A method and system for verifying beam spots in multi-laser 3D printing

Abstract

This application discloses a method and system for verifying beam spots in multi-laser 3D printing, belonging to the field of laser processing control technology. The method includes: synchronously fusing multi-physical field signal data such as thermal response, stress, and pressure from multiple laser sources to obtain comprehensive physical field data; extracting thermal response data from the laser beam spot region and fusing it with the comprehensive data to obtain positional offset feature data; performing Fourier transform on the thermal response data to extract frequency domain feature data, and further processing it to obtain beam spot frequency feature data; generating verification standard data by combining adaptive thermal response thresholds; fusing external temperature, humidity, and interference information to generate environmental adaptability data, obtaining optimized signal processing data; and finally obtaining laser beam spot accuracy data. This solution improves the stability, consistency, and intelligent response capability of multi-laser systems in high-precision manufacturing scenarios, improving printing quality, reducing defect rates, and extending equipment lifespan.

Description

Technical Field

[0001] This application belongs to the field of laser processing control technology, specifically relating to a method and system for multi-laser 3D printing beam spot verification. Background Technology

[0002] In multi-laser 3D printing technology, the accuracy and stability of the laser beam spot directly determine the quality and performance of additively manufactured parts. With the increasing demand for high-precision and high-efficiency manufacturing of complex metal components in aerospace, biomedical, and other fields, multi-laser collaborative printing has become a key technological direction. However, during dynamic scanning, multi-laser beam spots are susceptible to the coupling effects of multiple physical fields such as thermal stress, powder splashing, and substrate deformation, leading to problems such as beam spot position shift and uneven energy density distribution, which in turn cause printing defects (such as interlayer misalignment and increased porosity). Traditional single-laser beam spot calibration methods cannot meet the real-time and global accuracy requirements of multi-laser collaborative operations. There is an urgent need for a technology that can simultaneously monitor multi-physical field signals and dynamically calibrate beam spot parameters to solve the core challenges of low yield and poor process repeatability caused by beam spot inaccuracy in multi-laser 3D printing.

[0003] Currently, independent sensors are used to collect multi-physical field signals such as thermal response, stress, and pressure in a time-division manner. For example, infrared thermal imagers monitor the temperature field, and displacement sensors measure structural deformation. Subsequently, the collected signals are simply superimposed using post-processing algorithms, such as directly adding thermal imaging data and displacement data or weighted averaging. Finally, a fixed threshold judgment method is used, such as preset temperature gradient threshold or stress upper limit value, to perform offline analysis on the superimposed data. If the threshold is exceeded, an alarm is triggered or parameters are adjusted.

[0004] Because the independent sensors collect data in a time-division manner, the timing of the data is misaligned, making it difficult to accurately correlate transient phenomena such as thermal expansion and beam spot shift. Beam spot position monitoring relies solely on optical image analysis without integrating frequency domain features and environmental adaptability parameters, thus missing key information such as beam spot vibration modes. Fixed thresholds and static environmental models cannot adapt to dynamic working conditions, resulting in low real-time calibration accuracy. Furthermore, the technology is essentially a "data-stacking analysis" approach, without constructing a multi-physics coupling model and a closed-loop feedback mechanism, making it difficult to meet the stability requirements of long-term printing of complex components. Summary of the Invention

[0005] This application provides a method and system for multi-laser 3D printing beam spot calibration, which solves the problems of existing technologies where data timing misalignment caused by time-sharing acquisition by independent sensors makes it difficult to accurately correlate transient phenomena such as thermal expansion and beam spot shift; beam spot position monitoring relies solely on optical image analysis without integrating frequency domain features and environmental adaptability parameters, thus omitting key information such as beam spot vibration modes; fixed thresholds and static environmental models cannot adapt to dynamic working conditions, resulting in low real-time calibration accuracy; furthermore, the technology is essentially "data stacking analysis" without constructing a multi-physics coupling model and closed-loop feedback mechanism, making it difficult to meet the stability requirements of long-term printing of complex components.

[0006] In a first aspect, embodiments of this application provide a method for verifying beam spots in multi-laser 3D printing, the method comprising:

[0007] Multiphysics field signal data from multiple laser sources are acquired, and the multiphysics field signal data is synchronously acquired and fused to obtain comprehensive physical field data; wherein, the multiphysics field signal data includes thermal response signal, stress signal and pressure signal;

[0008] Thermal response data of beam spot regions from multiple laser sources are acquired, and the thermal response data is fused and analyzed with comprehensive physical field data to obtain multi-laser beam spot position offset characteristic data.

[0009] Fourier transform is performed on the thermal response signal data of multiple laser sources to obtain the frequency domain feature data of multiple laser sources. Fourier transform is performed on the beam spot position offset feature data and the frequency domain feature data of the multiple laser sources to obtain the beam spot frequency feature data of multiple laser sources.

[0010] Obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data, and frequency domain characteristic data to obtain the verification standard data;

[0011] Acquire external temperature data, external humidity data, and external interference data; determine environmental adaptability data based on the external temperature data, external humidity data, and external interference data; combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data.

[0012] Real-time laser beam spot accuracy data from multiple laser sources is acquired, and the real-time laser beam spot accuracy data is combined with optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

[0013] Secondly, embodiments of this application provide a multi-laser 3D printing beam spot verification system, the system comprising:

[0014] The physical field fusion module is used to acquire multi-physics field signal data from multiple laser sources, and to synchronously acquire and fuse the multi-physics field signal data to obtain comprehensive physical field data; wherein, the multi-physics field signal data includes thermal response signals, stress signals and pressure signals;

[0015] The offset recognition module is used to acquire thermal response data of the beam spot regions of multiple laser sources, and to perform fusion analysis on the thermal response data and comprehensive physical field data to obtain multi-laser beam spot position offset feature data.

[0016] The frequency domain extraction module is used to perform Fourier transform processing on the thermal response signal data of multiple laser sources to obtain the frequency domain feature data of multiple laser sources, and to perform Fourier transform processing on the multiple laser beam spot position offset feature data and the frequency domain feature data to obtain the beam spot frequency feature data of multiple laser sources.

[0017] The verification standard generation module is used to obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data and frequency domain characteristic data to obtain the verification standard data.

[0018] The environmental adaptation module is used to acquire external temperature data, external humidity data, and external interference data, determine environmental adaptability data based on the external temperature data, external humidity data, and external interference data, and combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data.

[0019] The accuracy calibration module is used to acquire real-time laser beam spot accuracy data from multiple laser sources, and combine the real-time laser beam spot accuracy data with optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

[0020] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0021] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0022] In this embodiment, multi-physics field signal data from multiple laser sources are acquired, and the multi-physics field signal data is synchronously acquired and fused to obtain comprehensive physical field data. The multi-physics field signal data includes thermal response signals, stress signals, and pressure signals. Thermal response data of the beam spot regions of multiple laser sources are acquired, and the thermal response data is fused and analyzed with the comprehensive physical field data to obtain multi-laser beam spot position offset characteristic data. Fourier transform processing is performed on the thermal response signal data of multiple laser sources to obtain frequency domain characteristic data of multiple laser sources. Fourier transform processing is then performed on the multi-laser beam spot position offset characteristic data and the frequency domain characteristic data. The method involves obtaining beam frequency characteristic data from multiple laser sources; acquiring an adaptive thermal response threshold; combining the adaptive thermal response threshold, beam frequency characteristic data, and frequency domain characteristic data to obtain calibration standard data; acquiring external temperature data, external humidity data, and external interference data; determining environmental adaptability data based on these data; combining the environmental adaptability data with the calibration standard data to obtain optimized signal processing data; and acquiring real-time laser beam accuracy data from multiple laser sources. This real-time laser beam accuracy data is then combined with the optimized signal processing data to obtain calibrated multi-laser beam accuracy data. This multi-laser 3D printing beam accuracy calibration method improves the stability, consistency, and intelligent response capabilities of multi-laser systems in high-precision manufacturing scenarios, contributing to improved printing quality, reduced defect rates, and extended equipment lifespan. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the multi-laser 3D printing beam spot verification method provided in Embodiment 1 of this application;

[0024] Figure 2 This is a flowchart illustrating the multi-laser 3D printing beam spot verification method provided in Embodiment 2 of this application;

[0025] Figure 3 This is a schematic diagram of the structure of the multi-laser 3D printing beam spot verification system provided in Embodiment 3 of this application;

[0026] Figure 4 This is a schematic diagram of the structure of the electronic device provided in Embodiment 4 of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0028] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0029] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0030] The following description, in conjunction with the accompanying drawings, details a multi-laser 3D printing beam spot verification method and system provided in this application through specific embodiments and application scenarios.

[0031] Example 1

[0032] Figure 1 This is a schematic flowchart of the multi-laser 3D printing beam spot verification method provided in Embodiment 1 of this application. Figure 1 As shown, the specific steps include the following:

[0033] S101, acquire multi-physics field signal data from multiple laser sources, and perform synchronous acquisition and fusion processing on the multi-physics field signal data to obtain comprehensive physical field data; wherein, the multi-physics field signal data includes thermal response signal, stress signal and pressure signal.

[0034] A laser source can refer to a light source device that can emit laser light (which has high directionality, monochromaticity, high brightness and coherence).

[0035] Multiphysics signal data can refer to multiple measurement signals from different physical properties or action fields of an object or system, and is often used for the coupled behavior analysis of complex systems.

[0036] Comprehensive physical field data can be a dataset obtained by synchronously acquiring, aligning, and fusing the above-mentioned multiple physical signals (heat, force, pressure, etc.).

[0037] Thermal response signals can refer to signals caused by temperature changes or thermal conductivity characteristics of a system after it is subjected to laser or other energy input. These signals mainly include surface or internal temperature distribution, thermal infrared images or thermocouple data, and deformation signals caused by thermal expansion.

[0038] Stress signals can refer to the internal stress or strain data generated by a material under external forces (such as laser shock or thermal expansion caused by heating). They can be expressed as: stress waveform sensor data (such as strain gauges and laser interferometers), structural deformation, stress concentration, and can reflect changes in the mechanical properties of the material.

[0039] Pressure signals can refer to pressure changes in a system caused by external or internal forces, such as pressure changes in gases or liquids (measured using pressure sensors), internal pressure responses within enclosed structures, and transient pressure responses under laser shock (e.g., in explosive laser shock experiments).

[0040] During laser excitation, multiple physical signal sensors are deployed to acquire corresponding multiphysics field signal data. Specifically, these include:

[0041] Thermal response signal acquisition: Thermocouples, infrared thermal imagers, thermistors, or fiber optic temperature sensors are installed on or inside the object under test to enable real-time monitoring of the temperature rise process caused by laser heating. The thermal imager continuously acquires images of the target surface temperature distribution at a high frame rate (e.g., 30–120 fps) to capture transient temperature changes.

[0042] Stress signal acquisition: Strain gauges (resistance strain gauges) are attached to critical stress concentration areas of the structure or fiber Bragg grating (FBG) sensors are deployed to acquire strain responses caused by thermal expansion, impact loads, or local yielding of the material in real time. Strain data is acquired synchronously via a high-precision data acquisition card (DAQ), with a sampling rate typically above 10kHz to capture dynamic stress waves.

[0043] Pressure signal acquisition: For closed systems or gas-liquid interaction areas, miniature piezoelectric pressure sensors or MEMS pressure sensors are deployed to collect pressure changes caused by local gas expansion, liquid disturbance, or shock wave response induced by laser excitation. The pressure signal needs to share a clock source with the temperature and stress signal sampling systems to ensure time consistency of the sampled data.

[0044] To achieve time-synchronous acquisition of the aforementioned multi-physics field signals, a unified data acquisition and control platform is constructed. This platform controls all sensor acquisition modules through a unified clock and uses multi-channel high-speed acquisition cards (such as the NIPXI series) to achieve parallel acquisition. Data synchronization is ensured through timestamp marking and triggering mechanisms (such as TTL trigger signals) to guarantee that all sensors start acquisition synchronously when a laser excitation event occurs. After acquisition, the maximum temperature rise, heating rate, and temperature gradient are extracted from the thermal response signal; stress wave amplitude, strain rate, and dominant frequency components are extracted from the stress signal; and pressure peak value, waveform duration, and pulse characteristics are extracted from the pressure signal. Then, a fusion strategy is executed, specifically including feature-level fusion: concatenating the feature vectors of the above three types of signals according to a time window to construct a multi-dimensional feature matrix; or model fusion (such as using LSTM-CNN or Transformer): used for nonlinear multimodal feature modeling and comprehensive judgment. Finally, a unified format comprehensive data structure is output, including thermal / stress / pressure fields and fusion fields.

[0045] S102, acquire thermal response data of beam spot regions of multiple laser sources, and perform fusion analysis on the thermal response data and comprehensive physical field data to obtain multi-laser beam spot position offset characteristic data.

[0046] The beam spot region can refer to the high-energy concentrated area formed by the laser beam on the surface of the printing material or the processing area. It usually has clear spatial geometric characteristics, such as diameter, energy distribution profile, and focal position.

[0047] Thermal response data refers to the temperature change data generated in the beam spot region during laser irradiation, reflecting the heating, conduction and cooling behavior of materials under different laser effects.

[0048] Multi-beam spot position offset characteristic data can represent the set of offset characteristics of multiple laser beams relative to the theoretical path or expected position during actual operation. It reflects the comprehensive influence of system geometric accuracy, thermal interference coupling, vibration error, etc.

[0049] To acquire thermal response data of the beam spot regions from multiple laser sources, a high-resolution infrared thermal imager is first deployed above the printing area. The sampling frame rate must be synchronized with the laser scanning system (e.g., ≥100Hz) to record the thermal diffusion process of the laser beam spot on the printing substrate surface in real time. Simultaneously, a distributed temperature sensing array, such as a thermocouple grid or fiber Bragg grating (FBG) array, is deployed on the printing platform or component surface to accurately acquire temperature changes in the beam spot center and surrounding areas, improving the measurement accuracy of the spatial thermal response. The beam spot region of each laser source is located by associating the scanning path with control parameters (such as power, frequency, and velocity). In the thermal image sequence, the corresponding hot spot region is extracted using temperature threshold segmentation and region growing methods, and the thermal evolution trajectory of each beam spot region is tracked according to the time series. When different laser sources are excited simultaneously, the thermal field overlaps and diffuses nonlinearly in space. Therefore, time synchronization markers and spatial mapping algorithms (such as feature point registration and perspective transformation) are used to align the infrared data with a predefined laser path coordinate system, establishing a one-to-one mapping relationship between the beam spot and the thermal response. Next, to fuse and analyze the thermal response data with the integrated physical field data, the thermal response data and stress and pressure data from strain gauges and pressure sensors are first aligned by timestamps and uniformly mapped onto the coordinate system of the 3D printing model. The joint thermal-stress-pressure field distribution at each point is reconstructed using interpolation or interpolation-regression methods (such as Gaussian processes or cubic splines). Based on this, a multi-physics feature fusion algorithm (such as principal component analysis (PCA) or a multi-source data coupling model based on graph neural networks) is used to extract the coupled physical response characteristics of each laser beam spot region, thereby analyzing the beam spot trajectory offset behavior caused by thermal coupling, non-uniform material response, or beam interference during multi-laser collaborative printing. Finally, by comparing the spatial differences between the actual thermal response trajectory and the expected laser path, multi-laser beam spot position offset characteristic data is obtained. This data may include the offset direction, offset amplitude, offset time, and thermal distribution characteristics of the offset region, serving as the core input for subsequent calibration, compensation, and feedback control.

[0050] The training process for a 3D printed model is as follows:

[0051] The process involves collecting control parameters (laser power, scanning speed, path coordinates, emission angle, etc.) for each laser beam spot during multi-laser printing. Multiphysics data of the substrate surface, including thermal response (temperature field), stress data, and pressure data, are collected synchronously using a sensor array and a high-resolution thermal imager. Geometric error and mechanical property measurement data of the printed product (such as deformation, crack distribution, etc.) are collected as annotation data for the model. The collected sensor data undergoes denoising processing (filtering, wavelet transform, etc.). The thermal, stress, and pressure field data are spatiotemporally aligned and interpolated, mapped to a unified 3D mesh. Key features are extracted, such as local temperature gradients, locations of maximum stress points, and thermal-stress coupling coefficients. A suitable model structure is selected, such as a deep learning-based multiphysics coupling network or a hybrid physics model combining finite element simulation. The inputs (laser parameters, process parameters, multiphysics features) and outputs (beam spot position offset, stress field distribution, thermal field changes, etc.) are determined. The collected experimental and simulation data are used as training sets, and a supervised learning method is employed to train the model. Regularization techniques and cross-validation are introduced to prevent overfitting. The model parameters are optimized using a loss function (such as mean squared error) to ensure a high degree of consistency between the predicted results and actual measurements. Independent test sets are used to verify the model's accuracy and generalization ability. The predicted beam shift is compared with the actual measurement error to evaluate the model's performance. The trained model is integrated into the printing control system for real-time beam shift prediction and correction. Feedback data is collected online to periodically fine-tune and update the model, improving its adaptability.

[0052] S103, Fourier transform processing is performed on the thermal response signal data of multiple laser sources to obtain frequency domain feature data of multiple laser sources, and Fourier transform processing is performed on the multiple laser beam spot position offset feature data and frequency domain feature data to obtain beam spot frequency feature data of multiple laser sources.

[0053] Frequency domain characteristic data refers to the frequency domain representation of a signal obtained by performing a Fourier transform on the thermal response signal data of multiple laser sources. It reflects the amplitude and phase information of each frequency component in the thermal response signal, and can reveal the periodic characteristics, frequency distribution, and dynamic laws of thermal changes under the action of the laser beam spot. It helps to identify oscillations, periodic fluctuations, or abnormal frequency components in the thermal response.

[0054] Beam spot frequency characteristic data refers to the frequency characteristics obtained after further Fourier transform processing of multiple laser beam spot position offset characteristic data and frequency domain characteristic data, reflecting the frequency law of laser beam spot position change over time. These characteristics describe the dynamic frequency response of beam spot position offset, including the periodicity, amplitude and variation trend of the offset, and can be used to analyze the stability, vibration mode and offset type of laser beam spot, thus providing key basis for beam spot position correction and accuracy optimization.

[0055] For the thermal response signal data of each laser source, the Fast Fourier Transform (FFT) algorithm is applied to transform the time-domain signal into frequency-domain feature data. The FFT decomposes the original temperature signal into a superposition of several sinusoidal waveforms, outputting the amplitude and phase information corresponding to each frequency component. To improve the accuracy of the Fourier Transform, the thermal response signal data is typically detrended and weighted using window functions (such as Hamming or Hanning windows) to eliminate spectral leakage caused by edge effects. The transformed result is the frequency-domain feature data of multiple laser sources, containing information such as the dominant frequency, energy density concentration area, harmonic distribution, and spectral width of the thermal response in each beam spot region. These features reveal whether the thermal response signal data exhibits regular perturbations (such as repetitive temperature peaks) and whether it is affected by phenomena such as structural resonance, overlapping scanning paths, or multi-beam interference. Subsequently, the matched multi-laser beam spot position offset feature data (usually obtained through image recognition or laser displacement sensors) is aligned, that is, its sampling time axis is kept consistent with the thermal response signal data. To further reveal the frequency coupling relationship between thermal response signal data and multi-laser beam spot position offset feature data, short-time Fourier transform (STFT) or continuous wavelet transform (CWT) was used to jointly process these two types of data. For example, in the STFT processing, a sliding time window was used to truncate the multi-laser beam spot position offset feature data into multiple short time intervals, and Fourier transforms were performed on each interval to extract frequency features within different time intervals. These features were then cross-compared with the corresponding frequency domain feature data. This method can reveal whether certain specific frequency components coexist in both the thermal response signal data and the position offset feature data, and whether they exhibit a strong correlation. Finally, through frequency matching and correlation analysis, the frequency features that significantly affect the laser beam spot position offset and their performance in the thermal response signal data were extracted, resulting in beam spot frequency feature data for multiple laser sources.

[0056] Based on the above technical solution, optionally, Fourier transform processing is performed on the multi-laser beam spot position offset feature data and frequency domain feature data to obtain beam spot frequency feature data of multiple laser sources, including:

[0057] The frequency domain feature data and the multi-laser beam spot position offset feature data are processed using adaptive frequency domain filtering technology to obtain optimized frequency domain feature data;

[0058] Kalman filtering and wavelet transform were used to remove noise from the optimized frequency domain feature data, resulting in beam spot frequency feature data for multiple laser sources.

[0059] In this scheme, adaptive frequency domain filtering technology can be a signal processing method that automatically adjusts the filtering parameters based on the frequency characteristics of the input signal. It is often used to extract effective information from complex or dynamically changing frequency domain signals and suppress irrelevant frequency components or noise interference.

[0060] Kalman filtering can be a recursive optimization algorithm, primarily used to estimate the state of dynamic systems under noisy conditions. It integrates system model predictions with sensor observations to calculate the optimal estimate.

[0061] Wavelet transform is a signal analysis method that has both time and frequency resolution capabilities. It can decompose a signal into local waveforms of different scales, making it suitable for detecting transient changes, local anomalies, or edge features in a signal.

[0062] In the processing, the acquired frequency domain feature data and multi-laser beam spot position offset feature data are first synchronized in the time domain through frequency-time axis alignment to facilitate subsequent fusion analysis. After synchronization, adaptive frequency domain filtering technology is used to process the spectral information of the two signal sources. Specifically, power spectral density analysis is first performed on the two sets of signals to identify the concentrated spectral energy regions and dominant frequency components, and their correlation with the system target frequency (such as laser scanning frequency and beam spot offset oscillation frequency) is evaluated. Based on this, a dynamically adjustable bandpass filter window is constructed, with variable bandwidth and center frequency set. The least mean square (LMS) adaptive algorithm is used to adjust the filter parameters in real time, thereby retaining the frequency bands with high correlation to the analysis and filtering out unrelated frequency interference with low signal-to-noise ratio. At the same time, frequency domain feature entropy analysis and distributed energy threshold judgment mechanism can be introduced in this process to further enhance the filter's ability to identify multi-source interference, resulting in optimized frequency domain feature data with a clearer structure and more representative characteristics. Subsequently, Kalman filtering and wavelet transform are used in combination for noise cancellation and frequency domain compensation processing of the optimized signal data. First, the signal is decomposed using wavelet multi-scale decomposition (e.g., using Daubechies wavelets or the Symlet wavelet family) to separate different frequency sub-bands. Soft threshold compression is applied to the high-frequency portion to suppress short-term burst noise and systematic glitches. Simultaneously, a Kalman filter is introduced into the low-frequency portion. The current sub-band frequency trend is predicted using the state transition equation, and optimal estimation is performed based on current observations to dynamically correct signal trend drift and background interference. After the wavelet-filtered signal is reconstructed, it undergoes full-cycle smoothing using a Kalman smoother to improve spectral structure continuity and suppress residual oscillations. Finally, key indicators such as the dominant frequency distribution, harmonic location, and energy density concentration segment of the spectrum after multi-dimensional noise reduction are extracted to form beam spot frequency characteristic data corresponding to multiple laser sources. This provides crucial input for subsequent offset identification, frequency drift diagnosis, and control compensation strategies.

[0063] This solution effectively eliminates environmental interference and system noise, extracts frequency features that truly reflect the dynamic characteristics of the laser beam spot, thereby improving the accuracy and stability of multi-laser source beam spot offset identification and providing high-quality data support for subsequent precision compensation and dynamic control.

[0064] S104, obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data and frequency domain characteristic data to obtain the verification standard data.

[0065] The adaptive thermal response threshold can be a dynamic reference standard value that is dynamically adjusted based on real-time changes in thermal response during the laser printing process, material thermal properties, and environmental conditions (such as temperature, humidity, interference, etc.) to determine whether the thermal response is normal or deviates.

[0066] The calibration standard data can be a comprehensive set of benchmark data used to determine whether the laser beam spot accuracy meets the requirements. It is composed of multiple indicators such as adaptive thermal response threshold, beam spot frequency characteristic data, and frequency domain characteristic data, reflecting the thermal stability, vibration characteristics, and positional consistency of each beam spot in a multi-laser system.

[0067] During the execution of multiple laser source printing tasks, the system monitors the thermal response signal data of each beam spot region in real time, recording the temperature changes over time and forming a continuous thermal response curve. For the thermal response curve of each beam spot region, data preprocessing is first performed, including signal denoising, baseline drift correction, and outlier removal, to ensure the high reliability of the subsequently extracted features. Subsequently, by statistically analyzing key feature parameters of the thermal response curves in the current and historical printing cycles, including average temperature rise, peak response time, cooling slope, and thermal response oscillation period, a sample distribution of the thermal response behavior of that beam spot under stable processing conditions is formed. Based on this distribution, combined with statistical analysis methods (such as the sliding window averaging method, empirical distribution function, and autoregressive moving average model ARMA), the upper and lower limits of the allowable thermal response in the current cycle are dynamically calculated to obtain a time-dependent basic thermal threshold range. To achieve dynamic adjustment to adapt to different materials and environmental conditions, externally collected environmental temperature disturbance data, material thermal diffusivity, and processing path complexity are used as correction factors and input into a preset dynamic thermal response adjustment model. This model can employ a numerical simulation model based on the heat conduction equation or a trained regression network to output an adaptive thermal response threshold applicable under current conditions, which takes the form of a dynamic threshold function:

[0068]

[0069] in, This represents the average thermal response for the current period. The standard deviation of the thermal response; This refers to the change in ambient temperature. These are the preset material thermal property parameters; The system is adjusted to the current standard deviation of the thermal response based on the preset thermal response fluctuation sensitivity coefficient. The response strength is used to measure the system's tolerance to the amplitude of thermal fluctuations; The preset environmental disturbance compensation coefficient is used to adjust the adaptive threshold for changes in external ambient temperature. Its adaptability is used to dynamically compensate for thermal changes caused by air temperature, humidity or convection conditions; These are preset material thermal property weighting coefficients, used to adjust preset material thermal property parameters. Contribution to adaptive thermal response threshold calculation.

[0070] Next, the obtained adaptive thermal response threshold is fused with the frequency domain feature data and beam spot frequency feature data previously extracted using frequency domain analysis methods such as Fourier transform. Specifically, a fusion algorithm based on rule matching and Bayesian inference is used to compare whether the dominant frequency in the frequency domain exceeds the influence range of the adaptive threshold, and to identify whether there are abnormal features such as resonance, laser cross thermal coupling, and periodic thermal disturbance.

[0071] In this processing flow, a multivariate feature vector set F=[f1, f2, ..., fn] is constructed, where each dimension represents a numerical indicator of a certain thermal frequency domain feature, such as frequency offset, dominant frequency energy, amplitude mutation rate, etc. A trained classifier (e.g., support vector machine or random forest) is used to determine whether the thermal response features are in a normal state based on this vector, and its corresponding beam spot frequency response mode and offset type are labeled.

[0072] Ultimately, the system fuses adaptive thermal response thresholds, frequency domain characteristic data, and beam spot frequency characteristic data to form complete verification standard data. This data structure can include the thermal response safety range for each laser source; the dominant frequency range; offset trend prediction labels; and anomaly risk scores. This data will be written into the system's precision control logic in real time, serving as the standard basis for subsequent beam spot position correction or anomaly alarm judgment, thereby significantly improving the reliability and dynamic response capability of laser beam spot control.

[0073] Based on the above technical solution, optionally, obtaining the adaptive thermal response threshold includes:

[0074] Acquire power data, scanning speed data, material thermal conductivity data, material specific heat capacity data, material density data, and external temperature data from multiple laser sources. Based on these data and a preset adaptive thermal response threshold calculation formula, calculate the adaptive thermal response threshold. The preset adaptive thermal response threshold calculation formula is as follows:

[0075]

[0076] in, Adaptive thermal response threshold; This is the preset thermal response sensitivity coefficient; Power data; This refers to the thermal conductivity data of the material. This refers to the specific heat capacity data of the material. For material density data; This is for scan speed data; External temperature data; This is the preset ambient temperature suppression and adjustment factor.

[0077] In this scheme, power data refers to the energy emitted by each laser source per unit time, usually expressed in watts (W). It reflects the energy intensity of the laser when it acts on the material surface and is a key parameter affecting the material's temperature rise and thermal diffusion behavior.

[0078] Scanning speed data refers to the speed at which the laser beam moves through the processing area, typically measured in millimeters per second (mm / s). It affects the energy density received per unit area, thus determining heat accumulation and molten pool behavior.

[0079] Thermal conductivity data represents a material's ability to conduct heat, and is measured in W / (m·K). The higher the thermal conductivity, the faster the material dissipates heat.

[0080] Specific heat capacity of a material represents the amount of heat required to raise the temperature of a unit mass of material by a unit, expressed in J / (kg·K). It determines the rate at which the material's temperature rises after an input of heat energy.

[0081] Material density data refers to the mass per unit volume of a material, expressed in kg / m³. In thermal response calculations, it is used in conjunction with parameters such as specific heat capacity and volume to calculate the temperature change of a material.

[0082] External temperature data refers to the ambient temperature in the material processing environment, typically expressed in degrees Celsius (°C) or Kelvin (K). External temperature affects the initial thermal field distribution and heat dissipation paths.

[0083] To calculate the adaptive thermal response threshold, the emission power of each laser channel is first monitored in real time using an integrated photodetector array in the laser system, acquiring power data from multiple laser sources in watts (W). The photodetector utilizes the characteristic of converting incident light into electrical signals to achieve high-frequency sampling, and records this data as time-series energy data suitable for calculation via an analog-to-digital converter circuit.

[0084] Next, scanning speed data of multiple laser sources are obtained from the servo control encoder system. This system records the angular velocity and displacement trajectory of the scanning lens or galvanometer in real time based on the encoder, and converts it into the linear velocity of the beam spot on the processing surface (in mm / s) to reflect the scanning efficiency of the laser beam at different positions.

[0085] Then, depending on the type of printing material (e.g., metal powder, polymer, or ceramic), the system retrieves the corresponding thermal conductivity data (W / m·K), specific heat capacity data (J / kg·K), and density data (kg / m³) from the built-in material thermophysical property database. These data are derived from experimental data or material handbooks (such as ASTM and NIST) and are used to model thermal diffusion behavior.

[0086] For ambient temperature monitoring, high-precision thermocouple sensors or infrared temperature probes installed near the printing chamber or platform measure the ambient air temperature in real time to obtain external temperature data (in °C). This data is used to reflect the environmental thermal conduction boundary conditions.

[0087] The adjustment coefficient α, as a thermal response sensitivity coefficient, is used to adjust the output amplitude of the thermal response threshold. It is typically determined through experimental calibration, i.e., under typical material and process conditions, recording the actual measured maximum safe temperature rise and fitting it with the theoretical value; alternatively, it can be normalized based on the material's thermal density, or optimized using historical data as a trainable parameter. As an ambient temperature suppression and regulation factor, it is used to control the exponential decay effect of ambient temperature on the thermal response threshold, preventing misjudgments caused by drastic temperature changes. Its value is generally determined through empirical fitting or dynamic feedback mechanisms based on the ambient temperature range designed for the system, with common values ​​ranging from several hundred to several thousand. Reasonable settings of both ensure that the thermal response threshold formula accurately reflects the thermal dynamic characteristics of the laser processing process under different material, process, and environmental conditions, effectively assisting in overheat warning and coupled interference control. It can be set through empirical fitting or dynamic feedback mechanisms to control the exponential decay effect of ambient temperature on the thermal response threshold. The preset values ​​of both, combined with thermal conductivity characteristics, material parameters, and system operating conditions, ensure the accuracy and adaptability of the thermal response threshold under multi-source interference and complex environments, thereby improving the thermal control precision and stability during laser processing.

[0088] This solution effectively reflects the actual thermal state of materials and environment during laser processing, improves the real-time adaptability and accuracy of the thermal response threshold, helps to adjust laser parameters in a timely manner, prevents overheating and thermal deformation, ensures processing quality and equipment safety, and improves the overall stability and reliability of the system.

[0089] S105, acquire external temperature data, external humidity data and external interference data, determine environmental adaptability data based on the external temperature data, external humidity data and external interference data, combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data.

[0090] External temperature data refers to the real-time changes in the ambient air temperature around the printing environment, collected by temperature sensors when the multi-laser source system is running.

[0091] External humidity data can represent the relative humidity (RH) in the printing environment, usually expressed as a percentage (%).

[0092] External interference data can refer to disturbance information from non-ideal external environments, including airflow disturbances (cooling wind, external wind speed), vibration information (ground vibration, equipment resonance), electromagnetic interference (laser equipment or power supply fluctuations), and abnormal laser reflection / scattering (due to powder accumulation, wall interference).

[0093] Environmental adaptability data can be a multidimensional environmental stability assessment index calculated using a fusion algorithm (such as weighted average, fuzzy logic, or neural network regression) based on the aforementioned external temperature data, external humidity data, and external disturbance data. It reflects the system's ability to adapt to current external conditions.

[0094] The optimized signal processing data can be comprehensive laser control and calibration control reference data formed after integrating multi-source information such as environmental adaptability data, thermal response data, frequency domain characteristic data, and threshold parameters, and after processing such as filtering, compensation, modeling, and prediction.

[0095] To acquire external temperature, humidity, and interference data, the system first deploys high-precision environmental sensing modules in multiple laser processing areas. These modules include thermocouples or infrared temperature sensors, industrial-grade humidity sensors (such as DHT22 and SHT35), and multi-source interference monitoring devices (such as MEMS accelerometers, airflow sensors, and EMC electromagnetic interference monitors). External temperature data is collected at the laser printing area boundaries using high-sampling-frequency thermoelectric sensors to capture ambient temperature change curves, typically updating at frequencies above 10Hz, and recording instantaneous temperature and historical fluctuations. External humidity data is obtained by continuously measuring the water vapor content in the air and converting it to a relative humidity percentage, reflecting the potential impact of air on heat conduction and powder hygroscopicity. External interference data is collected using multi-source heterogeneous sensors to capture non-thermal environmental interference, including: base vibration spectrum (acquired by a triaxial accelerometer), airflow disturbance amplitude (obtained by an airflow differential pressure sensor), and electromagnetic fluctuations (monitored by a radio frequency interference analysis module).

[0096] Subsequently, the system utilizes a set of multivariate fusion models to jointly analyze the three types of environmental factors, generating environmental adaptability data. This process first performs normalization and noise reduction preprocessing on various types of raw data. For example, a moving average filter is used to remove instantaneous fluctuations in temperature and humidity data, and Kalman filtering is used to smooth acceleration data. Then, principal component analysis (PCA) or environmental index calculation based on a weighted model is performed. In the weighted model, the external temperature change rate, relative humidity fluctuation amplitude, and disturbance intensity index are assigned three dynamically adjusted weight coefficients λ3, λ4, and λ5, respectively, to calculate a unified environmental adaptability score index.

[0097] E_adapt=λ3·ΔT_norm+λ4·ΔRH_norm+λ5·D_env

[0098] Here, E_adapt is the environmental adaptability scoring index; ΔT_norm represents standardized temperature fluctuation, ΔRH_norm represents standardized humidity change, and D_env represents the interference intensity score; λ3 is a preset temperature sensitivity weighting coefficient, representing the intensity of the impact of environmental temperature changes on the thermal stability of laser processing; λ4 is a preset humidity coupling weighting coefficient, representing the degree of influence of humidity on laser-material interaction, powder moisture absorption deformation, and other effects; and λ5 is a preset interference robustness weighting coefficient, representing the comprehensive influence intensity of environmental interference (such as vibration, electromagnetic interference, and airflow). Finally, the system performs multi-dimensional feature fusion of this environmental adaptability data with the established calibration standard data, evaluates its impact level on laser accuracy through a fuzzy rule system or BP neural network, and outputs a correction parameter matrix. This matrix serves as the weight compensation input, is superimposed with the ideal frequency domain response and thermal threshold window recorded in the calibration standard data, and outputs the final optimized signal processing data as the real-time input of the laser beam spot control system, improving the laser processing system's resistance to environmental disturbances and response stability.

[0099] S106: Acquire real-time laser beam spot accuracy data from multiple laser sources, and combine the real-time laser beam spot accuracy data with the optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

[0100] Real-time laser beam spot accuracy data can be raw data that reflects the spatial characteristics and stability of the current output beam spot of each laser source, obtained through an online monitoring system (such as a high-speed camera system, displacement sensor, laser interferometer, etc.) during multi-laser 3D printing or processing.

[0101] The multi-laser beam accuracy data can be the optimal accuracy value of the multi-laser beam used for control system correction, obtained by fusing the above real-time data with the optimized signal processing data (such as thermal response prediction, interference compensation results, frequency domain characteristic trends, etc.) obtained from the previous modeling, and then correcting it with algorithms (such as Kalman filtering, Bayesian optimization, etc.).

[0102] High-frame-rate visual monitoring systems (such as industrial cameras or high-speed infrared imaging systems), laser displacement sensors, or laser interferometry devices can be deployed on laser processing or multi-laser printing platforms. These devices acquire real-time behavioral characteristics of each laser beam spot within the processing area, including spot position coordinates, spot size, energy density distribution, edge contours, and spot center drift trajectory. By processing the raw images or sensor data using image processing algorithms (such as edge detection, shape fitting, and sub-pixel-level feature extraction), the spatial geometric parameters and energy distribution characteristics of each laser beam spot at a specific time point can be extracted, forming structured real-time laser beam spot accuracy data.

[0103] Subsequently, the real-time laser beam spot accuracy data is fused with the optimized signal processing data previously obtained through modeling. This signal processing data typically includes the theoretical thermal response characteristics of the laser beam spot, frequency domain interference modes, historical offset trends, and environmental disturbance compensation factors. To fuse the two types of data, a time synchronization mechanism (such as aligning the timestamps of the two data sources by triggering a clock or coded signal) is required during the data alignment phase to ensure that the two sets of data have a consistent timing reference.

[0104] In the fusion processing stage, a multi-channel Kalman filter algorithm is used to filter and estimate the state of the accuracy data streams from each laser source. This algorithm can dynamically correct random errors caused by system delay, measurement noise, and external disturbances, thereby outputting a more stable laser beam spot accuracy state that closely approximates the true value. Furthermore, wavelet transform can be introduced to decompose and suppress anomalous frequency changes or noise components, further enhancing the robustness of data fusion. Finally, the filtered and corrected output is the calibrated multi-laser beam spot accuracy data.

[0105] In this embodiment, multi-physics field signal data from multiple laser sources are acquired, and the multi-physics field signal data is synchronously acquired and fused to obtain comprehensive physical field data. The multi-physics field signal data includes thermal response signals, stress signals, and pressure signals. Thermal response data of the beam spot regions of multiple laser sources are acquired, and the thermal response data is fused and analyzed with the comprehensive physical field data to obtain multi-laser beam spot position offset characteristic data. Fourier transform processing is performed on the thermal response signal data of multiple laser sources to obtain frequency domain characteristic data of multiple laser sources. Fourier transform processing is then performed on the multi-laser beam spot position offset characteristic data and the frequency domain characteristic data. The method involves obtaining beam frequency characteristic data from multiple laser sources; acquiring an adaptive thermal response threshold; combining the adaptive thermal response threshold, beam frequency characteristic data, and frequency domain characteristic data to obtain calibration standard data; acquiring external temperature data, external humidity data, and external interference data; determining environmental adaptability data based on these data; combining the environmental adaptability data with the calibration standard data to obtain optimized signal processing data; and acquiring real-time laser beam accuracy data from multiple laser sources. This real-time laser beam accuracy data is then combined with the optimized signal processing data to obtain calibrated multi-laser beam accuracy data. This multi-laser 3D printing beam accuracy calibration method improves the stability, consistency, and intelligent response capabilities of multi-laser systems in high-precision manufacturing scenarios, contributing to improved printing quality, reduced defect rates, and extended equipment lifespan.

[0106] Based on the above technical solution, optionally, after obtaining the calibrated multi-laser beam accuracy data, the method further includes:

[0107] Real-time accuracy data of each laser beam spot is acquired, and the real-time accuracy data and the optimized signal processing data are fused and corrected using the Kalman filter algorithm to obtain the corrected multi-laser beam spot accuracy data.

[0108] The corrected multi-laser beam spot accuracy data is input into the preset multi-laser beam spot interaction effect model to update the multi-laser beam spot accuracy data.

[0109] In this solution, real-time accuracy data refers to the specific performance indicators of each laser beam spot monitored in real time during laser 3D printing or multi-laser processing, such as beam spot position deviation, size variation, shape distortion, and offset. This data is typically acquired in real time using devices such as high-speed cameras, optical sensors, and displacement sensors, reflecting the actual state and accuracy of the current laser beam spot.

[0110] The Kalman filter algorithm can be considered an optimal filtering algorithm based on recursive estimation, used to fuse measurement data from multiple sources, filter out noise and errors, and obtain more accurate state estimates. In this application, the Kalman filter algorithm effectively corrects and adjusts the accuracy parameters of the laser beam spot by combining real-time accuracy data and optimized signal processing data, achieving dynamic data fusion and error compensation, and improving the overall accuracy of measurement and control.

[0111] The pre-defined multi-laser beam interaction model can be a pre-established mathematical or physical model used to describe the interaction and interference effects between multiple laser beams, including complex interaction mechanisms such as phase difference, energy overlap, and thermal coupling. Based on multi-dimensional data such as laser beam power, wavelength, spatial location, and thermal response, this model can predict and adjust the dynamic behavior and accuracy of beams in a multi-laser system, supporting subsequent calibration and control strategies.

[0112] Real-time accuracy data for each laser beam spot is acquired, specifically by using high-speed optical sensors, laser displacement sensors, and industrial cameras to collect key parameters such as spatial position, size, shape characteristics, and offset of each laser beam spot in real time. During the acquisition process, a preprocessing module denoises, corrects, and synchronizes the raw data to ensure accuracy and timeliness. Subsequently, the preprocessed real-time accuracy data, along with optimized signal processing data obtained through physical modeling, experimental correction, or signal processing, is input into the Kalman filter algorithm module. The Kalman filter algorithm, based on recursive estimation, first predicts the current state of the laser beam spot according to the system dynamic model, then updates the state by combining observation data. Through repeated iterations, it eliminates the influence of noise and measurement errors, achieving real-time data fusion and correction, and finally outputting corrected multi-laser beam spot accuracy data. Next, this corrected accuracy data is input into an established multi-laser beam spot interaction effect model, which comprehensively considers multi-physics interaction factors such as the spatial relative position, phase difference, interference effect, thermal coupling, and energy distribution between laser beam spots. Based on the current corrected data, the model dynamically calculates the interaction effects of laser beams, updates the accuracy parameters and position status of each beam, and achieves real-time optimization and adjustment of the overall accuracy of the multi-laser system.

[0113] The training process for the established multi-laser beam spot interaction effect model is as follows:

[0114] Extensive experimental data on multi-beam laser systems under various operating conditions were collected, encompassing multi-dimensional information such as the spatial location, optical power, wavelength, phase difference, temperature variation, scanning speed, and thermophysical properties of the materials. High-precision sensor equipment and a high-speed data acquisition system ensured the time-series synchronization and accuracy of the data. Secondly, physical simulation software (such as finite element analysis and optical simulation tools) was used to numerically simulate the interference effects, thermal coupling, and energy distribution of the laser beams, generating simulated data to supplement the experimental data and enhance the diversity and coverage of the training samples. Then, the experimental and simulated data underwent preprocessing, including denoising, normalization, feature extraction, and multi-physics signal fusion, resulting in a structured input feature set. Subsequently, appropriate machine learning algorithms, such as deep neural networks, convolutional neural networks, or graph neural networks, were selected, and the network structure was designed to effectively capture the complex nonlinear interactions between laser beams. During model training, a supervised learning approach was adopted, using historically measured laser beam accuracy data as labels, combined with loss functions (such as mean square error and cross-entropy) for iterative optimization. The model parameters were adjusted using a backpropagation algorithm to gradually improve prediction accuracy. Meanwhile, cross-validation and early stopping are used to avoid overfitting and ensure the model's generalization ability. After training, offline validation and online testing are performed to compare the model's predictions with actual measurement data. The model structure or training strategy is then adjusted to further improve the model's accuracy and robustness. Finally, the trained multi-laser beam spot interaction effect model is deployed to achieve accurate simulation and prediction of the dynamic behavior of laser beam spots, supporting real-time accuracy optimization and system adaptive control.

[0115] In this scheme, by acquiring real-time accuracy data of each laser beam and combining it with optimized signal processing data, a Kalman filter algorithm is used for fusion and correction. This effectively filters out noise and measurement errors, improving the accuracy and stability of the data. The corrected accuracy data is then input into a preset multi-laser beam interaction effect model to achieve dynamic prediction and correction of the mutual influence of laser beams. This significantly improves the overall system accuracy and response speed, ensuring the stability and processing quality of multi-laser source collaborative operation.

[0116] Example 2

[0117] Figure 2 This is a schematic flowchart of the multi-laser 3D printing beam spot verification method provided in Embodiment 2 of this application. Figure 2 As shown, the specific steps include the following:

[0118] S201, acquire multi-physics field signal data from multiple laser sources, and perform synchronous acquisition and fusion processing on the multi-physics field signal data to obtain comprehensive physical field data; wherein, the multi-physics field signal data includes thermal response signal, stress signal and pressure signal.

[0119] S202, acquire thermal response data of beam spot regions of multiple laser sources, and perform fusion analysis on the thermal response data and comprehensive physical field data to obtain multi-laser beam spot position offset characteristic data.

[0120] S203, Fourier transform is performed on the thermal response signal data of multiple laser sources to obtain frequency domain feature data of multiple laser sources, and Fourier transform is performed on the beam spot position offset feature data and frequency domain feature data of multiple laser sources to obtain beam spot frequency feature data of multiple laser sources.

[0121] S204, obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data and frequency domain characteristic data to obtain the verification standard data.

[0122] S205, acquire external temperature data, external humidity data, and external interference data, determine environmental adaptability data based on the external temperature data, external humidity data, and external interference data, and combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data.

[0123] S206: Acquire real-time laser beam spot accuracy data from multiple laser sources, and combine the real-time laser beam spot accuracy data with the optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

[0124] S207, acquire real-time velocity data of each laser beam spot, and calculate the velocity error of each laser beam spot based on the real-time velocity data.

[0125] A laser beam spot refers to the area of ​​light formed when a laser beam strikes a material surface. In laser processing, laser printing, or laser cladding, the size, shape, position, and energy distribution of the beam spot directly affect the processing quality and precision. Each laser source corresponds to one or more beam spot areas, and their behavior and precision are crucial for the coordination of multiple laser systems.

[0126] Real-time velocity data refers to the instantaneous velocity information acquired by the system based on displacement sensors, scanning control feedback, or image processing technology as the laser beam spot moves over time. The unit is typically millimeters per second (mm / s) or micrometers per second (μm / s). This data reflects the beam spot's movement speed in each printing cycle or scanning path, facilitating dynamic adjustment of the beam trajectory, error compensation, or analysis of vibration behavior.

[0127] Velocity error refers to the difference between the ideal (set) beam pattern motion velocity and the actual real-time velocity acquired. It is typically used to reflect the velocity deviation of the system caused by factors such as control, environmental disturbances, or mechanical lag. This error is an important basis for control compensation, dynamic adjustment, and trajectory correction.

[0128] During the printing process, the system uses high-precision displacement sensors (such as laser interferometers, encoders integrated into scanning galvanometers, or vision tracking systems) deployed near the laser processing head to monitor the position of the laser beam spot in the working area over time. By continuously recording the beam spot position information at multiple moments and combining it with a unified timestamp management mechanism within the system, the movement speed of the beam spot in different time periods can be calculated; this is the so-called real-time speed data. Simultaneously, the system pre-sets target speed values ​​for each laser beam spot on different processing path segments. These target speeds are automatically generated based on process requirements, material properties, and path planning, representing the ideal movement speed that each laser beam spot should achieve. After receiving the real-time speed data collected by the sensors, the control system compares this data with the target speed values ​​at the corresponding time points. The difference between the two is the speed error, reflecting the degree of deviation between the actual movement and the ideal movement.

[0129] S208, acquire real-time temperature change data, input the real-time temperature change data into a preset thermal response model, and obtain the thermal response coefficient.

[0130] Real-time temperature change data can refer to the data on the temperature change of the material surface or beam spot area over time, which is collected in real time by thermocouples, infrared thermal imagers, thermistors, thermopile or other temperature sensing devices during laser processing or printing.

[0131] The preset thermal response model can be a mathematical or simulation model used to describe the heat conduction, heat diffusion and heat accumulation processes of materials under laser irradiation. It is usually established in advance based on material properties (such as thermal conductivity, specific heat capacity and density) and used to predict the effect of temperature changes on the thermal behavior of materials.

[0132] The thermal response coefficient can be a parameter used to quantify the sensitivity of a material to laser thermal excitation, reflecting the degree of thermal expansion, property change, or structural deformation of the material under unit temperature change. It can be a scalar, vector, or polynomial set of coefficients.

[0133] Infrared thermal imagers (such as short-wave infrared cameras), miniature thermocouples, or thermopile sensors can be deployed in each laser beam spot area of ​​the processing platform. These sensors continuously acquire temperature data of the processed surface at high frequencies (e.g., 1~10kHz), thereby recording the temperature change curve of the laser-irradiated area over time in real time, i.e., acquiring real-time temperature change data. This data is typically represented as a temperature time series for each laser source beam spot, in a format such as T1(t), T2(t), ..., T n(t). The acquired raw temperature data undergoes preprocessing, including noise reduction (such as median filtering), sampling rate unification, and time alignment, to ensure spatiotemporal consistency of data across different sensors. This processed real-time temperature change data is then input into a predefined, preset thermal response model within the system. This thermal response model can take one of three typical forms:

[0134] Analytical model: Based on the one-dimensional / two-dimensional Fourier heat conduction equation, considering laser power density, beam radius, initial temperature boundary conditions of the material, etc., the material temperature rise rate per unit time is calculated;

[0135] Finite element heat conduction model: Two-dimensional or three-dimensional thermal models are constructed using software such as COMSOL and ANSYS. Based on input parameters such as material thermal conductivity, specific heat capacity, and density, the thermal diffusion behavior caused by laser irradiation is simulated.

[0136] The data-driven neural network model takes multi-channel time-series temperature data as input and outputs local thermal conductivity or temperature rise response index. This model can be obtained through supervised learning using historical processing samples.

[0137] When the model is running, real-time temperature change data is input in time series, and combined with the current thermophysical parameters of the material (such as thermal conductivity, density, and specific heat capacity), the response degree of the material under local unit temperature perturbation is solved according to the internal heat conduction mechanism or fitting relationship of the model, and then the thermal response coefficient is output.

[0138] The thermal response coefficient can be expressed as local thermal diffusivity (distance of temperature propagation per unit time); thermal inertia factor (characterizing the relationship between temperature rise and laser irradiation time); and sensitivity index of temperature change to material deformation or laser deflection.

[0139] S209: Obtain the current time data and the initial laser position of each laser beam spot; input the real-time velocity data, velocity error, thermal response coefficient, initial laser position and current time data into the preset beam spot position calculation formula; calculate the adjusted beam spot position of each laser beam spot; and transmit the adjusted beam spot position to the laser beam spot control system.

[0140] Current time data can refer to the system's real-time timestamp at the moment of beam spot position correction, usually in seconds (s), milliseconds (ms), or even microseconds (μs), recording the specific time point of the laser processing system within a certain processing cycle.

[0141] The initial laser position can refer to the reference coordinates in space of each laser beam spot set or measured by the system at the beginning of a certain printing / processing cycle, usually expressed as two-dimensional or three-dimensional coordinates.

[0142] The beam position can be adjusted based on factors such as real-time velocity data, velocity error, and thermal response coefficient to determine the current optimal target position of each laser beam spot, in order to counteract the effects of temperature drift, velocity disturbance, and thermal expansion during the processing.

[0143] The laser beam spot control system is a subsystem in the entire laser processing platform used to precisely control parameters such as the laser beam spot position, power, and scanning path. It consists of both hardware and software.

[0144] In the laser processing process, in order to achieve high-precision, multi-laser beam spot synchronous control, the system needs to acquire the current time data and the initial laser position of each laser beam spot, and input them together with the real-time sensing parameters into the beam spot position calculation model, so as to dynamically calculate the adjustment of the beam spot position and transmit it to the laser beam spot control system for real-time correction.

[0145] Specifically, the system first acquires current time data using an internal high-precision synchronous clock module or embedded time synchronization controller. This data records the moment when the current laser beam control command takes effect with a high-resolution timestamp (e.g., nanoseconds or microseconds), ensuring strict timing consistency between all data sampling and control signals. Simultaneously, at each printing cycle or initial processing segment, the system obtains the initial laser position of each laser beam spot through a laser path initialization program or displacement feedback mechanism. This initial position data is typically represented in three-dimensional coordinates and calibrated using a preset trajectory of the laser scanning path, derived from a laser position encoder, a high-resolution displacement sensor, or an image recognition-based beam spot repositioning algorithm.

[0146] After acquiring the aforementioned static and dynamic inputs, the system inputs real-time velocity data (real-time beam velocity obtained from scanning galvanometer feedback or laser scanning trajectory estimation), velocity error (calculated from the difference between the actual velocity and the desired velocity), thermal response coefficient (obtained from real-time temperature change data input into the thermal response model), initial laser position, and current time data into a preset beam position calculation formula. This calculation formula integrates nonlinear acceleration terms, temperature drift compensation terms, and time coupling terms, and can output the compensated adjusted beam position in real time, i.e., the precise coordinates to which the current laser beam should move, to compensate for deviations caused by factors such as temperature rise, inertia, and interference coupling.

[0147] Subsequently, the obtained adjusted beam position is transmitted to the laser beam control system via a system bus (e.g., EtherCAT, CAN, or FPGA channel) in the form of digital control commands (such as G-code or control voltage vector). This control system analyzes the input coordinate values ​​in real time and drives the scanning galvanometer system or motion platform actuator to perform micron-level position corrections, ensuring that each laser beam moves accurately within the processing area along the optimized path, thereby significantly improving overall processing quality and morphological consistency.

[0148] In this embodiment, by acquiring real-time speed and temperature change data of each laser beam spot, and combining thermal response model and time synchronization information, the adjustment position of the beam spot is accurately calculated, and the adjustment result is fed back to the control system in real time, thereby realizing dynamic compensation and precise control of the laser beam spot, effectively reducing positioning errors caused by speed fluctuations and temperature changes, and improving the accuracy, stability and finished product quality of laser processing.

[0149] Based on the above technical solution, the optional, preset formula for calculating the beam spot position is as follows:

[0150]

[0151] in, To adjust the beam spot position; This is the initial laser position; This is real-time speed data; For speed error; Data is for the current time. This is the preset acceleration adjustment coefficient; This is the preset speed adjustment coefficient.

[0152] In this scheme, during actual processing, multiple typical operating conditions (such as different powers, scanning speeds, and material types) can be selected to record the deviation between the actual beam spot position and the ideal trajectory. By comparing the trend of the deviation over time, a nonlinear trend curve of the beam spot offset is fitted. Curve fitting tools (such as least squares or polynomial fitting) are then used to determine... , The value that the formula predicts The location is as close as possible to the measured trajectory. It reflects the degree to which acceleration affects position. It reflects the degree of correction for linear velocity drift.

[0153] Based on the above technical solution, optionally, after transmitting the adjusted beam spot position to the laser beam spot control system, the method further includes:

[0154] After each preset update time interval, the real-time velocity data of each laser beam spot is reacquired, and the velocity error of each laser beam spot is calculated based on the real-time velocity data.

[0155] Reacquire real-time temperature change data and input the real-time temperature change data into a preset thermal response model to obtain the thermal response coefficient;

[0156] The system reacquires the current time data and the initial laser position of each laser beam spot. It inputs the real-time speed data, speed error, thermal response coefficient, initial laser position, and current time data into the preset beam spot position calculation formula to calculate the adjusted beam spot position of each laser beam spot. The adjusted beam spot position is then transmitted to the laser beam spot control system until the laser printing process is completed.

[0157] In this solution, the preset update time interval can refer to the time interval during which the system automatically triggers a dynamic parameter update and adjustment operation once according to a fixed time period (e.g., every 10 milliseconds, 50 milliseconds, or 1 second) during the laser printing process.

[0158] During laser printing, to ensure the laser beam position remains highly accurate and to dynamically respond to complex factors such as thermal deformation, speed deviation, and time lag, the system automatically initiates a closed-loop calibration update process every preset update interval (e.g., every 10ms, 20ms, or shorter real-time cycles). This process achieves continuous optimization and real-time correction of the laser beam position through high-frequency data acquisition, dynamic model calculation, and control command updates.

[0159] First, after the timer or real-time control thread in the control system triggers the update signal, the system immediately acquires the real-time velocity data of the current laser beam spot from each laser scanning path. This data can be acquired in real time through a high-speed optical encoder, galvanometer feedback device, or position-velocity closed-loop control module in the motor driver to obtain the actual movement velocity of each laser beam spot at the current sampling point in the XY axis or three-dimensional space. At this time, the previously stored target velocity curve (generated according to the printing path planning) is also called to calculate the current velocity error of each beam spot, that is, the deviation between the measured velocity and the theoretical velocity, which is used to compensate for dynamic drift in trajectory tracking.

[0160] Subsequently, the system synchronously collects real-time temperature change data, which comes from thermocouple sensors, infrared temperature probes, or high-frame-rate thermal imagers deployed in the printing work area, accurately reflecting the subtle changes in the temperature of the printing material and the local environment over time. The acquired temperature change curves are input into the system's embedded thermal response model. This model is constructed based on heat conduction-diffusion theory or experimental fitting curves, taking into account the thermal conductivity, heat capacity, density, and other properties of the material. By calculating the thermal response coefficient of each beam spot under the current conditions, it represents the degree of positional shift that the region where the beam spot is located may experience under thermal disturbance.

[0161] Next, the control system acquires the current system timestamp and the laser dot positions of each laser beam spot in the initial state (which can be obtained through initial displacement calibration or trajectory recording). This time information, the initial laser position of each beam spot, real-time velocity, velocity error, and thermal response coefficient are input as variables into the preset beam spot position calculation formula, and the coordinates of the position to be adjusted for each laser beam spot at the current sampling point are output, thus obtaining the adjusted beam spot position.

[0162] Finally, the system encodes these adjusted precise position coordinates into position signals or motion control commands in real time and transmits them to the laser beam spot control system, including the scanning galvanometer system, optical path drive platform, or direct drive motor module. This control system rapidly adjusts the projection position of the laser beam in space according to the received commands, achieving microsecond-level laser path correction and trajectory control. This process is continuously executed in each update cycle until the entire laser printing process is completely finished, thus ensuring that the beam spot position is highly consistent, the error is minimized, and the processing quality is optimal throughout the entire printing process.

[0163] In this scheme, by periodically collecting multi-source data such as velocity and temperature and dynamically correcting the laser beam position, the accuracy error caused by velocity drift, heat accumulation and path deviation can be effectively suppressed, and high-precision trajectory tracking can be maintained.

[0164] Example 3

[0165] Figure 3 This is a schematic diagram of the multi-laser 3D printing beam spot verification system provided in Embodiment 3 of this application. Figure 3 As shown, it specifically includes:

[0166] The physical field fusion module 301 is used to acquire multi-physics field signal data from multiple laser sources, and to synchronously acquire and fuse the multi-physics field signal data to obtain comprehensive physical field data; wherein, the multi-physics field signal data includes thermal response signal, stress signal and pressure signal;

[0167] The offset recognition module 302 is used to acquire thermal response data of the beam spot regions of multiple laser sources, and to perform fusion analysis on the thermal response data and comprehensive physical field data to obtain multi-laser beam spot position offset feature data.

[0168] The frequency domain extraction module 303 is used to perform Fourier transform processing on the thermal response signal data of multiple laser sources to obtain the frequency domain feature data of multiple laser sources, and to perform Fourier transform processing on the multiple laser beam spot position offset feature data and the frequency domain feature data to obtain the beam spot frequency feature data of multiple laser sources.

[0169] The verification standard generation module 304 is used to obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data and frequency domain characteristic data to obtain verification standard data.

[0170] The environmental adaptation module 305 is used to acquire external temperature data, external humidity data, and external interference data, determine environmental adaptability data based on the external temperature data, external humidity data, and external interference data, and combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data.

[0171] The accuracy calibration module 306 is used to acquire real-time laser beam spot accuracy data from multiple laser sources, and combine the real-time laser beam spot accuracy data with optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

[0172] The multi-laser 3D printing beam spot verification system provided in this application embodiment can achieve... Figure 1 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0173] Example 4

[0174] like Figure 4 As shown, this application embodiment also provides an electronic device 400, including a processor 401, a memory 402, and a program or instructions stored in the memory 402 and executable on the processor 401. When the program or instructions are executed by the processor 401, they implement the various processes of the above-described multi-laser 3D printing beam spot verification method embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0175] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0176] Example 5

[0177] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described cable installation process based on the tension-adaptive control system embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0178] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0179] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element. Furthermore, it should be noted that the scope of the methods and systems in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0180] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0181] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0182] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A method for verifying beam spots in multi-laser 3D printing, characterized in that, The method includes: Multiphysics field signal data from multiple laser sources are acquired, and the multiphysics field signal data is synchronously acquired and fused to obtain comprehensive physical field data; wherein, the multiphysics field signal data includes thermal response signal, stress signal and pressure signal; Thermal response data of beam spot regions from multiple laser sources are acquired, and the thermal response data is fused and analyzed with comprehensive physical field data to obtain multi-laser beam spot position offset characteristic data. Fourier transform is performed on the thermal response signal data of multiple laser sources to obtain the frequency domain feature data of multiple laser sources. Fourier transform is performed on the beam spot position offset feature data and the frequency domain feature data of the multiple laser sources to obtain the beam spot frequency feature data of multiple laser sources. Obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data, and frequency domain characteristic data to obtain the verification standard data; Acquire external temperature data, external humidity data, and external interference data; determine environmental adaptability data based on the external temperature data, external humidity data, and external interference data; combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data. Real-time laser beam spot accuracy data from multiple laser sources is acquired, and the real-time laser beam spot accuracy data is combined with optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

2. The method according to claim 1, characterized in that, in, To obtain the adaptive thermal response threshold, the following steps are taken: Acquire power data, scanning speed data, material thermal conductivity data, material specific heat capacity data, material density data, and external temperature data from multiple laser sources. Based on these data and a preset adaptive thermal response threshold calculation formula, calculate the adaptive thermal response threshold. The preset adaptive thermal response threshold calculation formula is as follows: in, Adaptive thermal response threshold; This is the preset thermal response sensitivity coefficient; Power data; This refers to the thermal conductivity data of the material. This refers to the specific heat capacity data of the material. For material density data; For scanning speed data; External temperature data; This is the preset ambient temperature suppression and adjustment factor.

3. The method according to claim 1, characterized in that, in, After obtaining the calibrated multi-laser beam accuracy data, the method further includes: Real-time accuracy data of each laser beam spot is acquired, and the real-time accuracy data and the optimized signal processing data are fused and corrected using the Kalman filter algorithm to obtain the corrected multi-laser beam spot accuracy data. The corrected multi-laser beam spot accuracy data is input into the preset multi-laser beam spot interaction effect model to update the multi-laser beam spot accuracy data.

4. The method according to claim 1, characterized in that, in, The multi-laser beam spot position offset feature data and frequency domain feature data are subjected to Fourier transform processing to obtain beam spot frequency feature data of multiple laser sources, including: The frequency domain feature data and the multi-laser beam spot position offset feature data are processed using adaptive frequency domain filtering technology to obtain optimized frequency domain feature data; Kalman filtering and wavelet transform were used to remove noise from the optimized frequency domain feature data, resulting in beam spot frequency feature data for multiple laser sources.

5. The method according to claim 1, characterized in that, in, After obtaining the calibrated multi-laser beam accuracy data, the method further includes: Acquire real-time velocity data for each laser beam spot, and calculate the velocity error of each laser beam spot based on the real-time velocity data; Acquire real-time temperature change data, input the real-time temperature change data into a preset thermal response model, and obtain the thermal response coefficient; The system acquires the current time data and the initial laser position of each laser beam spot. It inputs the real-time velocity data, velocity error, thermal response coefficient, initial laser position, and current time data into a preset beam spot position calculation formula to calculate the adjusted beam spot position of each laser beam spot. The adjusted beam spot position is then transmitted to the laser beam spot control system.

6. The method according to claim 5, characterized in that, in, The preset formula for calculating the beam spot position is: in, To adjust the beam spot position; This is the initial laser position; This is real-time speed data; For speed error; Data is for the current time. This is the preset acceleration adjustment coefficient; This is the preset speed adjustment coefficient.

7. The method according to claim 5, characterized in that, in, After transmitting the adjusted beam position to the laser beam control system, the method further includes: After each preset update time interval, the real-time velocity data of each laser beam spot is reacquired, and the velocity error of each laser beam spot is calculated based on the real-time velocity data. Reacquire real-time temperature change data and input the real-time temperature change data into a preset thermal response model to obtain the thermal response coefficient; The system reacquires the current time data and the initial laser position of each laser beam spot. It inputs the real-time speed data, speed error, thermal response coefficient, initial laser position, and current time data into the preset beam spot position calculation formula to calculate the adjusted beam spot position of each laser beam spot. The adjusted beam spot position is then transmitted to the laser beam spot control system until the laser printing process is completed.

8. A multi-laser 3D printing beam spot verification system, characterized in that, The system includes: The physical field fusion module is used to acquire multi-physics field signal data from multiple laser sources, and to synchronously acquire and fuse the multi-physics field signal data to obtain comprehensive physical field data; wherein, the multi-physics field signal data includes thermal response signals, stress signals and pressure signals; The offset recognition module is used to acquire thermal response data of the beam spot regions of multiple laser sources, and to perform fusion analysis on the thermal response data and comprehensive physical field data to obtain multi-laser beam spot position offset feature data. The frequency domain extraction module is used to perform Fourier transform processing on the thermal response signal data of multiple laser sources to obtain the frequency domain feature data of multiple laser sources, and to perform Fourier transform processing on the multiple laser beam spot position offset feature data and the frequency domain feature data to obtain the beam spot frequency feature data of multiple laser sources. The verification standard generation module is used to obtain the adaptive thermal response threshold, and combine the adaptive thermal response threshold, beam spot frequency characteristic data and frequency domain characteristic data to obtain the verification standard data. The environmental adaptation module is used to acquire external temperature data, external humidity data, and external interference data, determine environmental adaptability data based on the external temperature data, external humidity data, and external interference data, and combine the environmental adaptability data with the calibration standard data to obtain optimized signal processing data. The accuracy calibration module is used to acquire real-time laser beam spot accuracy data from multiple laser sources, and combine the real-time laser beam spot accuracy data with optimized signal processing data to obtain calibrated multi-laser beam spot accuracy data.

9. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the multi-laser 3D printing beam spot verification method as described in any one of claims 1-7.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the multi-laser 3D printing beam spot verification method as described in any one of claims 1-7.

Images