Method for monitoring laser power instability

Abstract

This application discloses a method for monitoring laser power instability, relating to the field of laser state monitoring. The method includes: first, performing a fast Fourier transform on the power sequence of the laser output to obtain a spectrum; then calculating the power spectral density at each frequency point k in the spectrum to obtain a power spectral density map; next, scanning the target frequency band in the power spectral density map to extract target frequency peaks with peak values ​​greater than a peak threshold and their characteristic parameters; finally, determining the disturbance source causing laser power instability based on the center frequency of the target frequency peak, thus achieving the identification of the disturbance source causing laser power instability.

Description

Technical Field

[0001] This application relates to the field of laser status monitoring technology, and in particular to a method for monitoring laser power instability. Background Technology

[0002] During laser processing, the stability of the laser's output power directly affects the processing quality. In actual operation, various disturbance sources (such as power supply ripple, switching noise, mechanical resonance, temperature fluctuations, etc.) can couple into the laser power, causing power fluctuations. Among these, high-frequency power disturbances caused by power supply ripple are particularly common and difficult to suppress using conventional methods.

[0003] Existing laser power monitoring methods typically perform time-domain statistical analysis on the acquired power time series (e.g., calculating the average, standard deviation, peak-to-peak value, etc.) to output an overall amplitude index of power fluctuations. When power instability is detected, operators can only know that "power fluctuations exist," but cannot determine whether the fluctuations originate from specific disturbance sources such as power supply ripple, switching power supply noise, or mechanical vibration. Since different disturbance sources have different physical mechanisms, their suppression measures are also completely different (e.g., power supply ripple requires adjusting power supply filter parameters, while mechanical resonance requires changing structural damping). Monitoring methods lacking the ability to identify disturbance sources are difficult to guide on-site troubleshooting and process optimization. Summary of the Invention

[0004] The main objective of this application is to provide a method for monitoring laser power instability, which aims to solve the technical problem that existing technologies cannot identify sources of disturbance for laser power instability.

[0005] To achieve the above objectives, this application proposes a method for monitoring laser power instability, comprising: Perform a Fast Fourier Transform on the power sequence P[t] of the laser output to obtain the spectrum F[k], and calculate the power spectral density PSD at each frequency point k in the spectrum. k To obtain the power spectral density map PSD(f); The target frequency band in the power spectral density map PSD(f) is scanned, and the target frequency peaks with peak values ​​greater than the peak value threshold and their characteristic parameters are extracted. The characteristic parameters include the center frequency f. peak ; Based on the center frequency f of the target frequency peak peak The source of disturbance that causes laser power instability was identified.

[0006] In one embodiment, the characteristic parameter further includes amplitude PSD. peak and bandwidth (BW); The method further includes: based on the amplitude PSD of the target frequency peak. peakGiven the bandwidth BW, the effective value P of the power fluctuation caused by the target frequency peak is calculated. rms This is to quantify the disturbance of laser power by the corresponding interference source.

[0007] In one embodiment, the method further includes: selecting a frequency band without significant features in the power spectral density map, and calculating the peak threshold based on the 3σ threshold method.

[0008] In one embodiment, the step of determining the center frequency f of the target frequency peak... peak The disturbance sources that caused the laser power instability were identified as including: If the center frequency of the target frequency peak is f peak If the frequency is 50Hz or 100Hz, the interference source corresponds to power supply ripple. If the center frequency of the target frequency peak is f peak If the frequency is 1kHz, 5kHz, or 10kHz, the interference source corresponds to switching power supply noise or PWM modulation frequency. If the center frequency of the target frequency peak is f peak If the frequency is greater than 100Hz but less than 1000Hz, the source of interference is mechanical resonance.

[0009] In one embodiment, the method further includes: acquiring the power sequence P[t] of the laser output from the laser, and calculating the average value P within a preset time window. avg Standard deviation σ, peak-to-peak value P pp and volatility coefficient CV; Based on the average value P within the preset time window avg Standard deviation σ, peak-to-peak value P pp Fluctuation coefficient CV and effective value of power fluctuation P rms The laser power instability I is calculated using the corresponding weighting coefficients. index .

[0010] In one embodiment, the method further includes: adjusting the corresponding weighting coefficients according to process constraints. If the process constraints include average power consistency constraints, then increase the weighting factor of the fluctuation coefficient CV. If the process constraints include welding thickness accuracy constraints, then the peak-to-peak value P is increased. pp Weighting coefficients; If the process constraints include periodic stripe defect constraints, then increase the effective value P of the power fluctuation corresponding to the power ripple. rms The weighting coefficients.

[0011] In one embodiment, it further includes: The power sequence of the laser output is collected using a laser power generator; The laser power unit includes an optical signal acquisition module and a signal conversion module. The optical signal acquisition module is used to acquire the optical signal of the laser output by the laser, and the signal conversion module is used to convert the electrical signal into a power signal to obtain the power sequence of the output laser.

[0012] In one embodiment, the optical signal acquisition module includes an optical homogenizer and a photodetector; The light homogenizer includes: A substrate, wherein an optical homogenization cavity is disposed inside the substrate, and an observation port communicating with the optical homogenization cavity is disposed on the substrate; When the optical fiber passes through the optical homogenization cavity, the reflecting surface and the observation port of the optical homogenization cavity are located on both sides of the optical fiber, and the laser emitted from the cladding region on the side of the optical fiber facing the reflecting surface is directed towards the observation port after passing through the reflecting surface, while the laser emitted from the cladding region on the side of the optical fiber facing the observation port is directed directly towards the observation port.

[0013] In one embodiment, the reflecting surface is an arc-shaped diffuse reflecting surface. The laser light emanating from the cladding region on one side of the optical fiber is directed toward the arc-shaped diffuse reflecting surface, and after being scattered and homogenized, it is directed toward the observation port.

[0014] One or more technical solutions proposed in this application have at least the following technical effects: A method for monitoring laser power instability is proposed. First, a fast Fourier transform is performed on the power sequence of the laser output to obtain the spectrum. The power spectral density at each frequency point k in the spectrum is calculated to obtain a power spectral density map. Then, a fast Fourier transform is performed on the power sequence of the laser output to obtain the spectrum. The power spectral density at each frequency point in the spectrum is calculated to obtain a power spectral density map. Finally, based on the center frequency of the target frequency peak, the disturbance source causing laser power instability is identified, thus realizing the identification of the disturbance source causing laser power instability. Attached Figure Description

[0015] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the first embodiment of the laser power instability monitoring method. Figure 2This is an example of a power spectral density plot; Figure 3 This is a schematic diagram of the three-dimensional structure of the optical homogenizer. Figure 4 This is a schematic diagram of the planar structure of an optical homogenizer; Explanation of icon numbers: 10. Substrate; 20. Light homogenization cavity; 21. Reflecting surface; 30. Observation port; 41. Slot 1; 42. Slot 2. The realization of the purpose of this application, its functional characteristics and advantages will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application. To better understand the technical solutions of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0019] This application provides a method for monitoring laser power instability. In a first embodiment of the laser power instability monitoring method of this application, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the laser power instability monitoring method of this application. The laser power instability monitoring method may include steps S20 to S40: Step S20: Perform a Fast Fourier Transform on the power sequence P[t] of the laser output to obtain the spectrum F[k], and calculate the power spectral density PSD at each frequency point k in the spectrum. k To obtain the power spectral density map PSD(f).

[0020] It should be noted that the power sequence P[t] of the laser output can be acquired using a power meter, including but not limited to an integrating sphere power meter or a thermally stacked laser power meter.

[0021] In one feasible implementation, to reduce frequency leakage, the power sequence P[t] is windowed to obtain P w [t]; for the windowed power sequence P w Perform a Fast Fourier Transform on each point of [t] to obtain the spectrum F[k] of the complex number.

[0022] In one feasible implementation, the power density spectrum (PSD) is calculated. k The mathematical expression is:

[0023] The DC component is:

[0024] In the formula, |F[k]| represents the amplitude at the k-th point in the spectrum, where k=1,2,…,N / 2-1, N represents the number of sample points in the power sequence, fs represents the sampling frequency, and S represents the equivalent noise bandwidth of the window function.

[0025] Step S30: Scan the target frequency band in the power spectral density map PSD(f) and extract the target frequency peaks with peak values ​​greater than the peak value threshold and their characteristic parameters. The characteristic parameters include the center frequency f. peak .

[0026] It should be noted that the target frequency band is the frequency band of interest to the user. In this embodiment, the target frequency band is configured as (0~5kHz). The PSD is scanned within the target frequency band, and all continuous regions are searched. Frequency peaks with peak values ​​greater than a peak threshold are recorded as target frequency peaks. The characteristic parameters of all target frequency peaks are also recorded. These characteristic parameters may include the center frequency f. peak Amplitude PSD peak One or more of bandwidth BW.

[0027] In one feasible implementation, the monitoring method further includes the following steps: The peak threshold is calculated by selecting frequency bands with no significant characteristics in the power spectral density plot and using the 3σ threshold method.

[0028] It should be noted that the peak threshold is set based on the background noise level, and in this embodiment, the peak threshold is calculated using the 3σ thresholding method. Specifically, the mean μ of the frequency band without significant characteristics (e.g., 3kHz~4.5kHz) is calculated. noise and standard deviation σ noise Threshold PSD Configured as μ noise and 3σ noise The sum of.

[0029] Step S40, based on the center frequency f of the target frequency peak peak The source of disturbance that causes laser power instability was identified.

[0030] In one feasible implementation, step S40 may include the following steps: If the center frequency of the target frequency peak is f peak If the frequency is 50Hz or 100Hz, the interference source corresponds to power supply ripple. If the center frequency of the target frequency peak is f peak If the frequency is 1kHz, 5kHz, or 10kHz, the interference source corresponds to switching power supply noise or PWM modulation frequency. If the center frequency of the target frequency peak is f peak If the frequency is greater than 100Hz but less than 1000Hz, the source of interference is mechanical resonance.

[0031] The laser power instability monitoring method in this embodiment first performs a fast Fourier transform on the power sequence P[t] of the laser output to obtain the spectrum F[k], and then calculates the power spectral density PSD at each frequency point k in the spectrum. k To obtain the power spectral density map PSD(f), a fast Fourier transform is then performed on the power sequence P[t] of the laser output to obtain the spectrum F[k]. The power spectral density PSD at each frequency point k in the spectrum is then calculated. k To obtain the power spectral density map PSD(f); finally, based on the center frequency f of the target frequency peak... peak The disturbance source causing laser power instability was identified, thus achieving the identification of the disturbance source causing laser power instability.

[0032] In one feasible implementation, the characteristic parameter may include the center frequency f. peak Amplitude PSD peak Following step S40, and with bandwidth BW, the monitoring method also includes step S50: Step S50, based on the amplitude PSD of the target frequency peak peak Given the bandwidth BW, the effective value P of the power fluctuation caused by the target frequency peak is calculated. rms This is to quantify the disturbance of laser power by the corresponding interference source.

[0033] In one feasible implementation, the mathematical expression for calculating the effective value of power fluctuation is:

[0034] In the formula, P rms，est This represents the effective value of the power fluctuation caused by the target frequency peak.

[0035] In this embodiment, the interference of the interference source on the laser power can be quantified by the effective value of the power fluctuation.

[0036] The laser power instability monitoring method in this embodiment identifies the disturbance source that causes laser power instability and quantifies the degree of interference, so that users can understand the degree of interference based on the value, which is helpful for subsequent equipment adjustment and workflow control.

[0037] In one feasible implementation, after step S50, the monitoring method further includes step S60: Step S60: Obtain the power sequence P[t] of the laser output, and calculate the average value P within a preset time window. avg Standard deviation σ, peak-to-peak value P pp And the volatility coefficient CV; based on the average value P within a preset time window. avg Standard deviation σ, peak-to-peak value P ppFluctuation coefficient CV and effective value of power fluctuation P rms The laser power instability I is calculated using the corresponding weighting coefficients. index .

[0038] In this embodiment, the mathematical expression for calculating the volatility coefficient CV is:

[0039] Laser power instability I index The mathematical expression for the calculation is:

[0040] In the formula, α, β, and γ are weighting coefficients, and CV ref P represents the volatility threshold. pp,ref P represents the peak-to-peak threshold. rms，ref This represents the effective value threshold for power fluctuations.

[0041] The laser power instability monitoring method in this embodiment calculates the average value P within a time window in the time domain. avg Standard deviation σ, peak-to-peak value P pp The fluctuation coefficient CV is calculated, and the effective value of power fluctuation is obtained in the frequency domain. Then, based on the above parameters, a quantized value for quantifying laser power instability is calculated, so that users can understand the power instability of the laser output laser based on this value.

[0042] In one feasible implementation, after step S60, the monitoring method further includes step S70: Step S70: Adjust the corresponding weighting coefficients according to process constraints. If the process constraints include average power consistency constraints, then increase the weighting factor of the fluctuation coefficient CV, i.e., α. If the process constraints include welding thickness accuracy constraints, then the peak-to-peak value P is increased. pp The weighting coefficient, β; If the process constraints include periodic stripe defect constraints, then increase the effective value P of the power fluctuation corresponding to the power ripple. rms The weighting coefficient, i.e., the corresponding γ.

[0043] Specifically, process constraints are input by the user. For example, if the welding process is extremely sensitive to the consistency of average power, then an average power consistency constraint is input. If the welding process has strict control over the welding thickness, and extreme power peaks can lead to over-melting or incomplete melting, then a welding thickness accuracy constraint is input. If it is known that power ripple will cause molten pool oscillation to form periodic stripe defects, then a periodic stripe defect constraint is input, and the weighting coefficient corresponding to the effective value of power fluctuation at the 100Hz frequency is increased.

[0044] In one feasible implementation, prior to step S20, the monitoring method further includes step S10: Step S10: Collect the power sequence of the laser output by the laser using a laser power device.

[0045] Optionally, the laser power unit includes an optical signal acquisition module and a signal conversion module. The optical signal acquisition module is used to acquire the optical signal of the laser output by the laser, and the signal conversion module is used to convert the electrical signal into a power signal to obtain the power sequence of the output laser.

[0046] In one feasible implementation, refer to Figure 3 and Figure 4 The optical signal acquisition module includes an optical homogenizer and a photodetector; The substrate 10 has an optical homogenization cavity 20 disposed inside it, and an observation port 30 communicating with the optical homogenization cavity 20 is disposed on the substrate 10. When the optical fiber passes through the optical homogenization cavity 20, the reflecting surface 21 and the observation port 30 of the optical homogenization cavity 20 are located on both sides of the optical fiber, and the laser emitted from the cladding region on the side of the optical fiber toward the reflecting surface 21 is directed toward the observation port 30 after passing through the reflecting surface 21, and the laser emitted from the cladding region on the side of the optical fiber toward the observation port 30 is directed directly toward the observation port 30.

[0047] It should be noted that the optical homogenization cavity 20 is used to direct the laser emitted from the cladding region of the optical fiber toward the reflecting surface 21 toward the observation port 30 through specular reflection or diffuse reflection.

[0048] In one feasible implementation, refer to Figure 4 The diagram shows a planar structure of an optical homogenizer. An optical fiber horizontally passes through an optical homogenization cavity 20. The cavity 20 is divided into an upper and lower half based on the fiber's location. A reflecting surface 21 is positioned in the upper half, and an observation port 30 is located in the lower half. During laser transmission through the optical fiber, laser light emanating from the upper cladding passes through the reflecting surface 21 and then strikes the observation port 30. Laser light emanating from the lower cladding directly strikes the observation port 30. It should be noted that the terms "upper cladding" and "lower cladding" in this embodiment are only approximate descriptions.

[0049] Preferably, the reflecting surface 21 is an arc-shaped diffuse reflecting surface 21. The laser light emanating from the cladding region on one side of the optical fiber toward the reflecting surface 21 is directed toward the arc-shaped diffuse reflecting surface 21, and after being scattered and homogenized, it is directed toward the observation port 30.

[0050] It should be noted that in output laser power sampling technologies, direct-receiving power meters are typically used for sampling. For example, the output laser beam is split and directly incident on a photodiode (PD). However, this method requires pausing the welding operation, significantly impacting production efficiency. Existing technologies include non-invasive detection using an integrating sphere. The principle is to use the integrating sphere to reflect the laser emanating from the periphery of the optical fiber through multiple diffuse reflections to the power meter, and then calculate the output laser power using a calibrated mapping model. However, except in cases of poor fiber quality, the energy of the laser emanating from the fiber is extremely low, and after multiple diffuse reflections, the energy attenuates significantly. Therefore, ultra-high sensitivity photodiodes are required for accurate signal acquisition.

[0051] It is understandable that the optical signal acquisition module in this embodiment can reflect a portion of the laser light escaping from the fiber cladding through the optical homogenization cavity before entering the photodetector, while the remaining portion directly enters the photodetector. By performing diffuse reflection homogenization on only a portion of the light, energy loss and scattering noise are significantly reduced, ensuring that the laser has sufficient intensity to be detected by the photodetector. This achieves high-precision acquisition of optical signals under weak signal conditions, thereby improving the accuracy of results such as the identification of interference sources related to laser power instability mentioned above. It is worth noting that this homogenizer can be directly placed on the output optical path of the laser generator module, allowing for real-time acquisition of the laser output power sequence during laser operation. Furthermore, it is small in size and does not occupy excessive space.

[0052] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for monitoring laser power instability, characterized in that, include: Perform a Fast Fourier Transform on the power sequence P[t] of the laser output to obtain the spectrum F[k], and calculate the power spectral density PSD at each frequency point k in the spectrum. k To obtain the power spectral density map PSD(f); The target frequency band in the power spectral density map PSD(f) is scanned, and the target frequency peaks with peak values ​​greater than the peak value threshold and their characteristic parameters are extracted. The characteristic parameters include the center frequency f. peak ; Based on the center frequency f of the target frequency peak peak The source of disturbance that causes laser power instability was identified.

2. The laser power instability monitoring method as described in claim 1, characterized in that, The characteristic parameters also include amplitude PSD. peak and bandwidth (BW); The method further includes: Based on the amplitude PSD of the target frequency peak peak Given the bandwidth BW, the effective value P of the power fluctuation caused by the target frequency peak is calculated. rms This is to quantify the disturbance of laser power by the corresponding interference source.

3. The laser power instability monitoring method as described in claim 2, characterized in that, Also includes: The peak threshold is calculated by selecting frequency bands with no significant characteristics in the power spectral density plot and using the 3σ threshold method.

4. The laser power instability monitoring method as described in claim 2, characterized in that, The center frequency f of the target frequency peak peak The disturbance sources that caused the laser power instability were identified as including: If the center frequency of the target frequency peak is f peak If the frequency is 50Hz or 100Hz, the interference source corresponds to power supply ripple. If the center frequency of the target frequency peak is f peak If the frequency is 1kHz, 5kHz, or 10kHz, the interference source corresponds to switching power supply noise or PWM modulation frequency. If the center frequency of the target frequency peak is f peak If the frequency is greater than 100Hz but less than 1000Hz, the source of interference is mechanical resonance.

5. The laser power instability monitoring method as described in claim 2, characterized in that, Also includes: Obtain the power sequence P[t] of the laser output, and calculate the average value P within a preset time window. avg Standard deviation σ, peak-to-peak value P pp and volatility coefficient CV; Based on the average value P within the preset time window avg Standard deviation σ, peak-to-peak value P pp Fluctuation coefficient CV and effective value of power fluctuation P rms The laser power instability I is calculated using the corresponding weighting coefficients. index .

6. The laser power instability monitoring method as described in claim 5, characterized in that, Also includes: Adjust the corresponding weighting coefficients according to process constraints: If the process constraints include average power consistency constraints, then increase the weighting factor of the fluctuation coefficient CV. If the process constraints include welding thickness accuracy constraints, then the peak-to-peak value P is increased. pp Weighting coefficients; If the process constraints include periodic stripe defect constraints, then increase the effective value P of the power fluctuation corresponding to the power ripple. rms The weighting coefficients.

7. The laser power instability monitoring method as described in claim 1, characterized in that, Also includes: The power sequence of the laser output is collected using a laser power generator; The laser power unit includes an optical signal acquisition module and a signal conversion module. The optical signal acquisition module is used to acquire the optical signal of the laser output by the laser, and the signal conversion module is used to convert the electrical signal into a power signal to obtain the power sequence of the output laser.

8. The laser power instability monitoring method as described in claim 1, characterized in that, The optical signal acquisition module includes an optical homogenizer and a photodetector; The light homogenizer includes: A substrate, wherein an optical homogenization cavity is disposed inside the substrate, and an observation port communicating with the optical homogenization cavity is disposed on the substrate; When the optical fiber passes through the optical homogenization cavity, the reflecting surface and the observation port of the optical homogenization cavity are located on both sides of the optical fiber, and the laser emitted from the cladding region on the side of the optical fiber facing the reflecting surface is directed towards the observation port after passing through the reflecting surface, while the laser emitted from the cladding region on the side of the optical fiber facing the observation port is directed directly towards the observation port.

9. The laser power instability monitoring method as described in claim 8, characterized in that, The reflecting surface is an arc-shaped diffuse reflecting surface. The laser emitted from the cladding region on one side of the optical fiber toward the reflecting surface is directed toward the arc-shaped diffuse reflecting surface, and after being scattered and homogenized, it is directed toward the observation port.

Images