A real-time monitoring data processing method based on multi-source beam parameter fusion

Abstract

The present application relates to the field of data analysis and mining technology, and more particularly to a real-time monitoring data processing method based on multi-source light beam parameter fusion, which comprises: synchronously collecting original signals through a position sensitive detector, an image sensor and a power monitoring unit and converting them into digital feature vectors; performing spatio-temporal alignment and weighted fusion preprocessing on heterogeneous data to form a standardized light beam state data set; deeply analyzing the data set using a binary Gaussian fitting algorithm to extract core quality characteristics such as centroid position, diameter and quality factor; comparing real-time parameters with target values based on a PID control algorithm to drive an actuator to dynamically compensate for pointing deviation and power fluctuation; and monitoring the state of real-time parameters and triggering an early warning and data archiving when the tolerance threshold is exceeded. The present application eliminates the limitations of single sensor monitoring and significantly improves the recognition accuracy of complex light field changes.

Description

Technical Field

[0001] This invention relates to the field of data analysis and mining technology, and in particular to a real-time monitoring data processing method based on multi-source beam parameter fusion. Background Technology

[0002] Real-time assessment and stability control of beam quality are core aspects of ensuring the performance of optical systems. By digitally characterizing the physical properties of the beam, precise control over complex optical processes can be achieved. However, existing technologies often employ independent sampling and serial analysis modes when processing multi-source beam data, resulting in insufficient timestamp alignment accuracy between heterogeneous sensors and making it difficult to achieve deep parameter fusion.

[0003] Meanwhile, existing linear filtering algorithms are unable to effectively eliminate nonlinear interference caused by environmental noise and stray light, resulting in the fused data failing to accurately reflect the true transient changes of the beam, which seriously affects the monitoring accuracy and control robustness in precision optical tasks. Summary of the Invention

[0004] The purpose of this invention is to provide a real-time monitoring data processing method based on multi-source beam parameter fusion to solve the technical problems mentioned in the background art.

[0005] This invention provides a real-time monitoring data processing method based on multi-source beam parameter fusion, comprising: Step 1. Using a position-sensitive detector to acquire the transient position offset signal of the beam, using an image sensor to acquire the two-dimensional intensity distribution image data of the beam, using a power monitoring unit to acquire the instantaneous power signal of the beam, and synchronously sampling the transient position offset signal, the two-dimensional intensity distribution image data and the instantaneous power signal, and converting them into digital feature vectors; Step 2. Based on the high-precision hardware clock triggering mechanism, the heterogeneous data from different sensors are time-stamped and corrected. The image pixel coordinate system of the image sensor is mapped to the physical position coordinate system of the position-sensitive detector using a pre-calibrated spatial transformation matrix. Based on the real-time signal-to-noise ratio of each sensor, a weighted fusion algorithm with dynamic weight allocation is used to eliminate redundancy and conflict between multi-source data and generate a standardized beam state dataset. Step 3. Perform a binary Gaussian fitting operation on the standardized beam state dataset, extract the centroid position coordinates, diameter of a specific multiple standard deviation, and ellipticity characterizing the symmetry of the beam by nonlinear iterative solution, and calculate the beam quality factor and far-field divergence angle parameters by combining multi-section scanning data with hyperbolic fitting algorithm. Step 4. Compare the real-time extracted beam core quality characteristics with the preset target value, generate control commands based on the proportional-integral-derivative control algorithm with parameter self-tuning function, drive the electric adjustment frame and electric attenuator actuator to dynamically compensate for beam pointing deviation and power fluctuation, so that the beam state is locked within the reference range. Step 5. Monitor whether the parameters of the entire process exceed the preset tolerance threshold. If any parameter is in an abnormal state, execute the safety interlock protection action and write the processed monitoring data into a structured database with a time series architecture in real time.

[0006] In some embodiments, the induced current generated by the incident beam is sensed by a position-sensitive detector, and after impedance transformation by a preamplifier circuit, it enters a synchronous sampling analog-to-digital converter to capture the transient position shift of the beam within a predetermined time range within a preset frequency range, forming a sub-vector containing the instantaneous axial displacement. Using an image sensor, a light spot image is acquired in global shutter mode to ensure that all pixels in each frame are exposed at the same time to eliminate motion blur and distortion. The original grayscale image data stream is then transmitted to the memory buffer of the main controller. Using a power monitoring unit, the beam energy is sensed within a predetermined wavelength range. After linearizing the electrical signal output by the sensor, the instantaneous power value of the beam is output at a predetermined sampling period.

[0007] In some embodiments, step 2 includes: in the time dimension, the main controller sends a synchronization trigger pulse to each detector to control the sampling delay of different sensors within a predetermined time deviation, and performs an interpolation compensation algorithm for sensors with inconsistent sampling frequencies to ensure that the multi-source data achieves strict point-to-point correspondence on the time axis; In the spatial dimension, the calibration program drives the actuator to make the beam translate along a preset path. The least squares method is used to fit the mapping relationship between the centroid coordinates of image pixels and physical displacement data, and a spatial transformation matrix containing translation factor, rotation factor and scaling factor is constructed, ensuring that the calibration residual is within the preset residual threshold. In terms of fusion logic, the saturation and contrast of image data and the voltage fluctuation rate of the position detector are evaluated in real time. The weight ratio of the image sensor is increased in a stable light environment, and the weight ratio of the position sensitive detector is increased when high-frequency mechanical vibration or rapid light spot jitter is detected.

[0008] In some embodiments, step 3 includes: processing the two-dimensional intensity distribution data using a binary Gaussian fitting algorithm, wherein the fitting parameters include peak intensity, horizontal standard deviation, vertical standard deviation and background noise level, and using the Levenberg-Marquardt algorithm to perform parameter optimization iteration until the goodness of fit meets the preset goodness of fit threshold. When calculating the diameter at a specific multiple of the standard deviation, background noise interference is eliminated by setting a grayscale threshold, and the second moment of the light intensity distribution is integrated to define the effective beam width. When calculating the ellipticity parameter, it is defined as the ratio of the minor axis diameter to the major axis diameter of the spot. When this ratio is lower than the preset ratio threshold, the system logic determines that the beam has an abnormal distortion.

[0009] In some embodiments, the calculation of beam quality factor and far-field divergence angle parameters includes: the collaborative scanning module performing beam width acquisition tasks at different axial positions in the beam propagation direction to acquire multiple sets of experimental data points containing position information and corresponding beam width information; performing hyperbolic fitting on the experimental data points to extract the beam waist position, beam waist diameter and far-field divergence angle, and evaluating the output mode stability of the laser based on the product relationship between the beam waist diameter and the far-field divergence angle.

[0010] In some embodiments, step 4 includes: the main controller automatically optimizes the proportional coefficient, integral coefficient, and derivative coefficient according to the transfer function of the optical system, and generates an adjustment variable by calculating the deviation between the real-time monitoring parameters and the user-set target value; The electric adjustment frame uses a high-resolution stepper motor or piezoelectric ceramic driver. The main controller converts the angle compensation amount into a pulse sequence to drive the electric adjustment frame to adjust the deflection angle of the reflector, so that the center of mass of the beam is locked at the reference position. The electric attenuator uses a continuously variable neutral density filter or polarization control unit to adjust the transmittance or polarization state in real time according to the feedback signal of the power monitoring unit, control the power fluctuation within the preset fluctuation range, and monitor the stroke limit of the actuator in real time during the adjustment process.

[0011] In some embodiments, step 5 includes: the early warning module compares the position deviation, power fluctuation, beam width change and ellipticity parameters in real time, and when any parameter exceeds the tolerance threshold set by the user, it initiates an alarm action within a predetermined response time; the alarm action includes displaying warning information on the graphical interface, activating the audio-visual output of the main controller, and sending a safety interlock command to the external laser power supply through the general input / output interface to cut off the laser output; The structured database records a complete historical record, including timestamps, location coordinates, power values, and system status codes, at a preset write speed. It supports continuous recording mode and triggered recording mode, and automatically saves a snapshot of the original data before and after the anomaly occurs in triggered recording mode.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates a position-sensitive detector, an image sensor, and a power monitoring unit to achieve comprehensive characterization of the physical properties of a light beam. Utilizing a high-precision spatiotemporal alignment algorithm, it deeply fuses heterogeneous sensor data, eliminating the limitations of single-sensor monitoring. In position monitoring, its repeatability is better than a preset order of magnitude; in power measurement, its accuracy meets the predetermined requirements. 2. Based on an embedded real-time operating system and an optimized proportional-integral-derivative control algorithm, this invention achieves an extremely short response speed. The system can calculate the beam centroid and morphological parameters in real time and quickly generate compensation commands to drive the actuator, significantly improving the beam pointing stability. 3. This invention not only provides basic parameter monitoring but also integrates advanced analysis tools such as Gaussian fitting and beam quality factor calculation, enabling automatic evaluation of the far-field characteristics of the beam. Through flexible alarm settings and safety interlock functions, the system can achieve a safe response in a very short time when parameters exceed the tolerance range, effectively preventing damage to optical components from high-power lasers. Simultaneously, comprehensive data recording and export functions provide detailed data support for process optimization and fault tracing.

[0013] 4. Through modular algorithms, the system of this invention has good scalability and can adapt to the monitoring needs of different wavelengths, power levels and beam shapes. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a flowchart of the real-time monitoring data processing method based on multi-source beam parameter fusion of the present invention; Figure 2 This is the core principle diagram of the closed-loop feedback control of this invention. Detailed Implementation

[0016] The following will be based on embodiments of the present invention. Figures 1-2 The technical solutions in the embodiments of the present invention will be clearly and completely described together. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] Example 1 This embodiment provides a real-time monitoring data processing method based on multi-source beam parameter fusion, which includes at least the following steps: Step 1. Acquiring the transient position offset signal of the beam using a position-sensitive detector, acquiring the two-dimensional intensity distribution image data of the beam using an image sensor, and acquiring the instantaneous power signal of the beam using a power monitoring unit. The transient position offset signal, the two-dimensional intensity distribution image data, and the instantaneous power signal are synchronously sampled and converted into a digital feature vector. The digital feature vector includes horizontal displacement, vertical displacement, two-dimensional intensity distribution, and instantaneous power. Specifically, the execution process of Step 1 involves concurrent data acquisition from multiple heterogeneous sensing terminals. The position-sensitive detector uses a high-sensitivity photodiode array, and its physical structure is configured as a four-quadrant detector or a lateral effect position sensor. After the beam is incident on the photosensitive surface of the detector, the induced current is impedance-transformed by a preamplifier circuit and then enters a multi-channel synchronous sampling analog-to-digital converter. The sampling frequency is set within a preset frequency range of 1000Hz to 5000Hz, which is sufficient to capture the transient position offset of the beam in the microsecond-level predetermined time range. The digitized original position signal is initially filtered by a field-programmable gate array (FPGA) to form a sub-vector containing instantaneous X-axis and Y-axis displacements.

[0018] The image sensor employs a complementary metal-oxide-semiconductor (CMOS) photosensitive element with an effective pixel count of no less than 2 million pixels. The pixel size is precisely controlled within a preset range of 3.45 micrometers to 5.5 micrometers. Understandably, to eliminate the ghosting effect of fast-moving light spots, the image sensor is configured in global shutter mode, ensuring that all pixels in each frame begin and end exposure simultaneously. Each frame of image data from the image sensor includes two-dimensional intensity distribution information, i.e., the grayscale value of each pixel. These grayscale values ​​constitute the high-dimensional spatial information portion of the digital feature vector.

[0019] The power monitoring unit integrates a thermopile sensor or photodetector with a spectral response range covering a predetermined wavelength range of 200 nm to 1100 nm or wider. The internal circuitry of the power monitoring unit precisely amplifies and linearizes the weak electrical signal output by the sensor, achieving a power measurement accuracy of ±0.5%. The power monitoring unit outputs the instantaneous power value of the beam at a predetermined sampling period, using it as a scalar dimension in the feature vector.

[0020] Step 2. Based on a high-precision hardware clock triggering mechanism, timestamp synchronization correction is performed on heterogeneous data from different sensors. A pre-calibrated spatial transformation matrix maps the image pixel coordinate system of the image sensor to the physical position coordinate system of the position-sensitive detector. Then, based on the real-time signal-to-noise ratio of each sensor, a weighted fusion algorithm with dynamic weight allocation eliminates redundancy and conflicts between multi-source data, generating a standardized beam state dataset. Specifically, the timestamp synchronization correction process in Step 2 employs a high-precision hardware clock triggering mechanism. The main controller sends synchronization trigger pulses to each detector, controlling the sampling delay of different sensors within a predetermined time deviation of less than 10 microseconds. For sensors with inconsistent sampling frequencies, the system executes an interpolation compensation algorithm.

[0021] During the calibration phase, the electric adjustment frame is driven to translate the light beam along a preset path within the detection range. The pixel centroid coordinates output by the image sensor and the physical displacement data output by the position-sensitive detector are recorded. The mapping relationship between the two is fitted using the least squares method to obtain a 3x3 spatial transformation matrix that includes translation, rotation, and scaling factors. The calibration process calculates the sum of squared residuals to ensure that the calibration residual is less than a preset residual threshold of 0.01 pixels, thereby achieving accurate projection from the image coordinate system to the physical coordinate system.

[0022] The weighted fusion algorithm dynamically assigns weights based on the real-time signal-to-noise ratio of each sensor. The system evaluates the saturation and contrast of the image data, as well as the voltage fluctuation rate of the position detector, in real time. Specifically, in a stable environment with moderate light intensity, the weight of the image sensor is increased to obtain richer morphological features; when high-frequency mechanical vibration or rapid beam jitter is detected, the system automatically increases the weight of the position-sensitive detector, utilizing its extremely high response bandwidth to maintain the continuity of position monitoring, thereby forming a standardized beam state dataset with high robustness.

[0023] Step 3. Perform a binary Gaussian fitting operation on the standardized beam state dataset. Through nonlinear iteration, extract the beam's centroid coordinates, diameter at a specific multiple of the standard deviation, and ellipticity representing the beam's symmetry. Combined with multi-section scanning data, calculate the beam quality factor and far-field divergence angle parameters using a hyperbolic fitting algorithm. Specifically, the binary Gaussian fitting algorithm in Step 3 performs a nonlinear iterative solution on the two-dimensional intensity distribution data. The mathematical model is defined as follows: in, I ( x, y ) represents a pixel ( x, y The light intensity at point ( ); A is the peak intensity; x 0 , y 0 () represents the centroid coordinates obtained from the fitting. and , representing the standard deviations in the horizontal and vertical directions, respectively; B represents the background noise level. The system uses the Levenberg-Marquardt algorithm for parameter optimization, with a goodness-of-fit greater than the preset goodness-of-fit threshold of 0.99.

[0024] The calculation of the diameter at a specific multiple of the standard deviation follows the ISO standard, obtaining a diameter four times the standard deviation by integrating the second moment of the light intensity distribution. This calculation process eliminates background noise interference by setting a grayscale threshold, accurately defining the effective beam width. The ellipticity parameter is defined as the ratio of the fitted minor axis diameter to the major axis diameter. When this ratio is lower than a preset threshold of 0.85, the system logic determines that the beam is distorted.

[0025] For the calculation of beam quality factor, the system's collaborative scanning module collects beam width data at different positions along the beam propagation direction. By performing hyperbolic fitting on these data points, the beam waist position, beam waist diameter, and far-field divergence angle are extracted. These parameters together constitute a full-dimensional quality profile of the beam.

[0026] Step 4. Compare the real-time extracted beam core quality characteristics with the preset target value. Based on the proportional-integral-derivative (PID) control algorithm with parameter self-tuning function, generate control commands to drive the electric adjustment frame and electric attenuator actuators, dynamically compensating for beam pointing deviation and power fluctuations, thus locking the beam state within the reference range. Specifically, the system automatically optimizes the proportional coefficient K based on the optical system's transfer function. p Integral coefficient K i and differential coefficient K d .

[0027] The electric adjustment frame is driven by a high-resolution stepper motor, with a repeatability better than the preset accuracy of 0.5 microradians. The main controller converts the calculated angle compensation into a pulse sequence, driving the adjustment frame to precisely adjust the deflection angle of the reflector, ensuring the beam centroid is always locked at the reference position. For power control, the electric attenuator utilizes a continuously variable neutral density filter, adjusting the transmittance in real time based on feedback signals from the power monitoring unit, keeping power fluctuations within a preset range of ±0.5%.

[0028] Step 5. Monitor whether the parameters throughout the process exceed the preset tolerance thresholds. If any parameter is in an abnormal state, execute a safety interlock protection action and write the processed monitoring data into a structured database with a time-series architecture in real time. Specifically, the early warning logic in Step 5 compares position deviation, power fluctuation, beam width change, and ellipticity in real time. When any parameter exceeds the user-set upper or lower limit, the system immediately activates the early warning module, with an alarm response time of less than 10 milliseconds. Early warning actions include popping up a prominent red warning box on the graphical interface, activating the buzzer output of the main controller, sending an email to designated management personnel, and sending a safety interlock command to the laser power supply through the general input / output interface to forcibly cut off the laser output to protect downstream optical components.

[0029] The structured database adopts a high-performance time-series database architecture, supporting a preset write speed of over 1000 records per second. Each record contains a complete timestamp, location coordinates, power value, fitting parameters, and system status code.

[0030] Example 2 This embodiment, based on Embodiment 1, further refines the specific application of this method in laser processing quality assurance scenarios. During laser welding or cutting, the stability of the laser beam directly determines the consistency of the molten pool and the quality of the cut.

[0031] In this embodiment, the acquisition process in step 1 is specifically optimized for high-power laser environments. A high-magnification attenuation plate group is placed in front of the position-sensitive detector to ensure that the incident light intensity is within the linear operating range of the detector. The image sensor uses a CMOS photosensitive element with ultra-high dynamic range, which can simultaneously capture the main laser spot and the weak scattered light around it. This is of great significance for identifying plasma plume interference during the processing.

[0032] In step 2, the system introduces a digital compensation algorithm for industrial environment vibrations. The accelerometer built into the main controller senses the vibration frequency and amplitude of the machine tool base in real time and uses this as a compensation factor input to the spatial transformation matrix. Through this active vibration cancellation mechanism, the system is able to extract a clean beam pointing drift signal from the high-noise raw data.

[0033] In step 3, a customized algorithm is added for non-Gaussian beam shapes. For flat-top beams or ring beams commonly used in laser processing, the system no longer relies solely on binary Gaussian fitting, but instead uses edge detection and moment analysis methods to calculate the equivalent diameter and uniformity of the beam spot. The system calculates the root mean square error of the intensity distribution within the beam spot in real time, thereby characterizing the energy flatness of the beam.

[0034] In step 4, the control commands not only drive the internal electric adjustment frame but also send real-time position corrections to the machine tool controller via industrial Ethernet protocols (such as EtherCAT or Profinet). This linkage control mechanism ensures that the laser beam always precisely coincides with the predetermined machining path during complex machining trajectory movements. For power compensation, the adjustment speed of the electric attenuator is synchronized with the machine tool's feed speed, ensuring a constant laser energy input per unit length.

[0035] In step 5, during data archiving, the system associates and stores beam monitoring data with the serial number of the processed part. When a decrease in beam quality is detected (such as a sudden change in ellipticity), the system records the precise processing location coordinates in the database. This provides crucial information for subsequent quality traceability. By analyzing the historical trends of beam parameters, the damage progression of optical lenses can be predicted, enabling preventative maintenance.

[0036] In specific high-precision laser welding tasks, the system first performs an automatic search and connection to the equipment. After successful connection, the operator sets the target power value and position tolerance range required for welding through the configuration wizard. During the welding process, the monitoring view displays the two-dimensional shape of the laser spot in real time. When the laser spot diameter increases beyond a preset proportional threshold due to the thermal lensing effect caused by heat accumulation at the welding head, the system immediately triggers an audible and visual alarm and automatically adjusts the electric position of the collimating lens group to compensate, restoring the laser spot diameter to the design range.

[0037] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the scope of the invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0038] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A real-time monitoring data processing method based on multi-source beam parameter fusion, characterized in that, include: Step 1. Use a position-sensitive detector to acquire the transient position offset signal of the beam, use an image sensor to acquire the two-dimensional intensity distribution image data of the beam, use a power monitoring unit to acquire the instantaneous power signal of the beam, and simultaneously sample the transient position offset signal, the two-dimensional intensity distribution image data, and the instantaneous power signal, and convert them into digital feature vectors. Step 2. Based on the high-precision hardware clock triggering mechanism, the heterogeneous data from different sensors are time-stamped and corrected. The image pixel coordinate system of the image sensor is mapped to the physical position coordinate system of the position-sensitive detector using a pre-calibrated spatial transformation matrix. Based on the real-time signal-to-noise ratio of each sensor, a weighted fusion algorithm with dynamic weight allocation is used to eliminate redundancy and conflict between multi-source data and generate a standardized beam state dataset. Step 3. Perform a binary Gaussian fitting operation on the standardized beam state dataset, extract the centroid position coordinates, diameter of a specific multiple standard deviation, and ellipticity characterizing the symmetry of the beam by nonlinear iterative solution, and calculate the beam quality factor and far-field divergence angle parameters by combining multi-section scanning data with hyperbolic fitting algorithm. Step 4. Compare the real-time extracted beam core quality characteristics with the preset target value, generate control commands based on the proportional-integral-derivative control algorithm with parameter self-tuning function, drive the electric adjustment frame and electric attenuator actuator to dynamically compensate for beam pointing deviation and power fluctuation, so that the beam state is locked within the reference range. Step 5. Monitor whether the parameters of the entire process exceed the preset tolerance threshold. If any parameter is in an abnormal state, execute the safety interlock protection action and write the processed monitoring data into a structured database with a time series architecture in real time.

2. The method according to claim 1, characterized in that, The induced current generated by the incident beam is sensed by a position-sensitive detector. After impedance transformation by a preamplifier circuit, it enters a synchronous sampling analog-to-digital converter to capture the transient position shift of the beam within a predetermined time range within a preset frequency range, forming a sub-vector containing the instantaneous axial displacement. Using an image sensor, a light spot image is acquired in global shutter mode to ensure that all pixels in each frame are exposed at the same time to eliminate motion blur and distortion. The original grayscale image data stream is then transmitted to the memory buffer of the main controller. Using a power monitoring unit, the beam energy is sensed within a predetermined wavelength range. After linearizing the electrical signal output by the sensor, the instantaneous power value of the beam is output at a predetermined sampling period.

3. The method according to claim 1, characterized in that, Step 2 includes: in the time dimension, the main controller sends a synchronization trigger pulse to each detector to control the sampling delay of different sensors within a predetermined time deviation, and performs an interpolation compensation algorithm for sensors with inconsistent sampling frequencies to ensure that the multi-source data achieves strict point-to-point correspondence on the time axis; In the spatial dimension, the calibration program drives the actuator to make the beam translate along a preset path. The least squares method is used to fit the mapping relationship between the centroid coordinates of image pixels and physical displacement data, and a spatial transformation matrix containing translation factor, rotation factor and scaling factor is constructed, ensuring that the calibration residual is within the preset residual threshold. In terms of fusion logic, the saturation and contrast of image data and the voltage fluctuation rate of the position detector are evaluated in real time. The weight ratio of the image sensor is increased in a stable light environment, and the weight ratio of the position sensitive detector is increased when high-frequency mechanical vibration or rapid light spot jitter is detected.

4. The method according to claim 1, characterized in that, Step 3 includes: processing the two-dimensional intensity distribution data using a binary Gaussian fitting algorithm, with fitting parameters including peak intensity, horizontal standard deviation, vertical standard deviation and background noise level, and using the Levenberg-Marquardt algorithm to perform parameter optimization iteration until the goodness of fit meets the preset goodness of fit threshold. When calculating the diameter at a specific multiple of the standard deviation, background noise interference is eliminated by setting a grayscale threshold, and the second moment of the light intensity distribution is integrated to define the effective beam width. When calculating the ellipticity parameter, it is defined as the ratio of the minor axis diameter to the major axis diameter of the spot. When this ratio is lower than the preset ratio threshold, the system logic determines that the beam has an abnormal distortion.

5. The method according to claim 1, characterized in that, The calculation of beam quality factor and far-field divergence angle parameters includes: the collaborative scanning module performing beam width acquisition tasks at different axial positions in the beam propagation direction to obtain multiple sets of experimental data points containing position information and corresponding beam width information; performing hyperbolic fitting on the experimental data points to extract the beam waist position, beam waist diameter and far-field divergence angle, and evaluating the output mode stability of the laser based on the product relationship between the beam waist diameter and the far-field divergence angle.

6. The method according to claim 1, characterized in that, Step 4 includes: the main controller automatically optimizes the proportional coefficient, integral coefficient and derivative coefficient according to the transfer function of the optical system, and generates adjustment variables by calculating the deviation between the real-time monitoring parameters and the user-set target values. The electric adjustment frame uses a high-resolution stepper motor or piezoelectric ceramic driver. The main controller converts the angle compensation amount into a pulse sequence to drive the electric adjustment frame to adjust the deflection angle of the reflector, so that the center of mass of the beam is locked at the reference position. The electric attenuator uses a continuously variable neutral density filter or polarization control unit to adjust the transmittance or polarization state in real time according to the feedback signal of the power monitoring unit, control the power fluctuation within the preset fluctuation range, and monitor the stroke limit of the actuator in real time during the adjustment process.

7. The method according to claim 1, characterized in that, Step 5 includes: the early warning module compares the position deviation, power fluctuation, beam width change and ellipticity parameters in real time. When any parameter exceeds the tolerance threshold set by the user, an alarm action is initiated within a predetermined response time. The alarm action includes displaying warning information on the graphical interface, activating the audio-visual output of the main controller, and sending a safety interlock command to the external laser power supply through the general input / output interface to cut off the laser output. The structured database records a complete historical record, including timestamps, location coordinates, power values, and system status codes, at a preset write speed. It supports continuous recording mode and triggered recording mode, and automatically saves a snapshot of the original data before and after the anomaly occurs in triggered recording mode.

Images