A laser automatic frequency locking method based on similarity recognition spectrum

Abstract

The present application relates to a kind of laser automatic frequency locking methods based on similarity recognition spectrum, comprising: S1, digital PID feedback controller is to laser output sweep voltage;S2, obtain the spectrum signal of laser output under current sweep voltage;S3, whether there is target spectrum in current spectrum signal based on similarity recognition algorithm is determined, if exists, execute step S4;If target spectrum is not identified, adjust sweep voltage and return to execute step S1;S4, based on the locking point of similarity coefficient peak position determination current characteristic spectrum;S5, after obtaining locking point, based on the frequency locking parameter of locking point determination current characteristic spectrum, control digital PID feedback controller is based on the control amount of the frequency locking parameter and carries out PID control locking.This method can achieve the finding of various spectral characteristic peaks and locking point, wide applicability, small environmental interference and method recognition efficiency is high.

Description

Technical Field

[0001] This invention relates to the field of laser frequency locking technology, and in particular to an automatic laser frequency locking method based on similarity recognition spectrum. Background Technology

[0002] Lasers play a vital role in both scientific research and product development. For example, quantum gyroscope systems use Doppler or Raman cooling techniques to cool and trap atoms; by applying precisely timed laser pulses to an atomic beam, the quantum states of atoms can be manipulated to induce interference, and high-precision rotational acceleration values ​​can be calculated from the interference patterns. While commercially available lasers produce monochromatic light, their linewidths typically range from tens to hundreds of megahertz, and their wavelengths exhibit significant long-term drift, which does not meet the demands of quantum precision measurement research and engineering applications. Therefore, it is necessary to lock the laser frequency to a specific point, a technique known as laser frequency locking. The stability of the laser frequency directly determines the accuracy of atomic interferometry measurements.

[0003] The main function of laser frequency locking is to control the frequency of the laser and lock it to a characteristic frequency reference of the spectrum, such as an optical reference cavity or atomic transition frequency, so as to maintain a stable output of the laser frequency.

[0004] In practical applications, laser frequency locking typically employs various locking techniques, such as PHD locking, heterodyne offset locking, dithering locking, saturated absorption spectral locking (SAS), modulation transfer spectral locking (MTS), and frequency modulation spectral locking (FMS). These techniques use feedback control mechanisms to ensure that the laser exhibits a fixed frequency.

[0005] Currently, laser frequency locking methods typically involve manually observing the spectral signal in the optical feedback loop and determining the locking point to set the corresponding locking voltage for the PID controller, thereby locking the laser frequency. However, because laser frequencies are highly sensitive to external environmental factors such as mechanical vibration and thermal fluctuations, these factors can affect the laser's stability, causing it to break out of the locked state. Consequently, laser frequency locking is often unstable, limiting the opportunities for effective signal detection and characterization.

[0006] Therefore, if the laser can automatically lock its frequency and then automatically find the required locking point after unlocking, it will greatly improve the stability and efficiency of the laser system and increase its automation level. In many applications requiring high laser frequency stability, such as long-term continuous unattended observations by high-precision quantum gravimeters at seismic network stations and continuous dynamic absolute gravity mapping on marine vessels, a stable laser frequency can significantly improve work efficiency and the reliability of experimental data.

[0007] Application CN 116345287 A discloses an automatic laser signal frequency locking method, system, and device. This method first performs a frequency sweep on the laser until an error signal is acquired. Then, it outputs a triangular wave voltage signal to drive the laser. After obtaining the error signal curve, it selects the peak of the curve to activate PID control, keeping the laser frequency at zero to ensure the laser output frequency matches the rubidium atom transition frequency. However, this invention's method for finding the locking point is too simplistic and easily affected by interference when searching for signal peaks. For example, the presence of peaks from other spectral lines or noise signals may lead to misjudgment, resulting in a locked laser frequency that does not match the required frequency, still requiring manual confirmation of whether it is locked on the desired spectral line. Furthermore, this invention does not provide an automatic relocking function after the laser is unlocked, making it prone to losing the locked state in environments with interference and vibration, resulting in low reliability.

[0008] Patent CN 109301687 B discloses a laser automatic frequency stabilization system based on intelligent recognition technology of saturated absorption spectrum. The system includes a laser closed-loop locking circuit and an automatic relocking circuit. The closed-loop locking circuit consists of an atomic absorption gas chamber, a frequency discriminator based on the saturated absorption spectrum, and a feedback controller, achieving closed-loop locking of the laser frequency. The automatic relocking circuit combines pattern recognition technology from artificial intelligence with classical optoelectronic technology, enabling accurate and automatic tuning of the laser's operating state and searching for locking points in the saturated absorption spectrum. When the laser frequency is locked, the laser closed-loop locking circuit remains operational to lock the laser frequency; when the laser frequency is unlocked, the relocking circuit relocks the laser frequency through a series of automated operations. This system enables one-click automatic locking of the laser frequency and automatic relocking after unlocking.

[0009] This system achieves automatic laser frequency locking while intelligently determining the locking point, enabling one-click automatic laser frequency locking and automatic relocking after loss of lock. The pattern recognition technology used is essentially based on saturated absorption spectrum recognition using support vector machines. A two-stage classifier categorizes feature vectors in the saturated absorption spectrum into positive and negative values, thus identifying the vector to be locked. However, after identifying the corresponding saturated absorption peak, this method requires a convolution operation. By convolving different absorption peaks with the entire spectrum, the peak position of the result determines the locking point. Therefore, this method cannot achieve completely intelligent laser frequency locking. Furthermore, support vector machines have limitations. When laser frequency locking is required for other spectra such as modulation transfer spectrum (MTS, which generates monotonic error signals without Doppler background, avoiding additional noise from direct modulation of the laser; it has advantages such as high sensitivity, high resolution, and high error signal slope) and frequency modulation spectrum (FMS, which has strong anti-interference capabilities and stable transmission quality, effectively resisting external noise interference and enabling long-distance transmission), this method cannot be directly adapted. Summary of the Invention

[0010] To address the aforementioned problems, the present invention aims to provide a laser automatic frequency locking method based on similarity-based spectral recognition.

[0011] A laser automatic frequency locking method based on similarity-based spectral recognition, the method comprising the following steps:

[0012] S1 controls the digital PID feedback controller to output a sweep voltage to the laser.

[0013] S2, acquire the spectral signal output by the laser at the current sweep voltage;

[0014] S3, perform similarity recognition between the preset feature spectrum and the acquired spectral signal using a similarity algorithm to determine whether the target spectrum exists in the current spectral signal. If it exists, proceed to step S4; if the target spectrum is not identified, adjust the sweep frequency voltage and return to step S1.

[0015] S4, determine the locking point of the current feature spectrum based on the peak position of the similarity coefficient;

[0016] S5. After obtaining the lock point, determine the frequency locking parameters of the current characteristic spectrum based on the lock point, and control the digital PID feedback controller to determine the control quantity based on the frequency locking parameters to perform PID control locking.

[0017] Preferably, step S3 includes:

[0018] The preset feature spectral information is slid across the obtained spectral signal to calculate the similarity, resulting in a set of similarity coefficients with a length equal to the sum of the lengths of the preset feature spectral information and the spectral signal read in step S2.

[0019] The peak value of the similarity coefficient is used to determine whether the target spectrum appears in the current spectral signal.

[0020] Preferably, the step of determining whether the target spectrum appears in the current spectral signal based on the peak value of the similarity coefficient specifically includes:

[0021] The peak value of the similarity coefficient is determined from a set of similarity coefficients, and the peak value is compared with a preset threshold. If the peak value of the similarity coefficient matches the preset threshold, it indicates that the target spectrum exists.

[0022] Preferably, the adjustment of the sweep frequency voltage in step S3 specifically includes:

[0023] For different types of characteristic spectra, a frequency sweep voltage range is preset, and each frequency sweep voltage range is divided into multiple frequency sweep voltage segments in advance.

[0024] If the target spectrum is not identified in the current sweep voltage segment, the automatic step-by-step selects the sweep voltage segment adjacent to the current sweep voltage segment as the sweep voltage.

[0025] Preferably, step S4 specifically includes:

[0026] Determine the spectral information at the sliding position corresponding to the similarity coefficient peak;

[0027] The signal power is obtained from the spectral information and compared with the power information of the currently used feature spectrum. If the difference between the two is within a preset power range, the signal power is considered to match the power of the feature spectrum, and the sliding position corresponding to the current similarity coefficient peak is the locking point.

[0028] Preferably, the method further includes the step of:

[0029] S6. After entering the locked state, determine whether the laser is unlocked based on the currently received spectral signal. If the laser is determined to be unlocked, return to step S1.

[0030] Preferably, step S6 specifically includes:

[0031] If the variance, standard deviation, and maximum value of the received spectral signal voltages in the same group are within a preset range, the laser frequency is considered to be still locked. If they exceed the preset range, the laser is considered to be unlocked.

[0032] Preferably, step S6 further includes:

[0033] Record the spectral signal voltage under the successful frequency locking state. When the laser is determined to be unlocked and return to step S1, determine the sweep voltage of step S1 based on the spectral signal voltage under the successful frequency locking state.

[0034] Preferably, this method uses a state machine to control the execution of its steps.

[0035] Preferably, before executing step S5, it is determined whether the current working mode is automatic locking mode; if so, step S5 is executed.

[0036] This invention, by employing the aforementioned scheme, enables the identification of various spectral characteristic peaks and locking points. Through PID control of the locking point voltage, it achieves a fully intelligent automatic laser frequency locking function, eliminating a significant amount of manual work and meeting the demand for stable laser frequencies in various fields. Furthermore, based on similarity recognition, it automatically identifies the target spectrum and locking point, resulting in higher identification efficiency. It also automatically monitors the status, providing real-time monitoring of the laser frequency locking state. Even under conditions of significant external environmental interference, it promptly detects laser unlocking and performs a second scan and relocking. Recording the position of the last locking point effectively reduces relocking time, increases the total time the laser remains locked, and improves the efficiency of experiments or work. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the laser automatic frequency locking system involved in this application;

[0038] Figure 2 This is a schematic diagram of the structure of the spectral portion involved in this application;

[0039] Figure 3 This is a flowchart of the laser automatic frequency locking method involved in this application. Detailed Implementation

[0040] The embodiments of the present invention are described in detail below.

[0041] Definitions:

[0042] Electro-optic modulator (EOM): An electro-optic modulator is a modulator that utilizes the electro-optic effect of an electro-optic crystal. The electro-optic effect refers to the change in the refractive index of an electro-optic crystal when a voltage is applied, resulting in a change in the optical wave characteristics passing through the crystal, thereby modulating the phase, amplitude, intensity, and polarization state of the optical signal.

[0043] Laser frequency locking: This method uses feedback control technology to compare the laser's output frequency with a reference frequency, and then adjusts the laser's output frequency through feedback control to keep it consistent with the reference frequency.

[0044] Example 1:

[0045] This embodiment provides a laser automatic frequency locking system based on similarity-based spectral recognition. For example... Figure 1 As shown, the laser automatic frequency locking system based on similarity-based spectral recognition includes a laser, a spectral module, a digital PID feedback controller, and an automatic frequency locking operation system. The output of the laser is connected to the input of the spectral module, the output of the spectral module is connected to one input of the digital PID feedback controller, and one output of the digital PID feedback controller is connected to the input of the laser. The digital PID feedback controller interacts with the automatic frequency locking operation system.

[0046] The laser includes a laser source and a laser controller. The laser controller controls the output light frequency signal of the laser source to the spectral module.

[0047] The spectral module modulates and demodulates the optical frequency signal output from the laser, then outputs a spectral signal to the digital PID feedback controller. For example... Figure 2 As shown, the spectral module includes an optical fiber collimator, a first beam splitter, a second beam splitter, and an electro-optic modulator (EOM) and a rubidium gas cell connected in parallel between the first and second beam splitters. The emitting end of the second beam splitter is connected to the modulation and demodulation module via a photodiode. The first output end of the modulation and demodulation module is connected to the photoelectric modulator, and the first output end outputs an RF modulation signal to the photoelectric modulator. The second output end of the modulation and demodulation module is connected to a digital PID feedback controller, and the second output end outputs a spectral signal to the digital PID feedback controller.

[0048] The spectral module handles both the optical path and the circuitry. The optical path processing is as follows: (e.g.) Figure 2 As shown, the spectral module is connected to the laser's output at its fiber collimator. The spectral module guides the laser's emitted light source into free space via the fiber collimator. A first beam splitter splits the light into two paths: one directly enters the rubidium gas chamber, and the other is processed by an electro-optic modulator (EOM). The two paths are then combined in the rubidium gas chamber via a second beam splitter to eliminate Doppler frequency shift. The beams split by the first beam splitter are combined at the second beam splitter, and the combined spectral signal is converted into an electrical signal by a photodiode (PD).

[0049] The circuit section is processed as follows: Figure 2As shown, the spectral module includes a modulation / demodulation module, which comprises an operational amplifier and a demodulation chip. The operational amplifier performs active bandpass filtering and voltage amplification on the crystal oscillator signal of the corresponding frequency. One input terminal of the modulation / demodulation module is an input terminal for an external reference signal. The input reference signal is divided into two paths: the first path is output to the electro-optic modulator as the signal with the required frequency; the second path serves as the reference signal for demodulation by the modulation / demodulation module. Specifically, for the electrical signal input to the modulation / demodulation module from the photodiode, the demodulation chip demodulates the electrical signal based on the second path signal. The demodulated spectral signal is output to a digital PID controller.

[0050] The digital PID feedback controller includes a microcontroller (MCU), a digital-to-analog converter (DAC), and an analog-to-digital converter (ADC). The microcontroller is connected to both the DAC and ADC. The output of the DAC is connected to the laser, and the MCU drives the DAC to lock the laser's frequency. The input of the ADC is connected to the output of the spectral signal, and the microcontroller simultaneously drives the ADC to read the spectral signal at high speed in parallel. After converting the read spectral signal into a digital signal, the data is packaged according to a communication protocol and sent to the automatic frequency locking system. In this embodiment, a 16-bit ADC is used to read the spectral signal. The digital PID controller communicates with the automatic frequency locking system via a Universal Serial Bus (USB), ensuring high-speed and stable signal transmission. Simultaneously, the digital PID feedback controller outputs a sweep frequency signal to the laser based on the currently determined sweep voltage range.

[0051] The automatic frequency locking operation system includes a spectral signal receiving module, a spectral scanning module, a similarity recognition module, and a voltage locking module.

[0052] The spectral signal receiving module receives the spectral signal processed by the digital PID feedback controller. Specifically, as described above, the digital PID controller converts the read spectral signal into a digital signal and packages it according to the communication protocol. It is understood that the automatic frequency locking system includes a data parsing module, used to parse the spectral signal data from the digital PID controller according to the aforementioned communication protocol, and to parse the data for subsequent processing.

[0053] The spectral scanning module is used to control the digital PID feedback controller to output a sweep voltage to the laser; the sweep voltage range corresponds to a preset characteristic spectrum of different types;

[0054] The similarity recognition module is used to perform similarity recognition between the spectral signal read under the current frequency sweep voltage and the preset feature spectral signal to determine whether the target spectrum exists in the spectral signal, and when the target spectrum is detected, to locate the target spectrum and obtain the locking point.

[0055] The voltage locking module is used to analyze the corresponding automatic frequency locking parameters based on the locking point determined by the similarity recognition module, and send the automatic frequency locking parameters to the digital PID feedback controller. The digital PID feedback controller controls the spectral signal voltage of the locking point as the control quantity based on the current frequency locking parameters to perform PID control locking and enter the locking state.

[0056] With the cooperation of the above modules, after obtaining the spectral signal data, the automatic frequency locking system performs a similarity judgment on the signal, automatically identifying which spectral segment needs to be locked, and simultaneously identifying the locking point and automatic frequency locking parameters. In this embodiment, the automatic frequency locking parameters include the sweep voltage of the locking point and the obtained spectral signal voltage, so that the digital PID feedback controller can use the spectral signal voltage of the locking point as a control quantity for PID control locking.

[0057] As a preferred embodiment, the automatic frequency locking operation system further includes a sweep frequency voltage adjustment module, which is used to automatically adjust the sweep frequency voltage range output by the digital PID feedback controller when it is determined that there is no target spectrum under the current sweep frequency voltage.

[0058] In a preferred embodiment, the automatic frequency locking system also includes a monitoring module, which determines whether the system is still in a locked state based on the currently received spectral signal after entering the locked state. The monitoring module is configured to quickly control the spectral scanning module to re-scan the frequency after detecting that the laser has unlocked, and determine the scanning voltage within a preset range based on the voltage at the time of the previous successful locking, allowing for faster entry into the automatic frequency locking state during the re-locking process.

[0059] The aforementioned laser automatic frequency locking system based on similarity-based spectral recognition can automatically modulate the laser's sweep voltage. Even under conditions of laser mode hopping or significant environmental interference, it can still effectively locate characteristic spectra and locking points, exhibiting less susceptibility to environmental interference and more stable operation. Furthermore, due to the use of a similarity recognition module, it can intelligently identify characteristic peaks in various spectra, including saturated absorption spectra, modulation-transfer spectra, and frequency modulation spectra, thus broadening its application range. Moreover, the method of determining locking points based on similarity recognition allows for faster identification and higher work efficiency.

[0060] Example 2:

[0061] This embodiment provides a laser automatic frequency locking method based on similarity recognition spectrum, which operates in the laser automatic frequency locking system of Embodiment 1. Combined with... Figure 3 As shown, the method includes the following steps:

[0062] S1 controls the digital PID feedback controller to output a sweep voltage to the laser; the current state is the scanning state.

[0063] S2, acquire the spectral signal output by the laser at the current sweep voltage;

[0064] S3, perform similarity recognition between the preset feature spectrum and the acquired spectral signal using a similarity algorithm to determine whether the target spectrum exists in the current spectral signal. If it exists, proceed to step S4; if the target spectrum is not identified, adjust the sweep frequency voltage and return to step S1.

[0065] S4, determine the locking point of the current feature spectrum based on the peak position of the similarity coefficient;

[0066] S5. After obtaining the lock point, determine the frequency locking parameters of the current characteristic spectrum based on the lock point, and control the digital PID feedback controller to determine the control quantity based on the frequency locking parameters to perform PID control locking. The current state is the locked state.

[0067] The laser automatic frequency locking method based on similarity-based spectral recognition achieves automatic frequency locking control of the laser through the steps described above. By using a similarity recognition algorithm to identify characteristic spectra and locking points, it can intelligently identify characteristic peaks of various spectra, including saturated absorption spectra, modulation-transfer spectra, and frequency modulation spectra, thus broadening its application range. Furthermore, it can automatically identify the required locking point location; the locking point identification method is simple and highly efficient. In addition, the method can automatically adjust the laser's frequency sweep range, demonstrating strong adaptability.

[0068] In step S1 above, after receiving the sweep voltage, the laser can output a corresponding optical frequency signal. The optical frequency signal output by the laser is then processed as follows: Figure 1 The spectral portion shown, after processing by the digital PID feedback controller, yields a spectral signal that the automatic frequency-locking operation system can recognize and process. The sweep voltage output by the laser in S1 corresponds to different types of characteristic spectra, such as saturated absorption spectra, modulation transfer spectra, and frequency modulation spectra, with preset sweep voltage ranges expected to be obtained for each type of characteristic spectrum. Preferably, in this embodiment, the scanning state is monitored. If an error occurs during the laser's output sweep voltage, step S2 is not executed, and a scanning fault message is output. Monitoring the scanning state specifically involves monitoring preset parameters during the laser's output sweep voltage process. The specific implementation method is a conventional technique in the art and will not be described in detail here.

[0069] Furthermore, those skilled in the art will understand that before performing the above step S1, it is necessary to initialize the parameters and set necessary parameters such as PID control coefficient, sweep voltage amplitude, and sweep voltage range.

[0070] In step S2 above, the spectral signal corresponding to the optical frequency signal output by the laser after receiving the sweep voltage is read. This signal is a voltage signal. The error signal can be determined by combining the spectral signal and then controlled by feedback.

[0071] Step S3 above, specifically adjusting the sweep frequency voltage, includes:

[0072] For different types of characteristic spectra, a pre-set sweep voltage range is established. Each sweep voltage range is pre-divided into multiple sweep voltage segments, for example, the first segment is 0.1V to 0.3V, then after a step, it becomes 0.2V to 0.4V, and so on. If the target spectrum is not detected in the current sweep voltage segment, the sweep voltage is automatically stepped to the next adjacent segment and step S1 is re-executed. This automatic setting allows for automatic adjustment of the sweep voltage without manual intervention, reducing human involvement and improving efficiency. If the target spectrum is still not detected after using all sweep voltage segments within the current characteristic spectrum's sweep voltage range, it indicates an error in the pre-set sweep voltage range for that spectrum, requiring manual adjustment. For example, if the target spectrum is not found at 0.1V to 0.3V, the sweep voltage is stepped to 0.2V to 0.4V for sweeping.

[0073] Step S3 specifically includes sliding the preset feature spectral information across the obtained spectral signal to calculate similarity, resulting in a set of similarity coefficients whose length is the sum of the lengths of the two information segments (the preset feature spectral information and the spectral signal read in step S2). In this embodiment, the similarity coefficients and their corresponding sliding positions can be stored in an array, and the spectral information at the corresponding position of the spectral signal can be determined based on the sliding positions.

[0074] The presence of a target spectrum in the current spectral signal is determined based on the peak value of the similarity coefficient. Specifically, when the target spectrum appears in the obtained spectrum, the similarity coefficient will show a peak value at the corresponding sliding position. It is then determined whether this similarity coefficient peak matches a preset threshold to indicate the presence of the target spectrum. If the peak value matches the preset threshold, the target spectrum is determined to exist; otherwise, the current similarity coefficient peak cannot indicate the presence of the target spectrum. The preset threshold corresponds one-to-one with the characteristic spectral type, and the threshold can be a point value or a range value. In this embodiment, a specific matching rule is preset to determine whether the peak value of the similarity coefficient matches the preset threshold. As one possible implementation, the preset threshold is a range value; if the peak value of the similarity coefficient falls within the range value, it is determined to match the preset threshold. As another possible implementation, the threshold is a point value; the specific matching rule is that if the peak value of the similarity coefficient is within 10% before or after the threshold, it is determined to match the preset threshold.

[0075] In step S4 above, determining the lock point of the current feature spectrum based on the peak position of the similarity coefficient specifically involves, after confirming the existence of the target spectrum, using the spectral information at the sliding position corresponding to the similarity coefficient peak on the spectral signal as the lock point of the current feature spectrum. Further, it is necessary to determine whether the current sliding position is a lock point by combining the signal power of the spectral signal. Specifically, the signal power of the spectral signal at the sliding position corresponding to the similarity coefficient peak (similarity coefficient peak position, target position) is obtained and compared with the power information of the currently used feature spectrum. If the difference between the two is within a preset power range, the signal power is considered to match the power of the feature spectrum, and the current position is considered a lock point.

[0076] In this embodiment, the similarity algorithm can employ cosine similarity, Pearson correlation coefficient, or other commonly used similarity algorithms. For example, when using the Pearson correlation coefficient, for a feature spectrum X and the obtained spectral signal Y; the feature spectrum X is slid across the spectral line of the obtained spectral signal Y, and the length of the feature spectrum X is taken. For each point: the product of the difference between X and the average of X and the difference between Y and the average of Y is summed, divided by the square of the difference between X and the average of X, and the summation at each point is squared. Then, the product is divided by the square of the difference between Y and the average of Y, and the summation at each point is squared. The resulting similarity coefficient is obtained; the formula is as follows. Signal power is calculated as follows: for a spectral signal Y, the square of the difference between Y and its average value is summed at each point; the formula is: The above calculations can achieve step S3.

[0077] The above step S5 specifically includes, after obtaining the required locking point, the automatic frequency locking operation system sends the automatic frequency locking parameters, including the scanning voltage of the locking point and the obtained spectral signal voltage, to the digital PID feedback controller, and uses the spectral signal voltage of the locking point as the control quantity for PID control locking.

[0078] Preferably, the method further includes the step: S6, after entering the locked state, determining whether it is still in the locked state based on the currently received spectral signal; the current state is the monitoring state. In this embodiment, the spectral signal is parsed into a voltage signal, and the laser frequency is determined to be in the locked state based on the voltage value of the spectral signal. Specifically, if the variance, standard deviation, and maximum value of a set of spectral signal voltages are within a preset range, the laser frequency is considered to be in the locked state; if they exceed the preset range, the laser is considered to be unlocked; if the laser is unlocked, the process returns to step S1. This step enables intelligent monitoring of the locked state and timely relocking after unlocking.

[0079] Preferably, step S6 further includes recording the spectral signal voltage in the successfully locked frequency state. When the laser is determined to have lost lock and returns to step S1, the sweep voltage is determined based on the spectral signal voltage in the successfully locked frequency state. In a specific embodiment, when returning to step S1 to rescan, the sweep voltage is updated to a certain positive and negative range of the spectral signal voltage at the time of the last successful automatic frequency locking, so as to enter the locked frequency state more quickly during the relocking process. Step S6 enables the automatic relocking function after loss of lock, making the laser's locked frequency state more stable.

[0080] Preferably, this method is executed by state machine control, which can achieve fully automatic frequency locking, saving a lot of manpower and improving the stability when the laser needs to work at a specific frequency.

[0081] Preferably, the system includes an automatic locking mode and a waiting-to-lock mode. The automatic locking mode controls the automatic frequency locking and re-locking of the laser after unlocking. In automatic locking mode, the complete method of steps S1-S6 is executed. In waiting-to-lock mode, steps S5-S6 are not executed; whether to execute steps S1-S4 can be determined according to actual needs. In this embodiment, in waiting-to-lock mode, the automatic frequency locking system only performs a voltage scan within the current range without frequency locking, facilitating spectrum observation. After initiating automatic frequency locking, it jumps back to step S1 to restart the automatic frequency locking process. Figure 3As shown, the method includes determining whether the current state is in automatic locking mode before executing step S5. If so, step S5 is executed; otherwise, step S5 is not executed. In the monitoring state, if a control signal is received to switch the current automatic locking mode to a waiting-to-lock mode, the digital PID feedback controller is controlled to disengage from the locking control. In a specific embodiment, the switching between automatic locking mode and waiting-to-lock mode is performed through a working mode control signal. This method achieves fully automatic frequency locking through the state machine control in the automatic frequency locking system, saving significant manpower and improving stability in scenarios requiring the laser to operate at specific frequencies. Through a similarity recognition algorithm, the required locking point for the characteristic spectral lines is intelligently determined, and the similarity recognition can identify various spectra, making it widely applicable. This method, by automatically modulating the laser's sweep voltage, can effectively locate characteristic spectra and lock points even under conditions of laser mode hopping or significant environmental interference. It intelligently monitors the lock status and promptly relocks after delocking. The invention utilizes a similarity recognition function to intelligently identify characteristic peaks in various spectra, including saturated absorption spectra, modulation transfer spectra, and frequency modulation spectra, thus having a wide range of applications. It can directly identify the required lock point location and quickly lock it. It intelligently adjusts the laser's sweep range, offering strong adaptability. It achieves a fully intelligent automatic laser frequency locking function, eliminating a significant amount of manual work and meeting the demand for stable laser frequencies in various fields. It automatically records the position of the last locked point, effectively reducing relocking time, increasing the total time the laser is in a locked state, and improving work efficiency.

[0082] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A laser automatic frequency locking method based on similarity-based spectral recognition, characterized in that, The method includes the following steps: S1 controls the digital PID feedback controller to output a sweep voltage to the laser. S2, acquire the spectral signal output by the laser at the current sweep voltage; S3, perform similarity recognition between the preset feature spectrum and the acquired spectral signal using a similarity algorithm to determine whether the target spectrum exists in the current spectral signal. If it exists, proceed to step S4; if the target spectrum is not identified, adjust the sweep frequency voltage and return to step S1. S4, determine the locking point of the current feature spectrum based on the peak position of the similarity coefficient; S5, after obtaining the lock point, determine the frequency locking parameters of the current characteristic spectrum based on the lock point, and control the digital PID feedback controller to determine the control quantity based on the frequency locking parameters to perform PID control locking; Step S3 includes: The preset feature spectral information is slid across the obtained spectral signal to calculate the similarity, resulting in a set of similarity coefficients with a length equal to the sum of the lengths of the preset feature spectral information and the spectral signal read in step S2. The peak value of the similarity coefficient is used to determine whether the target spectrum appears in the current spectral signal; The method of determining whether the target spectrum appears in the current spectral signal based on the peak value of the similarity coefficient specifically includes: The peak value of the similarity coefficient is determined from a set of similarity coefficients, and the peak value is compared with a preset threshold. If the peak value of the similarity coefficient matches the preset threshold, it indicates that the target spectrum exists. Step S4 specifically includes: Determine the spectral information at the sliding position corresponding to the similarity coefficient peak; The signal power is obtained from the spectral information and compared with the power information of the currently used feature spectrum. If the difference between the two is within a preset power range, the signal power is considered to match the power of the feature spectrum, and the sliding position corresponding to the current similarity coefficient peak is the locking point.

2. The laser automatic frequency locking method based on similarity recognition spectrum according to claim 1, characterized in that, The adjustment of the sweep frequency voltage in step S3 specifically includes: For different types of characteristic spectra, a frequency sweep voltage range is preset, and each frequency sweep voltage range is divided into multiple frequency sweep voltage segments in advance. If the target spectrum is not identified in the current sweep voltage segment, the automatic step-by-step selects the sweep voltage segment adjacent to the current sweep voltage segment as the sweep voltage.

3. The laser automatic frequency locking method based on similarity recognition spectrum according to claim 1, characterized in that, The method further includes the following steps: S6. After entering the locked state, determine whether the laser is unlocked based on the currently received spectral signal. If the laser is determined to be unlocked, return to step S1.

4. The laser automatic frequency locking method based on similarity recognition spectrum according to claim 3, characterized in that, Step S6 specifically includes: If the variance, standard deviation, and maximum value of the received spectral signal voltages in the same group are within a preset range, the laser frequency is considered to be still locked. If they exceed the preset range, the laser is considered to be unlocked.

5. The laser automatic frequency locking method based on similarity recognition spectrum according to claim 3, characterized in that, Step S6 also includes: Record the spectral signal voltage under the successful frequency locking state. When the laser is determined to be unlocked and return to step S1, determine the sweep voltage of step S1 based on the spectral signal voltage under the successful frequency locking state.

6. The laser automatic frequency locking method based on similarity recognition spectrum according to claim 1, characterized in that, This method uses a state machine to control the execution of its steps.

7. The laser automatic frequency locking method based on similarity recognition spectrum according to claim 1, characterized in that, Before executing step S5, determine whether the current working mode is automatic locking mode. If so, then execute step S5.

Images