controller. The widely used reference frequencies are atomic or molecular transition frequencies and optical resonator characteristic frequencies. The main difference is the different frequency discrimination methods. Using atomic or molecular transition frequency as reference frequency, the frequency discrimination signal is often obtained from atomic or molecular absorption spectrum or dispersion spectrum, and researchers have designed many frequency discrimination methods and frequency stabilization techniques, such as modulation spectrum frequency stabilization technology, modulation transfer spectrum frequency stabilization technology, bias spectrum frequency stabilization technology, magnetic dichroism frequency stabilization technology, etc. The transmission and reflection characteristics of high-Q F-P cavity have similar characteristics with atomic absorption spectrum, with extremely narrow linewidth. The typical frequency stabilization technology based on F-P cavity is pdh frequency stabilization technology.

[0003] The closed-loop frequency stabilization technology of laser has made very outstanding achievements after years of development, and has been widely used in many research fields. However, it has a very obvious deficiency, mainly manifested as instability of the laser frequency stabilization system, i.e. easy to appear out of lock. This is because the effective frequency locking range of the above-mentioned laser closed-loop frequency stabilization technology is usually narrow. When the laser system is subjected to inevitable interference, the laser frequency jumps out of the effective frequency locking range and appears out of lock. This phenomenon seriously limits the application scenarios of the laser closed-loop frequency stabilization system, such as long time, complex environment, unattended, and satellite-borne experiments. Therefore, a small-sized automatic frequency stabilization system without human intervention, which can be applied in various scenes, quickly find back and maintain long-term stability under the out-of-lock condition of laser frequency point loss or drift, is particularly important. However, many devices are based on the peak locking of individual typical Lamb dips without considering the locking mode of all saturated absorption peaks; the mode of calculating the differential signal of a specific transition frequency peak as a frequency locking point is prone to strong fluctuations and drift of the differential signal due to environmental changes in dynamic measurement of a cold atom gravity meter, and the locking point cannot be correctly found. The problem of this mode is that the system repeatedly locks and sweeps to find the locking point, resulting in a long cycle, which will inevitably introduce a long "dead time", causing abnormal gravity measurement. The traditional differential method of finding the locking point works well in static experiments, but has fatal defects in dynamic measurement environment. Artificial intelligence technology has made great progress in recent years, and the computing and perception capabilities of AI large models have been greatly improved. The application of artificial intelligence algorithm can provide a new way for frequency stabilization in complex field environment. The target tracking algorithm can quickly find the target area and lock in the complex strong interference background in real-time target movement, and can be well applied in spectral line identification and tracking. The target tracking algorithm based on kernel correlation filtering can maximize the distinction between foreground and background by constructing a kernel correlation filter, and the ridge regression loss function constructed can effectively separate the foreground and background, so the algorithm is introduced into the peak locking. However, the kernel correlation filtering algorithm only constructs a set of cyclic transfer matrix to find the candidate target in the real-time moving area, and cannot accurately locate. Therefore, an automatic frequency stabilization system based on double kernel correlation matching intelligent peak searching is proposed to solve the above problems. SUMMARY

[0004] The purpose of the present application is to solve the problems existing in the prior art and provide an automatic frequency stabilization system based on double kernel correlation matching intelligent peak searching.

[0005] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0006] An automatic frequency stabilization system based on double kernel correlation matching intelligent peak searching comprises the following steps:

[0007] S1, the saturated absorption spectrum optical path and the circuit are built and debugged and optimized;

[0008] S2, constructing a double matching template;

[0009] S3, starting a triangular wave voltage scanning and clock timing, collecting signals and transmitting;

[0010] S4, starting KCF algorithm, applying template tracking model matching, discriminating the results obtained by matching, finding peak points and calculating corresponding scanning voltage;

[0011] S5, recording the peak point scanning voltage and closing the scanning signal, and at the same time starting the sinusoidal signal modulation, and at the same time giving the peak corresponding triangular wave scanning voltage to the sinusoidal signal as the offset, setting the frequency and amplitude of the sinusoidal signal;

[0012] S6, photoelectric detector collects photoelectric signal, and the same frequency sinusoidal signal is sent into the multiplier through the phase shifter and the photoelectric signal collected by the photoelectric detector, to obtain the pre-differential error signal;

[0013] S7, the error signal is obtained through the low-pass filter with set cut-off frequency;

[0014] S8, constructing PID feedback, taking the direct current component as the differential error signal to send into the PID feedback to form a closed loop to realize the required transition frequency point locking;

[0015] S9, determining whether the error signal is greater than the set threshold value to determine whether to unlock, including: detecting module monitoring whether to unlock, once the error signal is greater than 1×10 -3 or PD detection signal is less than 0.1, indicating that the laser is unlocked, at this time, the laser triangular wave scanning is restarted, the KCF intelligently determines the frequency locking point, the modulation and demodulation, the PID feedback control stabilizes the frequency locking point operation, and the automatic frequency relocking is completed.

[0016] The present application can be used for miniaturization of optical path module and modulation transfer module related to laser driving circuit provides stable current control and intelligent PID temperature control module (current control accuracy 0.01mA, temperature control accuracy 0.01℃), simplify the miniaturization of optical path construction and open the triangular wave scanning cooperation laser driving circuit adjustment and optimization spectrum line, first build the standard curve template containing 6 saturated absorption peaks and the curve area of the specified peak constitute double template, open the triangular wave scanning to obtain real-time spectrum line, analog-digital conversion module collects spectrum line voltage and triangular wave voltage, runs double branch parallel core correlation filtering algorithm to find peak value, parallel running double branch core correlation filtering algorithm can not only effectively locate the suspected area of spectrum line peak value, but also double branch and correlation filtering tracking method can accurately find the suspected area for the second time, which greatly improves the locking accuracy. -12 Long-term stability of 10

[0017] The frequency locking monitoring module of the present application continuously monitors the error signal and the spectrum line voltage signal, when the measurement signal cannot meet the set requirement, it indicates that the laser frequency point drifts or the frequency is abnormal and is unlocked, at this time, the modulation transfer (modulation demodulation) module is closed, the triangular wave scanning module is opened to scan the scanning spectrum line, then the double branch KCF algorithm is opened to find the peak again, when the peak finding meets the set point requirement, it indicates that it is re-locked, the triangular wave scanning module is closed again and the peak point scanning voltage is recorded, the modulation transfer module is opened synchronously, the error signal is fed back by PID to fine tune the peak point scanning voltage to keep the frequency point locked, at the same time, the unlock monitoring module is opened to monitor whether it is unlocked in real time, the intelligent re-locking module can quickly find and complete the locking after unlocking, it can complete the re-locking operation within a few seconds, and it can not be unlocked for several months after re-locking, and it can guarantee the long-term stability of 10 -12 When the vibration occurs, the intelligent KCF peak finding algorithm has stable sample anti-drift interference ability, when the vibration signal is superimposed on the optical path spectrum line signal, it can still meet the double set requirements, therefore, when the vibration signal occurs, it has good anti-vibration interference ability. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 Figure is a laser wavefront probe for detecting laser wavefront condition diagram;

[0019] Figure 2 Figure is a laser temperature control and current driving interface diagram;

[0020] Figure 3 schematic diagram of hyperfine level structure for Rb atom;

[0021] Figure 4 schematic diagram of automatic frequency stabilization system of saturated absorption spectrum;

[0022] Figure 5 schematic diagram of physical optical path construction;

[0023] Figure 6 schematic diagram of triangular wave scanning spectrum;

[0024] Figure 7 schematic diagram of double-template construction of saturated absorption peak;

[0025] Figure 8 schematic diagram of signal generator scanning spectrum;

[0026] Figure 9 schematic diagram of double-branch KCF algorithm;

[0027] Figure 10 schematic diagram of double-template kernel correlation intelligent peak searching automatic frequency stabilization system;

[0028] Figure 11 schematic diagram of overall system flow;

[0029] Figure 12 schematic diagram of frequency stabilization;

[0030] Figure 13 schematic diagram of frequency stabilization spectrum and error signal before and after filtering;

[0031] Figure 14 schematic diagram of wavelength meter stability test;

[0032] Figure 15 schematic diagram of 100-hour stability test;

[0033] Figure 16 schematic diagram of relocking test;

[0034] Figure 17 schematic diagram of Allan variance. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only some of the embodiments of the application, but not all the embodiments of the application.

[0036] Reference Figures 1-17 An automatic frequency stabilization system based on double-kernel correlation matching intelligent peak searching includes the following steps:

[0037] S1, saturated absorption spectrum optical path and circuit construction and debugging optimization;

[0038] S2, constructing a double matching template;

[0039] S3, starting a triangular wave voltage scanning and clock timing, collecting signals and transmitting;

[0040] S4, starting KCF algorithm, applying template tracking model matching, discriminating the results obtained by matching, finding peak points and calculating corresponding scanning voltage;

[0041] S5, recording the peak scanning voltage and closing the scanning signal, and at the same time starting the sinusoidal signal modulation, and setting the frequency and amplitude of the sinusoidal signal as the offset of the peak corresponding triangular wave scanning voltage;

[0042] S6, photoelectric detector collects photoelectric signal, and the same frequency sinusoidal signal is sent into the multiplier through the phase shifter and the photoelectric signal collected by the photoelectric detector to obtain the pre-differential error signal;

[0043] S7, the error signal is obtained through the low-pass filter with set cut-off frequency;

[0044] S8, constructing PID feedback, sending the direct current component into PID feedback as the differential error signal to form a closed loop to realize the required transition frequency point locking;

[0045] S9, determining whether the error signal is greater than the set threshold value to determine whether to unlock, including: detecting module monitoring whether to unlock, once the error signal is greater than 1x10 -3 or PD detection signal is less than 0.1, indicating that the laser is unlocked, at this time, the laser triangular wave scanning is restarted, the KCF intelligently determines the frequency locking point, the modulation and demodulation, the PID feedback control stabilizes the frequency locking point operation, and the automatic frequency relocking is completed.

[0046] In the step S1, the optical path is built, including:

[0047] Laser drive current setting, temperature PID opening, driving the laser tube to emit light by adjusting the driver current, setting the current value to 110mA, temperature PID opening, at this time, the laser tube emits light normally, the spot range is good, and the effect is as shown in Figure 1 , Figure 1 The Sorebo wavefront detector detects the laser wavefront condition, and finds that the light emitted by the laser can well meet the telltale beam condition.

[0048] The laser drive current is tuned for many times, and the range of 110mA can better emit light and meet the subsequent spectral line requirements. The laser adjustment temperature PID sets an automatic parameter setting function, which meets the requirements of temperature adjustment in dynamic environment scene, Figure 2 the laser adjustment interface is given, Figure 2The working current of the middle drive interface is set to 110 mA, the maximum output current is set to 140 mA, the maximum output current of the laser tube is 180 mA, a 40 mA margin is required, and the laser output power of 47 mw meets the required current and power of the test. The temperature PID drives the PID parameters according to the required environment temperature and opens the temperature control module. The temperature control accuracy is 0.001℃, which basically meets the required temperature setting of the laser tube in the experiment.

[0049] Saturation absorption peak frequency calculation:

[0050] The coupling of the electron spin angular momentum L and the orbital angular momentum S (S = 1 / 2) (LS coupling) produces a fine energy level structure, and the total angular momentum J of the coupled electrons has a value range of |L-S|≤J≤|L+S|, that is, J = |L±1 / 2|, Figure 3 The hyperfine energy level structure of the Rb atom.

[0051] The atomic 5 2 S 1 / 2 →5 2 P 1 / 2 and 5 2 S 1 / 2 →5 2 P 3 / 2 transitions are called D1 line and D2 line respectively, and the numerical value (unit: MHz) between the energy levels is the energy level spacing. The Rb element contains 85 Rb and 87 Rb two isotopes, with natural abundance of 72.15% and 27.85% respectively. The fine energy levels of the ground state and the excited state of the atom are usually denoted as n 2S+1 S J and n 2S+1 P J , where the principal quantum number n determines the number of layers where the valence electron outside the nucleus is located. For Rb atom n = 5, when the ground state L = 0, J g = 1 / 2, corresponding to the fine energy level 5 2 S 1 / 2 ; when the lowest excited state L = 1, J e1 = 1 / 2 and J e2 = 3 / 2, where J e1 = 1 / 2 corresponds to the fine energy level 5 2 P 1 / 2 , and J e2 = 3 / 2 corresponds to the fine energy level 5 2 P 3 / 2 .

[0052] The coupling of the total angular momentum J of the electron and the total angular momentum M of the atomic nucleus (JM coupling) produces a hyperfine energy level structure, and the total angular momentum F of the coupled atom has a value range of |J-M|≤F≤|J+M|. When the ground state is J = 1 / 2, 85Rb has M = 5 / 2, therefore the ground state fine level is 5. 2 S 1 / 2 Split into F g =2 and F g = 3 two hyperfine levels; 87 Rb has M = 3 / 2, therefore the ground state fine level is 5. 2 S 1 / 2 Split into F g =1 and F g =2 hyperfine levels, and similarly, the excited state F can also be derived. e1 and F e2 The ultra-fine energy level.

[0053] The main causes of atomic spectral line broadening are: natural broadening of atoms in the absence of external influences; Doppler broadening caused by random thermal motion; and pressure broadening from collisions between absorbing atoms. Lorentz linear natural broadening is on the order of MHz, pressure broadening in a rarefied atomic cell without buffer gas is negligible, but Gaussian linear Doppler broadening can reach the order of hundreds of MHz. Assume a two-level atom with velocity v moves to the left and is subjected to a force of frequency ω. L The probe light shines to the right.

[0054] Due to the Doppler effect, the atoms sense a frequency. When ω dop The group velocity v of the atoms absorbing the probe light when the atomic resonance frequency ω0 is equal to the atomic resonance frequency. pro for: In the formula, Δ, These represent the detuning and wave vector of the laser, respectively. It can be seen that due to the Doppler effect, the laser light around the resonant frequency is also absorbed by atoms, causing the absorption spectral lines of the atoms to broaden. If a laser with the same frequency ω is introduced into the rubidium chamber... L However, it produces a higher-powered reverse-ejection pump beam. Similarly, from equation... It can be seen that when Δ≠0, the pump light will be deflected by the velocity... When the atomic groups absorb the light, the two beams of light are absorbed by atomic groups with different velocities and do not affect each other. When Δ = 0, both beams of light are absorbed by atomic groups with zero velocity. Atoms become saturated when absorbing the pump light with high power, so they absorb the probe light weakly, resulting in strong transmission of the probe light. This leads to a bulge peak observed in the absorption spectrum, which is called the saturation absorption peak. The corresponding spectral line is called the saturation absorption spectrum.

[0055] The Rb atom is modeled using a multi-level model; here, we will use a three-level model as an example for explanation. Figure 4 As shown in (a), if the resonance frequencies of the three-level atoms are ω1 and ω2, then the atom groups with two velocities will absorb photons.

[0056] When ωL = (ω1+ ω2) / 2, by formula and formula v1= v2, it is known that when the laser scanning frequency is equal to the intermediate frequency of the two hyperfine levels, the atom will also produce a saturated absorption peak when interacting with it, that is, a cross-saturated absorption peak.

[0057] (b) is the saturated absorption spectrum of the D2 line transition of Rb atom, and the saturated absorption peaks are 87 F g = 2→ F e2 = (1, 3) cross line (C 13 ), F g = 2→ F e2 = (2, 3) cross line (C 23 ) and F g = 2→ F e2 = 3 transition line (T3), and 85 F g = 3→ F e2 = (2, 4) cross line (C 24 ), F g = 3→ F e2 = (3, 4) cross line (C 34 ) and F g = 3→ F e2 = 4 transition line (T4).

[0058] The wavelength and frequency of the saturated absorption peak are shown in Table 1.

[0059] Table 1 Frequency and wavelength of saturated absorption peak of Rb atom

[0060]

[0061]

[0062] Stabilization system design;

[0063] The principle of laser frequency stabilization is as follows: a stable frequency is selected as the reference frequency, when the laser frequency deviates from the reference frequency, the error signal representing the deviation is generated through the discrimination deviation, and the error signal is fed back to the laser. The reference frequency is usually the center frequency of the atomic or molecular transition spectrum or the resonance frequency of the optical resonant cavity. The automatic frequency stabilization system takes the center frequency of the atomic saturated absorption spectrum as the frequency reference, and uses the wavelength modulation spectrum method to modulate and demodulate the laser, so as to obtain the frequency discrimination signal required for frequency stabilization. The main components of the automatic frequency stabilization system are shown in Figure 4 It mainly consists of optical part, single-chip microcomputer control loop, modulation and demodulation circuit, proportional integral derivative (PID) feedback circuit, host computer module and laser controller.

[0064] The physical diagram of the light path is shown in Figure 5 The physical diagram of the light path is shown in

[0065] Figure 5 For physical display of the light path, the light path is collimated by a yellow single-mode optical fiber, a small part of light is separated by a polarization beam splitter prism, and the pump light is transmitted to a 0-degree mirror through a rubidium bubble and a quarter-wave plate, and then reflected through the rubidium bubble. The pump light power is about 0.45 mw, and the probe light power is about 46 μw, which can effectively observe the spectral line information. The power ratio of the pump light to the probe light is about 10:1. The probe light is transmitted from the polarization beam splitter prism and focused by a 40 mm focusing lens to a photodetector for photoelectric conversion. The data acquisition card collects and transmits the spectral line information to the host computer for display and subsequent frequency stabilization processing. The spectral line information after triangular wave amplitude frequency optimization is shown in Figure 6 .

[0066] The S2 step of constructing a double matching template includes:

[0067] The spectral line information obtained by the photodetector is optimized to obtain real-time spectral lines. A plurality of cycles are intercepted by a LabView program, saved as an array, and simulated offline to obtain the required template. Since the rubidium bubble is of natural abundance, the spectral line information converted by the photodetector contains all spectral line information of 85 Rb and 87 Rb. Here, the 87 Rb spectral line is taken as the reference spectral line, which contains six basic saturated absorption peaks. The F g =2→F e2 =(2,3) transition frequency is taken as the reference frequency locking point, corresponding to the second saturated absorption peak of the spectral line. Therefore, the six basic saturated absorption peaks are taken as template 1 by array interception, and the F g =2→F e2 =(2,3) corresponding transition frequency is taken as the first template 2. Here, the F g =2→F e2 =(2,3) corresponds to the second saturated absorption peak of the template 1. The double template is made as shown in Figure 7 .

[0068] Figure 7The template creation process is presented. By comparing with the standard curve and optimizing the optical path, spectral information containing six saturated absorption peaks is obtained. Multiple sets of multi-period spectral data are collected and saved from real-time spectra. Offline simulation is used to extract a suitable spectral profile that effectively characterizes the current environment. From this, a set of data is selected, and a spectral segment containing six saturated absorption peaks is extracted as the standard template for the first branch, as shown in Template 1 in the figure. To more effectively locate the saturated absorption peak positions, one of the saturated absorption peaks is arbitrarily selected as Template 2. Considering the red detuning cooling light at the frequency point used in the cold atom gravimeter at F... g =2→F e2 = (2,3) is near the transition frequency, therefore the second saturated absorption peak of template 1 is selected as template 2, such as Figure 8 Template 2 is F g =2→F e2 = (2,3) corresponds to the saturated absorption peak of the spectral line.

[0069] Among them, step S3: starting triangular wave voltage scanning and clock timing, signal acquisition and transmission include;

[0070] A triangular wave scanning module was built, using a signal generator to construct a triangular wave scan. A wavelength meter was used to observe the frequency variation range for real-time simulation experiments. The signal generator was set to a frequency of 1Hz and a voltage amplitude of 1V. At this setting, the spectral information was good, and all six saturated absorption peaks were fully displayed. The saturated absorption peak segment and the signal generator settings interface are shown below. Figure 8 As shown.

[0071] Figure 8 The frequency was set to 1Hz, the voltage amplitude to 0.75V, and the offset and phase to 0. Since the driver requires an impedance of 50Ω, the impedance was adjusted to a low level. The spectral results were good, with sharp peaks, providing valuable spectral information for subsequent experiments.

[0072] To achieve miniaturization, this invention implements triangular wave scanning and parameter optimization through a program. The program is written in LabVIEW, facilitating later programming with DSP or FPGA. For signal acquisition, the triangular wave scanning signal is first sent to the driver's SMA interface, then output using the data acquisition card's AO interface, and finally the PD detector's spectral line signal is acquired via the AI ​​interface. Clock synchronization is set, and the program sequentially executes the output and acquisition operations. The LabVIEW display controls allow for real-time viewing of the spectral line information.

[0073] S4. Start the KCF algorithm, apply template tracking model matching, judge the matching results, find the peak point and calculate the corresponding scanning voltage, including:

[0074] Constructed based on kernel correlation filter tracking algorithm, such as Figure 9 As shown.

[0075] Figure 9 The double-branch KCF algorithm structure diagram is shown. Template 1 and the current cycle are collected by the acquisition card to calculate and update the correlation filter, and the response point area is obtained. After FFT operation, the correlation calculation is performed with the spectrum line information of the next t+1 cycle. The corresponding point position is calculated by inverse FFT operation. The maximum value of the response area is obtained by classification. Branch 2 maps the spectrum line area of branch 1 correlation operation. The mapped spectrum line area is correlated with template 2 to update the correlation filter 2. After FFT operation, the branch obtained by mapping branch 1 at t+1 time is correlated. The response point position is obtained by iFFT operation and sent to the classifier for discrimination operation. When the response values obtained by the discrimination calculation of the two branches are greater than the preset value, it indicates that the locking point is found. At this time, the locking point spectrum line voltage position is recorded, and the scanning voltage value corresponding to the scanning voltage is calculated.

[0076] S5, record the peak point scanning voltage and close the scanning signal, and simultaneously open the sinusoidal signal modulation. At the same time, load the peak value corresponding to the triangular wave scanning voltage to the sinusoidal signal as the offset, and set the frequency of the sinusoidal signal to 10 kHz and the amplitude to 0.01 v.

[0077] S6, the photodetector collects the photoelectric signal. The same frequency sinusoidal signal is sent to the multiplier through the phase shifter and the photoelectric signal collected by the photodetector to obtain the pre-differential error signal, which includes:

[0078] The same frequency signal modulation demodulation, the atomic resonance frequency can be used as a reference frequency source. The feedback system needs to receive an error signal. According to the error signal feedback adjustment, the laser is stably maintained near the resonance frequency point. The modulation demodulation method can remove the differential signal. The error signal is small fluctuation near 0, which is a first-order differential error signal sent to the PID feedback control.

[0079] The modulation method adds a small sinusoidal modulation signal Asin(Ωt) to the laser frequency w0. Ω is the modulation frequency, and A is the modulation amplitude, which is as small as possible. The modulation signal is loaded to the laser driver to modulate the laser.

[0080] In order to obtain the differential signal, the modulation signal is loaded to the laser for uninterrupted sinusoidal modulation. At this time, demodulation is needed to obtain the differential signal.

[0081] S7, the error signal is obtained by a low-pass filter with a set cutoff frequency to obtain the direct current component, which includes:

[0082] In the demodulation process, the same frequency sinusoidal signal is used as a reference signal after phase shifting through the phase shifter and multiplied by the spectrum line signal of the PD detector as a mixed frequency signal. Then, the direct current component is obtained by a low-pass filter. At this time, the direct current component is the differential signal, which fluctuates near 0.

[0083] S8, construct PID feedback, with the direct current component as the differential error signal to the PID feedback, form a closed loop to achieve the required transition frequency point locking;

[0084] S9, determine whether the error signal is greater than the set threshold to determine whether to unlock, including: detection module monitors whether to unlock, once the error signal is greater than 1x10 -3 or PD detection signal is less than 0.1, indicating that the laser is unlocked, at which time the laser triangular wave scanning is restarted, the KCF intelligently determines the frequency locking point, modulates and demodulates, and the PID feedback control stabilizes the frequency locking point operation, completing the automatic frequency relocking.

[0085] In order to overcome the peak locking based on individual typical Lamb dip without considering all saturated absorption peaks; by calculating the differential signal of the specific transition frequency peak as the mode of the frequency locking point, in the dynamic measurement of cold atom gravimeter, the differential signal is easily affected by environmental factors, resulting in strong fluctuation and drift, and the locking point cannot be correctly found, resulting in long "dead time" introduced by repeated unlocking and frequency scanning to find the locking point, which seriously affects the precision of dynamic gravity measurement. The present application provides an automatic frequency stabilization system based on double-core correlation matching intelligent peak searching. Figure 10 and Figure 11 A specific operation method is given, which first builds an easy-to-miniaturize optical path, sets Figure 2 The constant current is 110mA, the maximum current is 140mA, the temperature is set to 21℃, the current environmental temperature control PID parameter is automatically adjusted, the optical path is built to align the pump light and the detection light through the rubidium bubble, the pump light and the detection light power ratio is adjusted until the spectral line information is observed, and each link is optimized to achieve the best effect, Figure 4 and Figure 5 The specific operation of the optical path is given. The triangular wave scanning module is built on the host computer, the triangular wave frequency is adjusted to 1Hz, the amplitude is 750mv, the triangular wave scanning is started, and the data acquisition card is transmitted to the SMA drive. The data acquisition card collects the PD detector signal, and the host computer displays the spectral line information and adjusts the triangular wave scanning phase and bias to the optimum. Figure 9 The specific frequency point scanning voltage searching process is given, and the collected spectral line information is sent to the double-template matching core correlation algorithm intelligent peak searching and records the frequency point scanning voltage. Figure 12 The automatic frequency stabilization method after peak searching is given, the triangular wave scanning is closed after the locking peak is found, and the sinusoidal modulation module is started at the same time, the sinusoidal signal frequency is set to 10kHz, the amplitude is 0.01, the signal is sent to the SMA drive through the acquisition card, and the spectral line voltage is collected. When the PD detector and the same frequency sinusoidal signal are phase shifted through the phase shifter, they are sent into the multiplier to extract the error signal, the error signal is filtered through the low-pass filter to filter out the direct current component, and the error signal is identified through the unlocking identification module. Whether the differential signal is higher than the preset 10 -3and PD detection voltage is less than 0.1mv, judge whether to unlock. If not unlocked, send into PI module feedback fine-tuning frequency point scanning voltage, if unlocked, start triangle wave scanning module again, double template core correlation intelligent peak searching module searches peak and locks point scanning judgment, modulation and demodulation module again to determine whether to unlock, if not unlocked, PI feedback adjusts frequency point current, the system repeatedly executes the cycle process to ensure that the frequency is automatically stable, the frequency stability can reach 10 -12 orders of magnitude, and the unlock locking time is about 4 seconds. In a long-term locking test, the system was not unlocked for 108 hours.

[0086] The above is only the preferred specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can make equivalent replacements or changes according to the technical solution and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. An automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching, characterized in that, It comprises the following steps: S1, saturated absorption spectrum light path and circuit building and debugging optimization; S2, build double matching template; S3, open triangular wave voltage scanning and clock timing, signal acquisition and transmission; S4, start KCF algorithm, apply template tracking model matching, distinguish the results obtained by matching, find the peak point and calculate the corresponding scanning voltage; S5, record the peak scanning voltage and close the scanning signal, and at the same time, open the sinusoidal signal modulation, and set the sinusoidal signal frequency and amplitude; S6, photodetector collects photoelectric signal, and sends the photoelectric signal collected by the photodetector into the multiplier through the phase shifter with the same frequency sinusoidal signal to obtain the pre-differential error signal; S7, the error signal is obtained through the low-pass filter with set cut-off frequency; According to step S1, the light path building comprises: laser driving current setting, temperature PID opening, driving laser tube light emission by adjusting the driver current, setting the current value to 110mA, temperature PID opening, at this time the laser tube emits light normally, the light spot range is good, the laser driving current is adjusted for many times, the range is in the range of 110mA, which can better emit light and meet the subsequent spectral line requirements, the laser adjustment temperature control PID setting automatic parameter setting function meets the requirements of temperature adjustment in dynamic environment scene. S9, determine whether the error signal is greater than the threshold value to determine whether to unlock, including: detection module to monitor whether to unlock, once the error signal is greater than 1 x 10 -3 Or PD detection signal is less than 0.1 indicates that the laser is unlocked, at this time restart the laser triangle wave scanning, KCF intelligent determination of frequency locking point, modulation, PID feedback control stable frequency locking point operation, complete the automatic frequency relocking.

2. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S1, the principle of laser frequency stabilization is as follows: a stable frequency is selected as the reference frequency, when the laser frequency deviates from the reference frequency, the error signal which can represent the deviation is generated by discriminating the deviation, and the error signal is fed back to the laser, the reference frequency usually adopts the center frequency of atomic molecular transition spectrum or the resonance frequency of optical resonant cavity, the automatic frequency stabilization system takes the center frequency of atomic saturated absorption spectrum as the frequency reference, and uses the amplitude-frequency mixed modulation spectrum method to modulate and demodulate the laser, so as to obtain the frequency discrimination signal required for frequency stabilization.

3. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S1, the saturated absorption peak frequency is calculated: the coupling of the electron spin angular momentum L and the orbital angular momentum S (S = 1 / 2) produces a fine energy level structure, the coupling electron total angular momentum J takes the value range of |L-S|≤J≤|L+S|, that is, J = |L±1 / 2|, and the atomic 5 2 S 1 / 2 →5 2 P 1 / 2 and 5 2 S 1 / 2 →5 2 P 3 / 2 transitions are called D1 line and D2 line respectively, the energy level interval is the energy level interval, the Rb element contains 85 Rband 87 RbTwo isotopes, the natural abundance is 72.15% and 27.85% respectively, and the fine energy level of the ground state and the excited state of the atom is usually denoted as n 2S+1 S J and n 2S+1 P J , wherein the principal quantum number n determines the layer number of the valence electron outside the nucleus, for Rb atom n = 5, when the ground state L = 0, J g =1 / 2, corresponding to the fine energy level 5 2 S 1 / 2 , when the lowest excited state L = 1, J e1 =1 / 2 and J e2 =3 / 2, wherein J e1 =1 / 2 corresponds to the fine energy level 5 2 P 1 / 2 , J e2 =3 / 2 corresponds to the fine energy level 5 2 P 3 / 2 , the coupling of the total angular momentum J of the electron and the total angular momentum M of the atomic nucleus produces a hyperfine energy level structure, the total angular momentum F of the atom after coupling takes the value range of |J-M|≤F≤|J+M|, when the ground state J = 1 / 2, 85 M of Rb = 5 / 2, so the ground state fine energy level 5 2 S 1 / 2 is split into F g =2 and F g =3 two hyperfine energy levels, 87 M of Rb = 3 / 2, so the ground state fine energy level 5 2 S 1 / 2 is split into F g =1 and F g =2 two hyperfine energy levels, and the hyperfine energy levels of the excited state F e1 and F e2 can also be obtained.

4. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S3, the triangular wave scanning module is built, the triangular wave scanning is built through the signal generator, and the wavelength meter is used to check the scanning frequency change range, real-time simulation test is carried out, the signal generator is set to 1Hz, and the voltage amplitude is 1V, at this time the spectral line information is good, and 6 saturated absorption peaks can be displayed completely.

5. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S2, the construction of the double matching template includes: the spectral line information obtained by the photodetector is adjusted to obtain real-time spectral lines, a plurality of cycles are intercepted through a LabView program, and are saved as an array, and the required template is obtained through array offline simulation interception. Since the rubidium bubble is a natural abundance rubidium, the spectral line information converted by the photodetector contains 85 all spectral line information of Rb and 87 Rb. 87 Rb spectral line as a reference spectral line, contains 6 basic saturated absorption peaks, and F g = 2 → F e2 = (2, 3) transition frequency as a reference frequency locking point, corresponding to the second saturated absorption peak of the spectral line, so that 6 basic saturated absorption peaks are intercepted as template 1, and F g = 2 → F e2 = (2, 3) corresponding transition frequency is the first template 2, corresponding to the second saturated absorption peak of the template 1.

6. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S5, the sinusoidal signal frequency is set to 10kHz and the amplitude is 0.01v.

7. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, ​ 8. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S6, the same frequency signal modulation demodulation, atomic resonance frequency can be used as a reference frequency source, feedback system needs to receive an error signal, according to the error signal feedback adjustment so that the laser is stable and maintained near the resonance frequency point, modulation demodulation method can remove the differential signal, the error signal is small fluctuation near 0, as a first order differential error signal into the PID feedback control, modulation method by adding a small sinusoidal modulation signal A sin(Ωt) to the laser frequency w0, Ω is the modulation frequency, A is the modulation amplitude, as small as possible, loaded in the laser drive to do modulation signal to the laser with modulation, in order to obtain the differential signal, the modulation signal is loaded to the laser uninterrupted sinusoidal modulation, demodulation to obtain the differential signal.

9. The automatic frequency stabilization system based on dual-core correlation matching intelligent peak searching according to claim 1, characterized in that, According to step S7, during demodulation, the same frequency sinusoidal signal is used as the reference signal after phase shifting through the phase shifter and multiplied by the spectral line signal of the PD detector as the mixed frequency signal, and then the direct current component is taken out through the low pass filter. The direct current component is the differential signal, which fluctuates near 0.

Images