Method, apparatus and storage medium for three-dimensional dither compensation of laser cutting

Abstract

The application discloses a three-dimensional jitter compensation method and device for laser cutting and a storage medium, relates to the technical field of lasers, and comprises the following steps: performing fast Fourier transform on a vibration signal of a to-be-processed sheet in a Z-axis direction to obtain a frequency spectrum feature; performing frequency domain filtering processing on the frequency spectrum feature in combination with a preset jitter filtering parameter set to generate a filtered signal; reconstructing the filtered signal into a time domain compensation amount through inverse Fourier transform; performing weighted fusion on the time domain compensation amount in combination with a cutting acceleration and a cutting displacement deviation of a current cutting direction to generate a three-dimensional compensation vector; and superimposing the three-dimensional compensation vector on original cutting path coordinates through a motion controller to correct a following state of a cutting head. The application overcomes the technical problem that a nonlinear cutting force caused by the interaction between the high-speed motion of the cutting head and the laser and the material leads to the jitter of the sheet, and improves the stability of the sheet during laser cutting.

Description

Technical Field

[0001] This application relates to the field of laser technology, and in particular to a three-dimensional jitter compensation method, device and storage medium for laser cutting. Background Technology

[0002] Laser cutting technology, with its high precision and non-contact processing characteristics, has become a core process in precision manufacturing industries such as semiconductor packaging, flexible circuit board processing, and aerospace titanium alloy thin-walled parts forming.

[0003] When cutting ultra-thin materials at high speed, vibration is inevitable, especially in the processing of ultra-thin titanium alloys for aerospace and copper substrates for microelectronic devices. The nonlinear cutting force caused by the interaction between the high-speed movement of the cutting head and the laser and the material leads to vibration of the thin plate.

[0004] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this application is to provide a three-dimensional jitter compensation method, device, and storage medium for laser cutting, aiming to solve the technical problem of jitter in thin plates caused by nonlinear cutting forces resulting from the high-speed movement of the cutting head and the interaction between the laser and the material.

[0006] To achieve the above objectives, this application proposes a three-dimensional jitter compensation method for laser cutting, the method comprising:

[0007] The spectral characteristics are obtained by performing a fast Fourier transform on the vibration signal of the thin plate to be processed in the Z-axis direction.

[0008] Based on the spectral characteristics, frequency domain filtering is performed using a preset set of jitter filtering parameters to generate a filtered signal.

[0009] The filtered signal is reconstructed into a time-domain compensation quantity through inverse Fourier transform;

[0010] Based on the time-domain compensation amount, a three-dimensional compensation vector is generated by weighted fusion of the cutting acceleration and cutting displacement deviation in the current cutting direction.

[0011] The three-dimensional compensation vector is superimposed onto the original cutting path coordinates via a motion controller to correct the following state of the cutting head.

[0012] In one embodiment, the original signal is obtained by acquiring the signal of the sheet to be processed in the Z-axis direction at a certain moment and filtering the signal;

[0013] Based on the original signal, adaptive filtering is performed to obtain the vibration signal.

[0014] In one embodiment, a target frequency band related to the resonance of the thin plate to be processed is extracted based on the spectral characteristics;

[0015] Based on a preset set of jitter filtering parameters, the vibration components in the target frequency band are subjected to frequency domain filtering to generate the filtered signal.

[0016] In one embodiment, an accelerated signal is obtained by adaptive filtering based on the accelerated signal;

[0017] Based on the filtered acceleration signal, digital sampling is performed to obtain the cutting acceleration corresponding to the timestamp of the acceleration signal; and...

[0018] Based on the displacement data signal, displacement data information is obtained through calculation and processing;

[0019] The displacement data is compared with the preset displacement path to obtain the cutting displacement deviation.

[0020] In one embodiment, the time-domain compensation amount, the cutting acceleration, and the cutting displacement deviation are weighted and fused according to a preset weight to generate the deviation compensation amount;

[0021] The deviation compensation amount is broken down into compensation amounts for each axis according to a preset rule;

[0022] The axial compensation amounts are superimposed as spatial vectors using a three-dimensional coordinate transformation matrix to generate a three-dimensional compensation vector containing amplitude and direction information.

[0023] In one embodiment, the three-dimensional compensation vector is decomposed into translation compensation components and rotation compensation components through a coordinate transformation matrix;

[0024] The motion controller superimposes the translation compensation component and the rotation compensation component onto the original cutting path coordinates to obtain the corrected target coordinates.

[0025] The motion controller corrects the following state based on the target coordinates.

[0026] In one embodiment, when the cutting head collides with the plate during movement, a collision alarm is triggered and the cutting head is automatically raised to a safe height.

[0027] If the collision alarm does not stop, the cutting head will continue to be lifted.

[0028] If the collision alarm stops, the cutting head will no longer be raised, and the three-dimensional compensation vector will be generated based on the current following state to update the following state.

[0029] In one embodiment, the vibration signal, the cutting acceleration, and the cutting displacement deviation are used to generate a dataset according to their corresponding timestamps;

[0030] The dataset is divided using cross-validation, with half of the dataset used as the validation set and the remaining dataset used as the training set.

[0031] The validation set is used to generate a three-dimensional compensation validation vector through the three-dimensional jitter compensation method of laser cutting, and the training set is input into the basic frequency domain filtering model to output a three-dimensional compensation training vector;

[0032] The three-dimensional compensation verification vector is input into the basic frequency domain filtering model and compared with the three-dimensional compensation training vector, and feedback compensation parameters are generated based on the comparison results;

[0033] Based on the feedback compensation parameters, the model parameters of the basic frequency domain filtering model are adjusted, and the basic frequency domain filtering model is repeatedly trained to obtain the target frequency domain filtering model.

[0034] In addition, to achieve the above objectives, this application also proposes a three-dimensional jitter compensation device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the three-dimensional jitter compensation method for laser cutting as described above.

[0035] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the three-dimensional jitter compensation method for laser cutting as described above.

[0036] This application provides a three-dimensional jitter compensation method for laser cutting, including performing a fast Fourier transform on the vibration signal of the thin plate to be processed in the Z-axis direction to obtain spectral features; performing frequency domain filtering processing on the spectral features in combination with a preset jitter filtering parameter set to generate a filtered signal; reconstructing the filtered signal into a time-domain compensation amount through an inverse Fourier transform; weighting and fusing the time-domain compensation amount with the cutting acceleration and cutting displacement deviation in the current cutting direction to generate a three-dimensional compensation vector; and superimposing the three-dimensional compensation vector onto the original cutting path coordinates through a motion controller to correct the following state of the cutting head. By processing the vibration signal in the Z-axis direction and combining it with the motion data of the cutting head to generate a three-dimensional compensation vector, the cutting head can change its following state in a timely manner to compensate for the jitter of the thin plate, thereby improving the stability of the thin plate during laser cutting.

[0037] In summary, this application, by acquiring vibration signals and integrating the motion data of the cutting head, enables the generated three-dimensional compensation vector to more effectively compensate for the vibration of the thin plate. This overcomes the technical problem of the nonlinear cutting force caused by the high-speed movement of the cutting head and the interaction between the laser and the material, which leads to the vibration of the thin plate, and improves the stability of the thin plate during laser cutting. Attached Figure Description

[0038] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating the first embodiment of the three-dimensional jitter compensation method for laser cutting according to this application;

[0041] Figure 2 This is a flowchart illustrating the fourth embodiment of the three-dimensional jitter compensation method for laser cutting according to this application;

[0042] Figure 3 This is a flowchart illustrating the fifth embodiment of the three-dimensional jitter compensation method for laser cutting in this application;

[0043] Figure 4 This is a flowchart illustrating the sixth embodiment of the three-dimensional jitter compensation method for laser cutting in this application;

[0044] Figure 5 This is a flowchart illustrating the seventh embodiment of the three-dimensional jitter compensation method for laser cutting in this application.

[0045] Figure 6 This is a flowchart illustrating the eighth embodiment of the three-dimensional jitter compensation method for laser cutting in this application;

[0046] Figure 7 This is a schematic diagram of the structure of the three-dimensional jitter compensation device of this application.

[0047] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0048] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0049] When cutting ultra-thin materials at high speed, vibration is inevitable, especially in the processing of ultra-thin titanium alloys for aerospace and copper substrates for microelectronic devices. The nonlinear cutting force caused by the interaction between the high-speed movement of the cutting head and the laser and the material leads to vibration of the thin plate.

[0050] This application provides a solution: First, a fast Fourier transform is performed on the vibration signal of the sheet metal to be processed in the Z-axis direction to obtain the spectral characteristics; based on the spectral characteristics, frequency domain filtering is performed in combination with a preset jitter filtering parameter set to generate a filtered signal; then, the filtered signal is reconstructed into a time-domain compensation quantity through an inverse Fourier transform; then, based on the time-domain compensation quantity, the cutting acceleration and cutting displacement deviation in the current cutting direction are weighted and fused to generate a three-dimensional compensation vector; finally, the three-dimensional compensation vector is superimposed onto the original cutting path coordinates through a motion controller to correct the following state of the cutting head.

[0051] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or a three-dimensional jitter compensation device capable of performing the above functions. The following description uses a three-dimensional jitter compensation device as an example to illustrate this embodiment and the subsequent embodiments.

[0052] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0053] This application provides a three-dimensional jitter compensation method for laser cutting, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the three-dimensional jitter compensation method for laser cutting according to this application.

[0054] In this embodiment, the three-dimensional jitter compensation method for laser cutting includes steps S10 to S50:

[0055] Step S10: Perform a fast Fourier transform on the vibration signal of the thin plate to be processed in the Z-axis direction to obtain the spectral characteristics.

[0056] In this embodiment, the thin plate to be processed refers to the metal or non-metal thin plate material to be processed in the laser cutting process. The vibration signal in the Z-axis direction is time-domain acceleration data collected by the motor encoder of the cutting head in a direction perpendicular to the plane of the thin plate. Fast Fourier Transform (FFT) is an efficient algorithm for converting time-domain signals into frequency-domain signals, used to analyze vibration frequency components. Spectral characteristics refer to the key frequency components and their parameterized descriptions exhibited by the signal after Fourier transform decomposition in the frequency domain.

[0057] As an optional implementation, the vibration signal of the thin plate to be processed in the Z-axis direction during processing is collected, and the vibration signal is subjected to a fast Fourier transform to convert the time domain signal in the vibration signal into a frequency domain spectrum. Based on the frequency domain spectrum and combined with a preset jitter frequency identification rule, the spectral features related to jitter are identified. The jitter frequency identification rule includes limiting a preset frequency range.

[0058] By using Fast Fourier Transform to convert the time-domain signal into a frequency-domain spectrum, the characteristic frequencies related to jitter are identified. A preset jitter filter parameter set is called, and the filter transfer function is multiplied point by point with the Fast Fourier spectrum in the frequency domain to filter out the energy of the target frequency band and generate a filtered signal.

[0059] Step S20: Based on the spectral characteristics, frequency domain filtering is performed using a preset jitter filter parameter set to generate a filtered signal.

[0060] In this embodiment, the preset jitter filtering parameter set is a frequency range set based on historical device data or theoretical models. Frequency domain filtering refers to suppressing signal energy in a specific frequency band through multiplication or convolution operations in the frequency domain. The filtered signal is vibration data after frequency domain processing, retaining the effective components and removing noise or interference frequencies.

[0061] As an optional implementation, based on the spectral characteristics, a corresponding preset jitter filter parameter set is invoked, a frequency domain filter transfer function is generated based on the jitter filter parameter set, the frequency domain filter transfer function is multiplied point by point with the spectral characteristics and frequency domain filtering is performed, and the filtered signal within the preset frequency domain range is retained.

[0062] Step S30: The filtered signal is reconstructed into a time-domain compensation quantity through inverse Fourier transform.

[0063] In this embodiment, the inverse Fourier transform is the inverse operation of the fast Fourier transform, restoring the frequency domain signal to a time domain waveform. Reconstruction refers to restoring the time-series shape of the signal through mathematical operations. The time-domain compensation amount is an inverse control signal generated based on the filtered signal, used to counteract the original vibration.

[0064] As an optional implementation, the filtered signal is transformed and reconstructed into a time-domain waveform using inverse Fourier transform. The real part data aligned with the time axis of the original vibration signal is extracted as the compensation reference. The compensation amount is then calibrated for amplitude and corrected for phase shift based on dynamic response characteristics. The compensation amount is then adjusted for gain in the time domain using the transfer function matrix. Based on the compensation reference, amplitude calibration, phase shift correction, and gain adjustment, a time-domain compensation amount with the same amplitude but opposite phase as the original vibration is generated.

[0065] Step S40: Based on the time-domain compensation amount, a weighted fusion is performed by combining the cutting acceleration and cutting displacement deviation in the current cutting direction to generate a three-dimensional compensation vector.

[0066] In this embodiment, the cutting direction refers to the travel angle of the laser head's current trajectory in the X / Y plane. The cutting acceleration is the instantaneous acceleration value of the cutting head along the tangent direction of the path. The cutting displacement deviation is the real-time position error between the actual position and the theoretical path. Weighted fusion refers to the linear or nonlinear combination of different physical quantities according to preset weighting coefficients. The three-dimensional compensation vector is a spatial vector that fuses the X / Y / Z axis compensation quantities, used for multi-axis cooperative motion correction.

[0067] As an optional implementation, based on preset weights, the first weight is multiplied by the time-domain compensation amount to obtain the first vector, the second weight is multiplied by the cutting acceleration to obtain the second vector, the third weight is multiplied by the cutting displacement deviation to obtain the third vector, the second vector and the third vector are added together and converted into X / Y axis Cartesian coordinate system components through a homogeneous coordinate transformation matrix, and the X / Y axis Cartesian coordinate system components are superimposed with the first vector on the Z axis to generate a three-dimensional compensation vector.

[0068] Step S50: The three-dimensional compensation vector is superimposed onto the original cutting path coordinates through the motion controller to correct the following state of the cutting head.

[0069] In this embodiment, the motion controller is a hardware module used to drive the servo motor or actuator, responsible for parsing path instructions and outputting control signals. The original cutting path coordinates are the theoretical coordinate points of the cutting trajectory preset in the CNC system. Superposition refers to algebraically adding the compensation amount to the theoretical coordinates. The cutting head is the execution end that performs laser focusing and material processing. The following state is the actual relative position of the cutting head and the thin plate. When the thin plate vibrates, the cutting head follows in the same direction as the vibration of the thin plate through compensation.

[0070] As an optional implementation, the three-dimensional compensation vector is converted into a compensation command consistent with the coordinate system of the motion controller according to the control cycle. The compensation command is transmitted to the motion controller, and the motion controller superimposes the three-dimensional compensation vector corresponding to the compensation command onto the original cutting path coordinates to generate updated cutting path coordinates. The following state of the cutting head is then corrected to run according to the updated cutting path coordinates.

[0071] For example, a stainless steel sheet is used as the processing object, and a fiber laser cutting machine is used to cut it linearly along the X-axis. A piezoelectric accelerometer installed at the center of the sheet collects the Z-axis vibration signal in real time, and a 10-second time-domain waveform during the stable cutting phase is captured. A fast Fourier transform is performed to generate a spectrum, identifying the 450Hz main peak and the 120Hz subharmonic. A preset band-stop filter parameter set (center frequency 450Hz±20Hz, hysteresis attenuation -40dB) is used to suppress the target frequency band energy in the frequency domain. The time-domain compensation is then obtained by inverse Fourier transform reconstruction. The compensation signal is synchronously read from the servo system's X-axis tangential acceleration (peak value 1.2 m / s²) and Y-axis normal displacement deviation (±8 μm). After Kalman filtering and fusion, a three-dimensional compensation vector (X: -0.4 μm, Y: +5.6 μm, Z: -14.8 μm) is generated according to weighting coefficients (acceleration 0.3, deviation 0.7). The compensation amount is superimposed on the theoretical path coordinates of the motion controller with a 1 ms control cycle. A look-ahead interpolation algorithm is used to smooth abrupt trajectory changes. The closed-loop verification by the grating ruler shows that after compensation, the Z-axis vibration RMS (root mean square value) value is reduced from 0.21g to 0.06g, and the X / Y tracking error is ≤3 μm. Finally, a log file is generated to record the timestamp, compensation amount, and actual path data, achieving high-precision dynamic correction.

[0072] By combining the acquisition of vibration signals with the motion data of the cutting head, the generated three-dimensional compensation vector can more effectively compensate for the vibration of the thin plate, overcome the technical problem of the nonlinear cutting force caused by the high-speed movement of the cutting head and the interaction between the laser and the material, which causes the thin plate to vibrate, and improve the stability of the thin plate during laser cutting.

[0073] Based on any of the above embodiments, in Embodiment 2 of this application, before step S10, steps A11 to A12 are further included:

[0074] Step A11: The original signal is obtained by acquiring the signal of the thin plate to be processed in the Z-axis direction at a certain moment and filtering the signal.

[0075] In this embodiment, "a certain time" refers to a signal segment captured at a specific point in time. Filtering involves using a preprocessing algorithm to remove vibration signals along the Z-axis that do not belong to the sheet metal to be processed. The original signal is the original signal obtained after filtering out signals that do not belong to the sheet metal to be processed and are generated by vibration along the Z-axis.

[0076] As an optional implementation, vibration signals are acquired at a specific moment or in real time, and interference from signals that do not belong to the vibration of the thin plate on the Z-axis is filtered out by a low-pass filter to obtain the original signal.

[0077] Step A12: Based on the original signal, perform adaptive filtering to obtain the vibration signal.

[0078] In this embodiment, environmental noise is an unwanted signal generated by external interference. Noise removal is achieved by using signal processing algorithms to eliminate noise components. The vibration signal is the effective data reflecting the true vibration of the thin plate after noise removal.

[0079] As an alternative implementation, the original signal is subjected to adaptive filtering using an adaptive threshold to suppress environmental noise and obtain a basic signal. Then, a frequency domain bandpass filter is used to filter out the non-thin plate vibration frequency band to obtain a vibration signal.

[0080] For example, using an aluminum alloy sheet as the test object, a laser cutting machine was used to cut a square hole. A piezoelectric accelerometer was rigidly fixed to the geometric center of the sheet along the Z-axis using a magnetic base. At the moment the cutting started, the acquisition system was triggered to synchronously record 5 seconds of raw signal. A low-pass filter (cutoff frequency 5kHz, fourth order) was used to filter out high-frequency electromagnetic noise, generating a raw signal file (including timestamp and acceleration g value). Based on this signal, a 5-level decomposition was performed using a wavelet basis. Environmental noise was removed by improving the threshold function. After reconstruction, a 50Hz-2kHz bandpass filter was superimposed. Finally, the signal-to-noise ratio of the vibration signal was improved from 12dB to 28dB, and the time-domain waveform correlation coefficient reached 0.97.

[0081] By acquiring vibration signals in the Z-axis direction and removing noise, the generated compensation amount is more accurate, thus improving the reliability of thin plate vibration compensation during laser cutting.

[0082] Based on any of the above embodiments, in Embodiment 3 of this application, step S10 includes steps B11 to B12:

[0083] Step B11: Based on the spectral characteristics, extract the target frequency band related to the resonance of the thin plate to be processed.

[0084] In this embodiment, resonance is the phenomenon where the amplitude of forced vibration of an object increases significantly at a specific frequency. The target frequency band is a specific frequency range in the spectrum that overlaps with the material's natural frequency and the excitation frequency of the processing equipment.

[0085] As an optional implementation, frequency bands are located by using the inherent frequency to identify the peak frequency band of the target frequency band. The coupling frequency band of the target frequency band is identified by combining process parameters. Then, the amplitude and phase data of the target frequency band are extracted by using a bandpass filter to obtain the target frequency band including the peak frequency band, coupling frequency band, amplitude and phase data.

[0086] As one alternative method for obtaining the natural frequency, the natural frequency of the thin plate is obtained by striking the corresponding thin plate using an experimental striking method.

[0087] Step B12: Based on a preset set of jitter filtering parameters, frequency domain filtering is performed on the vibration components in the target frequency band to generate the filtered signal.

[0088] In this embodiment, the preset jitter filter parameter set is a filter parameter predefined based on the device vibration characteristics. The vibration component is the energy distribution corresponding to a specific frequency in the signal.

[0089] As an optional implementation, a frequency domain filter transfer function is generated based on the jitter filter parameter set. The frequency domain filter transfer function is then multiplied point by point with the vibration components in the target frequency band to perform frequency domain filtering, thus retaining the filtered signal within the preset frequency range.

[0090] The preset band-stop filter parameters are invoked to set the spectral amplitude of the target frequency band to zero or scale it according to the attenuation coefficient. The vibration amplitude attenuation rate after filtering is verified by comparing the time-domain waveform, and the output is a time-aligned filtered signal.

[0091] For example, a laser cutting experiment was conducted using a thin aluminum alloy sheet. Vibration signals during the cutting process were synchronously acquired using a triaxial piezoelectric accelerometer. A fast Fourier transform was performed on the Z-axis signal to separate the main peak of the spectrum at 520Hz (amplitude 0.6g), the secondary peak at 120Hz (amplitude 0.15g), and high-frequency noise (>5kHz). Combined with the results of the finite element modal analysis of the thin sheet, the target resonance frequency band of 510-530Hz was extracted. Preset hysteresis filter parameters (center frequency 520Hz, bandwidth ±15Hz, hysteresis attenuation -50dB) were applied to attenuate the amplitude of the target frequency band to below 5% in the frequency domain. The filtered signal was reconstructed by inverse Fourier transform. The time-domain waveform showed that the amplitude of the 520Hz component was reduced to 0.08g, the signal-to-noise ratio was increased from 18dB to 32dB, and the waveform correlation coefficient reached 0.98. The output was a data file containing timestamps, original signal, filtered signal, and spectrum comparison diagram, which was used for subsequent active vibration compensation control.

[0092] By combining the jitter filter parameter set with the vibration component, the corresponding compensation parameters can be generated according to the jitter parameter settings, thus improving the reliability of jitter compensation for thin plates during laser cutting.

[0093] Based on any of the above embodiments, in Embodiment 4 of this application, referring to Figure 2 , Figure 2 This is a flowchart illustrating the fourth embodiment of the three-dimensional jitter compensation method for laser cutting according to this application. Before step S30, steps C11 to C14 are also included:

[0094] Step C11: Based on the acceleration signal, an adaptive filter is used to obtain the filtered acceleration signal.

[0095] In this embodiment, the acceleration signal is an electrical signal output by the sensor that reflects changes in the object's acceleration. Environmental noise is a non-target acceleration component caused by external interference. Adaptive filtering removes noise and retains the valid signal through filtering or algorithms.

[0096] As an alternative implementation, the acceleration signal is obtained by adaptively filtering the acceleration signal using an adaptive threshold to suppress environmental noise, and then filtering out the non-cut head acceleration frequency band by combining frequency domain bandpass filtering to obtain the acceleration signal.

[0097] Step C12: Based on the filtered acceleration signal, perform digital sampling to obtain the cutting acceleration corresponding to the timestamp of the acceleration signal.

[0098] In this embodiment, digital sampling is a quantization process that converts a continuous analog signal into a discrete digital signal. A timestamp is a precise time marker for a data point. Cutting acceleration is the dynamic acceleration value of the cutting head or material during the processing.

[0099] As an optional implementation, the filtered acceleration signal is digitally sampled and converted into a discrete acceleration signal. The cutting acceleration corresponding to the timestamp is then obtained from the discrete acceleration signal.

[0100] Step C13: Based on the displacement data signal, displacement data information is obtained through calculation and processing.

[0101] In this embodiment, the displacement data signal is the real-time position information output by the position sensor. The processing converts the original signal into a physical displacement quantity through coordinate transformation or data filtering. The displacement data information refers to the actual displacement data quantity and actual coordinate position of the cutting head.

[0102] As an optional implementation, based on the collected X / Y axis displacement data signals, the displacement data signals are converted into actual coordinate positions according to the timestamps through coordinate transformation of the calculation process, so as to obtain displacement data information including the corresponding timestamps and actual coordinate positions.

[0103] Step C14: Compare the displacement data with the preset displacement path to obtain the cutting displacement deviation.

[0104] In this embodiment, the preset displacement path is the theoretical coordinate sequence of the cutting trajectory planned in the CNC system. The deviation is determined by calculating the difference between the actual position and the theoretical position. The cutting displacement deviation is the error amount by which the actual movement trajectory of the cutting head deviates from the theoretical path.

[0105] As an optional implementation, the Euclidean distance is calculated between the actual displacement coordinates corresponding to the timestamp in the actual displacement data and the coordinates in the preset displacement path at the same timestamp to obtain the cutting displacement deviation.

[0106] It should be noted that steps C11 and C12 are performed simultaneously with steps C13 and C14.

[0107] For example, taking a thin carbon steel sheet as the object, a fiber laser cutting machine is used for arc cutting. The acceleration signal of the cutting head is collected by an accelerometer at a sampling rate of 100kHz. The signal is decomposed into 6 layers using a wavelet basis and environmental noise, such as the 80Hz vibration of the machine tool motor, is removed by a preset threshold. Cutting acceleration data with GPS synchronization timestamp is generated. Simultaneously, the X / Y axis displacement signals are acquired in real time by a high-precision grating ruler. After eliminating mechanical backlash noise by Kalman filtering, the data is compared in real time with the preset arc path (radius 50mm, center coordinates X=300mm / Y=300mm) to calculate the displacement deviation (maximum normal deviation 0.12mm, tangential deviation 0.05mm). The acceleration and deviation data are then fused and output as a structured file.

[0108] Because the cutting acceleration and cutting displacement deviation of the cutting head are combined in the compensation calculation, three-dimensional compensation data can be generated, which improves the reliability of thin plate vibration compensation during laser cutting.

[0109] Based on any of the above embodiments, in Embodiment 5 of this application, referring to Figure 3 , Figure 3 This is a flowchart illustrating the fifth embodiment of the three-dimensional jitter compensation method for laser cutting according to this application. Step S30 includes steps D11 to D13:

[0110] Step D11: Based on the time-domain compensation amount, the cutting acceleration, and the cutting displacement deviation, a weighted fusion is performed according to a preset weight to generate a deviation compensation amount.

[0111] In this embodiment, the preset weights are the contribution ratio coefficients of the time-domain compensation amount, cutting acceleration, and cutting displacement deviation, set according to process requirements. The deviation compensation amount is a multi-correction amount after integrating acceleration and deviation. The deviation compensation amount is a correction amount generated based on position error.

[0112] As an optional implementation, according to preset weights, the first weight is multiplied by the time-domain compensation amount to obtain the first vector, the second weight is multiplied by the cutting acceleration to obtain the second vector, the third weight is multiplied by the cutting displacement deviation to obtain the third vector, the second vector and the third vector are added together and converted into X / Y axis Cartesian coordinate system components through a homogeneous coordinate transformation matrix, and the X / Y axis Cartesian coordinate system components are superimposed with the first vector of the Z axis to generate the deviation compensation amount.

[0113] Step D12: Decompose the deviation compensation amount into compensation amounts for each axis according to a preset rule.

[0114] In this embodiment, the preset rule refers to the decomposition formula that is adaptively set by using trigonometric functions based on the distribution relationship of the X, Y and Z axes, to decompose the deviation compensation amount into the axial compensation amount of the three axes.

[0115] As an optional implementation method, based on preset rules, the deviation compensation amount is decomposed according to the decomposition formula in the preset rules, and the deviation compensation amount is decomposed into axial compensation amounts on the X-axis, Y-axis and Z-axis through the formula.

[0116] Step D13: The axial compensation amounts are superimposed as spatial vectors using a three-dimensional coordinate transformation matrix to generate a three-dimensional compensation vector containing amplitude and direction information.

[0117] In this embodiment, the three-dimensional coordinate transformation matrix is ​​a mathematical tool that maps the compensation amounts along each axis to a unified coordinate system. Each axial compensation amount is a correction value in an independent X / Y / Z direction. Spatial vector superposition is an operation that synthesizes multi-axis components into a three-dimensional vector.

[0118] As an optional implementation, a three-dimensional coordinate transformation matrix is ​​constructed according to the weighting coefficients of the compensation amounts of each axis corresponding to the X-axis, Y-axis and Z-axis. The X / Y / Z axial compensation amounts are converted into spatial vectors in the workpiece coordinate system according to the three-dimensional coordinate transformation matrix. The spatial vectors are superimposed with the initial coordinate vectors to generate a three-dimensional compensation vector containing amplitude and direction information.

[0119] For example, using a stainless steel sheet as the test object, an S-curve cut was performed using a fiber laser cutting machine. Vibration signals were acquired by a piezoelectric accelerometer at a sampling rate of 50kHz. After band-stop filtering (center frequency 350Hz, bandwidth ±15Hz, hysteresis attenuation -45dB), a filtered signal was generated. The time-domain compensation was reconstructed by inverse Fourier transform. The X-axis cutting acceleration of the servo motor (peak ±2.5m / s2) and the Y-axis displacement deviation (-0.15mm) fed back by the grating ruler were read synchronously. The values ​​were then calculated according to preset weights (acceleration 0.4, deviation 0.4). 6) Weighted fusion generates deviation compensation amounts (X: +0.5μm, Y: -9.8μm). The time-domain compensation amounts are subjected to trapezoidal integration and Kalman filtering for smoothing. The X / Y / Z compensation amounts (X: +4μm, Y: -9.8μm, Z: -10.3μm) are mapped to three-dimensional compensation vectors (amplitude 15.2μm, direction angle θ = arctan(-9.8 / 4) = -68°, φ = arccos(-10.3 / 15.2) = 133°) through a homogeneous coordinate transformation matrix (including 45° tilt attitude compensation of the cutting head).

[0120] By using data from three axes, including the plate jitter on the Z-axis and the cutting head motion data on the X / Y axes, a three-dimensional compensation vector is obtained through weighted fusion, resulting in a significant compensation effect and improving the effectiveness of plate jitter compensation during laser cutting.

[0121] Based on any of the above embodiments, in Embodiment Six of this application, referring to Figure 4 , Figure 4 This is a flowchart illustrating the sixth embodiment of the three-dimensional jitter compensation method for laser cutting according to this application. Step S40 includes steps E11 to E13:

[0122] Step E11: Based on the three-dimensional compensation vector, decompose it into translation compensation components and rotation compensation components through a coordinate transformation matrix.

[0123] In this embodiment, the coordinate transformation matrix is ​​a conversion tool for mathematically modeling the translation and rotation relationships between different coordinate systems. Decomposition is a mathematical operation that breaks down a vector into independent motion components. The translation compensation component is the pure displacement that corrects positional deviations. The rotation compensation component is the rotation about an axis that corrects angular deviations.

[0124] As an optional implementation, a three-dimensional compensation vector containing three axes, X, Y and Z, is input into a coordinate transformation matrix for decomposition. The translation and rotation components are separated by singular value decomposition. The translation component is superimposed onto the interpolation path coordinates by a motion controller, and the rotation component is used to drive the rotation axis of the cutting head to deflect synchronously. The feedback of the compensated position is verified by a laser tracker, and the translation compensation component and rotation compensation component are generated based on the feedback.

[0125] Step E12: The translation compensation component and the rotation compensation component are superimposed onto the original cutting path coordinates by the motion controller to obtain the corrected target coordinates.

[0126] In this embodiment, the superposition is achieved by adding the compensation amount to the original value through mathematical operations. The original cutting path coordinates are the preset theoretical processing trajectory coordinates. The corrected target coordinates are the actual execution coordinates after the compensation amount is superimposed.

[0127] As an optional implementation, the motion controller reads the translation compensation component and the rotation compensation component, and algebraically adds the translation component to the original path coordinates in unit control cycles. The rotation component is then used to correct the rotation angle of the cutting head through a homogeneous coordinate transformation matrix to generate target coordinates including coordinate position and rotation angle.

[0128] Step E13: The motion controller corrects the following state based on the target coordinates.

[0129] In this embodiment, the target coordinates are the actual execution position coordinates after compensation and correction.

[0130] As an optional implementation, the motion controller receives the target coordinates, compares the target coordinates with the running coordinates, calculates the deviation between the actual position and the target coordinates by combining the closed-loop feedback signal from the encoder or grating ruler, and adjusts the operating parameters of the servo motor according to the deviation to achieve the correction of the following state.

[0131] For example, the three-dimensional compensation vector is decomposed into translational components (X = +5μm, Y = -3μm, Z = -8μm) and rotational components (θx = 0.05°, θy = 0.1°, θz = 0.02°) by a homogeneous coordinate transformation matrix (including Euler angles of the cutting head posture). The motion controller superimposes the translational components onto the original path coordinates at 1ms intervals (e.g., X is corrected from 150.000mm to 150.005mm, Y from 75.000mm to 74.997mm, and Z from 0.05mm to 0.02°). The rotation component drives the rotation axis (C-axis) to deflect by 0.1°. The following state correction is achieved through dynamic adjustment of operating parameters (proportional gain increased by 15%, integral time shortened by 20%) and acceleration feedforward control. After compensation, the position error is ≤2μm and the angle deviation is ≤0.005°. The vibration RMS value is reduced from 0.3g to 0.07g. The corrected coordinates, compensation parameters and grating ruler closed-loop data are output to the feedback file to meet the requirements of high-precision dynamic compensation.

[0132] By decomposing the three-dimensional compensation vector and superimposing it onto the cutting path coordinates, the state of the cutting head can be adjusted at any moment, thus improving the accuracy of thin plate vibration compensation during laser cutting.

[0133] Based on any of the above embodiments, in Embodiment Seven of this application, referring to Figure 5 , Figure 5 This is a flowchart illustrating the seventh embodiment of the three-dimensional jitter compensation method for laser cutting according to this application. Following step S40, steps F11 to F13 are also included:

[0134] In step F11, when the cutting head collides with the plate during its movement, a collision alarm will be triggered and the cutting head will be automatically raised to a safe height.

[0135] In this embodiment, a collision occurs when the cutting head makes unexpected contact with the workpiece or fixture. A collision alarm is an audible, visual, or software warning triggered upon detecting a collision. The safe height is a preset Z-axis lifting distance to prevent secondary collisions. Automatic lifting is achieved by using a servo motor or pneumatic device to drive the cutting head for emergency avoidance.

[0136] As an optional implementation, a force sensor installed at the end of the cutting head monitors the X / Y / Z triaxial forces and torques in real time. When the force in any direction exceeds a threshold and the duration exceeds a preset time threshold, a collision alarm is triggered. An emergency stop command is sent to the motion controller via the bus, and the safe height parameter is called simultaneously. The servo motor drives the cutting head to rise to a safe position with maximum acceleration. After the raising is completed, the motion axis is locked and the alarm is stopped.

[0137] Step F12: If the collision alarm does not stop, continue lifting the cutting head.

[0138] In this embodiment, continuing to lift the cutting head is an obstacle avoidance action that involves vertically raising the cutting head to a safe position along the Z-axis, as the risk of collision still exists.

[0139] As an optional implementation, if the collision alarm does not stop, the lifting position is verified in real time by an absolute encoder. If a collision risk is still found, the lifting of the cutting head continues.

[0140] Step F13: If the collision alarm stops, the cutting head lifting will no longer be performed, and the three-dimensional compensation vector will be generated based on the current following state to update the following state.

[0141] In this embodiment, "collision alarm stop" means that the collision detection system deactivates the alarm and resumes processing.

[0142] As an optional implementation, when the collision alarm is cleared, the motion controller terminates the lifting command and resumes processing. It acquires vibration signals, acceleration signals and displacement data signals of the current following state in real time through sensors. After noise is removed by Kalman filtering, the signals are fused according to preset weights to generate a three-dimensional compensation vector. The vector is then mapped to the workpiece coordinate system through a homogeneous coordinate transformation matrix. The operating parameters of the motion controller are updated with a preset control cycle, and the compensated target coordinates are sent out synchronously to update the following state of the cutting head.

[0143] For example, the cutting head is monitored in real time by a six-dimensional force sensor. When a Z-axis contact force > 60N (threshold) is detected and lasts for > 20ms, a collision alarm is triggered, the motion axis is stopped, and the servo motor moves at 3m / s 2 Acceleration will raise the cutting head to a safe height (Z-axis + 8mm). If the alarm does not stop (e.g., force continues > 50N), then at 1m / s²... 2The increment continues to rise until the Z-axis reaches +15mm; if the alarm is cleared (Z-axis force <10N for 50ms), the rising stops, and based on the current following state deviation (X-axis +0.01mm, Y-axis -0.008mm, Z-axis +0.002mm) fed back by the grating ruler and the accelerometer data, a three-dimensional compensation vector (X: +5μm, Y: -4μm, Z: -1μm) is generated according to weights (displacement 0.5, vibration 0.5). This vector is then decomposed into translation compensation through a homogeneous transformation matrix. For both compensation and rotational compensation, the motion controller updates the operating parameters (proportional gain +12%) at a period of 0.5ms, and superimposes the compensation amount onto the target coordinates (X = 200.005mm, Y = 150.996mm, Z = 0.001mm). The laser tracker verifies that the position error is ≤2μm and the vibration RMS value is ≤0.05g. The alarm log, compensation vector, and control parameters are output to the feedback file (including timestamp, coordinates, force curve, and spectrum) to complete the collision recovery closed-loop control.

[0144] The collision alarm system protects the thin plate from damage, providing better control over operational errors during the cutting process and improving the safety and reliability of laser cutting.

[0145] Based on any of the above embodiments, in Embodiment Eight of this application, referring to Figure 6 , Figure 6 This is a flowchart illustrating the eighth embodiment of the three-dimensional jitter compensation method for laser cutting according to this application. Following step C12, steps G11 to G15 are also included:

[0146] Step G11: Generate a dataset by taking the vibration signal, the cutting acceleration, and the cutting displacement deviation according to their corresponding timestamps.

[0147] In this embodiment, the timestamp is a precise time marker of the moment the data point was collected. The dataset is a multi-dimensional collection of data that is time-aligned and stored in a structured manner.

[0148] As an optional implementation, vibration signals, cutting acceleration, and displacement deviation are collected by a preset sampling rate. The time axis is unified by an interpolation algorithm. The vibration signals are low-pass filtered and detrended. The cutting acceleration is smoothed by moving average. The displacement deviation is denoised by Kalman filtering. The three types of data are merged into a structured dataset according to timestamps.

[0149] Step G12: The dataset is divided using cross-validation, with half of the dataset used as the validation set and the remaining dataset used as the training set.

[0150] In this embodiment, cross-validation is a method of dividing the dataset into multiple subsets to alternately train and validate the model. The validation set is the subset of data used to evaluate the model's generalization performance. The training set is the subset of data used to train the model parameters.

[0151] As an optional implementation, the dataset containing vibration signals, cutting acceleration, and displacement deviations with timestamp synchronization is divided into a training set and a validation set according to a certain ratio to ensure that the timestamps of the two types of data are continuous and have a consistent distribution.

[0152] Step G13: The validation set is used to generate a three-dimensional compensation validation vector through the three-dimensional jitter compensation method of laser cutting, and the training set is input into the basic frequency domain filtering model to output a three-dimensional compensation training vector.

[0153] In this embodiment, the basic frequency domain filtering model is a basic model built on the core algorithms of fast Fourier transform and frequency domain filtering.

[0154] As an optional implementation, the validation set is subjected to Fast Fourier Transform to obtain spectral features. Then, based on the spectral features, frequency domain filtering is performed in combination with a preset jitter filtering parameter set to generate a filtered signal. The filtered signal is reconstructed into a time domain compensation quantity through Inverse Fourier Transform. Then, based on the time domain compensation quantity, the cutting acceleration and cutting displacement deviation in the current cutting direction are weighted and fused to generate a three-dimensional compensation vector and a three-dimensional compensation validation vector. At the same time, the training set is input into the frequency domain filtering model, and after filtering and inverse transformation, a three-dimensional compensation training vector is output.

[0155] Step G14: The three-dimensional compensation verification vector is input into the basic frequency domain filtering model and compared with the three-dimensional compensation training vector, and feedback compensation parameters are generated based on the comparison results.

[0156] In this embodiment, the feedback compensation parameter is an adjustment parameter for optimizing the model or compensation strategy based on the comparison results.

[0157] As an optional implementation, the parameters of the frequency domain filtering model are dynamically adjusted by calculating the axial error and correlation coefficient between the three-dimensional compensation verification vector and the three-dimensional compensation training vector, using the gradient descent method. The model weights are iteratively updated by an optimizer with a preset learning rate, and feedback compensation parameters are generated.

[0158] Step G15: Based on the feedback compensation parameters, adjust the model parameters of the basic frequency domain filtering model and repeatedly train the basic frequency domain filtering model to obtain the target frequency domain filtering model.

[0159] In this embodiment, the model parameters are the core configuration of the filter. Adjustments are made by updating the parameters through gradient descent or heuristic algorithms. Repeated training involves iteratively optimizing the model weights. The target frequency domain filtering model is the optimized frequency filtering model that meets the accuracy requirements.

[0160] As an optional implementation, based on feedback compensation parameters, the filter parameter library of the basic frequency domain filtering model is updated by an optimizer. The root mean square error and correlation coefficient are used as loss functions. The training set is trained several times. The generalization performance is evaluated on the validation set in each round. When the validation loss decreases by a threshold for consecutive rounds, an early stopping mechanism is triggered, and the target frequency domain filtering model is finally obtained.

[0161] As an optional implementation using a target frequency filtering model, the vibration signal at a certain moment, as well as the cutting acceleration and cutting displacement deviation in the cutting direction of the cutting head, are acquired. The vibration signal, cutting acceleration, and cutting displacement deviation are then input into the target frequency domain filtering model to obtain the target coordinates of the cutting head at that moment. The target frequency domain filtering model then transmits these target coordinates to the motion controller, which corrects the following state of the cutting head based on the target coordinates.

[0162] For example, using stainless steel sheet as the processing object, a fiber laser cutting machine is used for grid path cutting. Vibration signals are collected by an accelerometer at a sampling rate of 50kHz, and displacement deviations are recorded by a grating ruler. A servo encoder collects X / Y axis cutting acceleration at 1kHz. A dataset containing 100,000 samples is generated by synchronizing timestamps (accuracy ±1μs) via the PTP protocol. (Fields include timestamp, triaxial vibration g-value, and X / Y acceleration m / s²). 2 The dataset, containing three-axis displacement deviations (μm), is divided into a training set (50,000 records) and a validation set (50,000 records) in a 5:5 ratio. The validation set is input into a 3D jitter compensation module (Kalman filter + weighted fusion, vibration weight 0.5, acceleration 0.3, deviation 0.2) to generate 3D compensation validation vectors (X: ±6μm, Y: ±4μm, Z: ±10μm). The training set is input into a basic frequency domain filtering model (band-stop filter bank: initial center frequency 300Hz / 500Hz, bandwidth ±20Hz, attenuation -30dB) to generate 3D compensation training vectors (X: ±7μm, Y: ±5μm, Z: ±12μm). Axial error analysis (X-axis root mean square error = 1.2μm, Y-axis 1.0μm, Z-axis 2.5μm) and correlation coefficient (R² = 0.82) were used to generate feedback compensation parameters (center frequency +15Hz, bandwidth reduced to ±15Hz, attenuation enhanced to -40dB). The Adam optimizer (learning rate 0.001, 200 iterations) was used to adjust the model parameters. After the early stopping mechanism was triggered, the target frequency domain filtering model (center frequency 315Hz / 515Hz) was obtained. After actual cutting verification, the vibration and jitter decreased by 50% and the displacement error was ≤2μm. The model parameters and verification data were stored in the database for adaptive retrieval.

[0163] By suppressing the influence of thin plate vibration through jitter filtering parameters and modeling the material vibration characteristics as a frequency domain filtering model, real-time vibration compensation is achieved, which improves the efficiency of generating thin plate jitter compensation during laser cutting.

[0164] This application provides a three-dimensional jitter compensation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the laser cutting three-dimensional jitter compensation method in Embodiment 1 above.

[0165] The following is for reference. Figure 7 This document illustrates a structural schematic diagram of a three-dimensional jitter compensation device suitable for implementing embodiments of this application. The three-dimensional jitter compensation device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, three-dimensional jitter compensation devices, personal digital assistants (PDAs), tablet computers (PADs), portable media players (PMPs), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as three-dimensional jitter compensation devices and laser cutting machines. Figure 7 The three-dimensional jitter compensation device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0166] like Figure 7As shown, the three-dimensional jitter compensation device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the three-dimensional jitter compensation device. The processing unit 1001, the read-only memory 1002, and the RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the 3D jitter compensation device to communicate wirelessly or wiredly with other devices to exchange data. Although a 3D jitter compensation device with various systems is shown in the figure, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0167] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0168] The three-dimensional jitter compensation device provided in this application employs the three-dimensional jitter compensation method for laser cutting described in the above embodiments. It can solve the technical problem of thin plate jitter caused by the nonlinear cutting force resulting from the high-speed movement of the cutting head and the interaction between the laser and the material. Compared with the prior art, the beneficial effects of the three-dimensional jitter compensation device provided in this application are the same as those of the three-dimensional jitter compensation method for laser cutting provided in the above embodiments. Furthermore, other technical features of this three-dimensional jitter compensation device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0169] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0170] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0171] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the three-dimensional jitter compensation method for laser cutting in the above embodiments.

[0172] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, radio frequency (RF), etc., or any suitable combination thereof.

[0173] The aforementioned computer-readable storage medium may be included in the three-dimensional jitter compensation device; or it may exist independently and not assembled into the three-dimensional jitter compensation device.

[0174] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the three-dimensional jitter compensation device, the device performs the following actions: 1) Fast Fourier Transform on the vibration signal of the sheet metal to be processed in the Z-axis direction to obtain spectral characteristics; 2) Frequency domain filtering is performed based on the spectral characteristics and a preset jitter filtering parameter set to generate a filtered signal; 3) The filtered signal is reconstructed into a time-domain compensation quantity through an inverse Fourier Transform; 4) Based on the time-domain compensation quantity, a weighted fusion is performed on the cutting acceleration and cutting displacement deviation in the current cutting direction to generate a three-dimensional compensation vector; 5) The three-dimensional compensation vector is superimposed onto the original cutting path coordinates through a motion controller to correct the following state of the cutting head.

[0175] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0176] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0177] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0178] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described three-dimensional jitter compensation method for laser cutting. This solves the technical problem of jitter in thin plates caused by the nonlinear cutting force resulting from the high-speed movement of the cutting head and the interaction between the laser and the material. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the three-dimensional jitter compensation method for laser cutting provided in the above embodiments, and will not be repeated here.

[0179] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A three-dimensional jitter compensation method for laser cutting, characterized in that, The method includes: The spectral characteristics are obtained by performing a fast Fourier transform on the vibration signal of the thin plate to be processed in the Z-axis direction. Based on the spectral characteristics, frequency domain filtering is performed using a preset set of jitter filtering parameters to generate a filtered signal. The filtered signal is reconstructed into a time-domain compensation quantity through inverse Fourier transform; Based on the time-domain compensation amount, a three-dimensional compensation vector is generated by weighted fusion of the cutting acceleration and cutting displacement deviation in the current cutting direction. The three-dimensional compensation vector is superimposed onto the original cutting path coordinates via a motion controller to correct the following state of the cutting head.

2. The three-dimensional jitter compensation method for laser cutting as described in claim 1, characterized in that, Before the step of performing a fast Fourier transform on the vibration signal of the thin plate to be processed in the Z-axis direction to obtain the frequency domain spectrum, the method further includes: By acquiring the signal of the sheet to be processed in the Z-axis direction at a certain moment and filtering the signal, the original signal is obtained; Based on the original signal, adaptive filtering is performed to obtain the vibration signal.

3. The three-dimensional jitter compensation method for laser cutting as described in claim 1, characterized in that, The step of generating a filtered signal by performing frequency domain filtering based on the spectral characteristics and a preset dithering filter parameter set includes: Based on the spectral characteristics, the target frequency band related to the resonance of the thin plate to be processed is extracted; Based on a preset set of jitter filtering parameters, the vibration components in the target frequency band are subjected to frequency domain filtering to generate the filtered signal.

4. The three-dimensional jitter compensation method for laser cutting as described in claim 1, characterized in that, Before the step of generating a three-dimensional compensation vector by weighted fusion of the cutting acceleration and cutting displacement deviation in the current cutting direction based on the time-domain compensation amount, the method further includes: Based on the acceleration signal, an adaptive filtering method is used to obtain the filtered acceleration signal; Based on the filtered acceleration signal, digital sampling is performed to obtain the cutting acceleration corresponding to the timestamp of the acceleration signal; and... Based on the displacement data signal, displacement data information is obtained through calculation and processing; The displacement data is compared with the preset displacement path to obtain the cutting displacement deviation.

5. The three-dimensional jitter compensation method for laser cutting as described in claim 1, characterized in that, The step of generating a three-dimensional compensation vector by weighted fusion of the time-domain compensation amount and the cutting acceleration and cutting displacement deviation in the current cutting direction includes: Based on the time-domain compensation amount, the cutting acceleration, and the cutting displacement deviation, a weighted fusion is performed according to a preset weight to generate a deviation compensation amount; The deviation compensation amount is broken down into compensation amounts for each axis according to a preset rule; The axial compensation amounts are superimposed as spatial vectors using a three-dimensional coordinate transformation matrix to generate a three-dimensional compensation vector containing amplitude and direction information.

6. The three-dimensional jitter compensation method for laser cutting as described in claim 1, characterized in that, The step of superimposing the three-dimensional compensation vector onto the original cutting path coordinates via a motion controller to correct the following state of the cutting head includes: Based on the three-dimensional compensation vector, it is decomposed into translation compensation components and rotation compensation components through a coordinate transformation matrix; The motion controller superimposes the translation compensation component and the rotation compensation component onto the original cutting path coordinates to obtain the corrected target coordinates. The motion controller corrects the following state based on the target coordinates.

7. The three-dimensional jitter compensation method for laser cutting as described in claim 1, characterized in that, After the step of superimposing the three-dimensional compensation vector onto the original cutting path coordinates via a motion controller to correct the following state of the cutting head, the method further includes: When the cutting head hits the plate during movement, a collision alarm will be triggered and the cutting head will be automatically raised to a safe height. If the collision alarm does not stop, the cutting head will continue to be lifted. If the collision alarm stops, the cutting head will no longer be raised, and the three-dimensional compensation vector will be generated based on the current following state to update the following state.

8. The three-dimensional jitter compensation method for laser cutting as described in claim 4, characterized in that, After the step of processing the displacement data signal and comparing it with a preset displacement path to obtain the cutting displacement deviation, the method further includes: Generate a dataset by taking the vibration signal, the cutting acceleration, and the cutting displacement deviation according to their corresponding timestamps; The dataset is divided using cross-validation, with half of the dataset used as the validation set and the remaining dataset used as the training set. The validation set is used to generate a three-dimensional compensation validation vector through the three-dimensional jitter compensation method of laser cutting, and the training set is input into the basic frequency domain filtering model to output a three-dimensional compensation training vector; The three-dimensional compensation verification vector is input into the basic frequency domain filtering model and compared with the three-dimensional compensation training vector, and feedback compensation parameters are generated based on the comparison results; Based on the feedback compensation parameters, the model parameters of the basic frequency domain filtering model are adjusted, and the basic frequency domain filtering model is repeatedly trained to obtain the target frequency domain filtering model.

9. A three-dimensional jitter compensation device, characterized in that, The three-dimensional jitter compensation device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the three-dimensional jitter compensation method for laser cutting as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the three-dimensional jitter compensation method for laser cutting as described in any one of claims 1 to 8.

Images