Intelligent optimization method and system for laser cutting process parameters

Abstract

The invention provides an intelligent optimization method and system for laser cutting process parameters, and relates to the technical field of laser cutting, and the method comprises the steps: firstly obtaining a radiance image sequence of a cutting region and a torque current signal of a servo driver, and extracting a temperature rise gradient sequence and a harmonic distortion sequence from the radiance image sequence and the torque current signal; performing time sequence alignment and vectorization combination on the two sequences after dimension unification to obtain a first vector sequence; then constructing a distribution mapping relation between the temperature rise gradient and the current harmonic based on the first vector sequence, and calculating a joint probability density value between the temperature rise gradient and the current harmonic by using an Archimedes Copula function to determine a coupling evaluation factor; mapping according to the coupling evaluation factor to obtain a target damping coefficient, and differentiating the target damping coefficient with a reference damping value to obtain an impedance deviation value; and finally, respectively adjusting an output duty ratio and a feeding speed value based on the impedance deviation value so as to output an optimal power parameter and an optimal speed parameter. According to the invention, closed-loop self-adaptive control of the cutting quality of the thick plate is realized.

Description

Technical Field

[0001] This application relates to the technical field of laser cutting, and in particular to a method and system for intelligent optimization of laser cutting process parameters. Background Technology

[0002] Laser cutting technology, with its advantages of high precision and high efficiency, has been widely used in high-end equipment manufacturing fields such as aerospace, automobile manufacturing, and construction machinery. As industrial manufacturing continues to demand higher processing quality and efficiency, how to achieve precise optimization of process parameters for high-power continuous laser cutting of extra-thick carbon steel plates has become a key focus of the industry.

[0003] Currently, the optimization of laser cutting process parameters usually adopts experimental modeling or empirical formulas. By collecting temperature information of the cutting area or equipment operating status information, the correspondence between process parameters and cutting quality is established. These methods generally process thermal field parameters and force field parameters separately and adjust them independently.

[0004] However, in the scenario of continuous cutting of thick plates with high power, there is a complex nonlinear coupling relationship between the changes in the thermophysical properties of the material caused by long-term heat accumulation and the mechanical load of the feed drive system. Existing methods are difficult to capture this coupling characteristic, which leads to quality defects such as slag on the cross-section, overheating, or incomplete cutting in the later stage of processing. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for intelligent optimization of laser cutting process parameters, so as to solve the problem of lack of deep integration of thermal coupling characteristics in the prior art.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a method for intelligent optimization of laser cutting process parameters, comprising: The radiometric image sequence of the cut area and the torque current signal of the servo driver are acquired. The temperature rise gradient sequence is extracted from the radiometric image sequence. The torque current signal is subjected to fast Fourier transform to obtain the harmonic distortion sequence of the current in each sampling period. By performing time-series alignment and vectorization combination on the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence, a first vector sequence is obtained. Based on the first vector sequence, a first distribution mapping relationship of temperature rise gradient and a second distribution mapping relationship of current harmonics are constructed, and the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship is calculated by using the Archimedes Copula function to determine the coupling evaluation factor between thermal stress and mechanical resistance. Based on the coupling evaluation factor, the target damping coefficient is determined according to the preset mapping relationship between the evaluation factor and the damping coefficient. The target damping coefficient is then differentially calculated with the preset reference damping value to obtain the impedance deviation. The power correction is calculated based on the impedance deviation and the preset power gain coefficient to adjust the output duty cycle, and the speed scaling factor is calculated based on the impedance deviation and the preset speed adjustment parameter to adjust the feed speed value, thus obtaining the target power parameter and the target speed parameter.

[0007] Optionally, the step of obtaining a first vector sequence by temporally aligning and vectorizing the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence to obtain a first vector sequence includes: By calculating the difference between the first timestamp sequence of the radiance image sequence and the second timestamp sequence of the torque current signal, time point matching is performed to establish the correspondence between the temperature rise gradient value in the temperature rise gradient sequence and the harmonic distortion value in the harmonic distortion sequence. The first range is obtained by calculating the difference between the maximum and minimum values ​​of the temperature rise gradient in the temperature rise gradient sequence. The ratio of the difference between each temperature rise gradient value and the minimum value in the temperature rise gradient sequence to the first range is calculated to obtain the temperature rise gradient value after dimension unification. The second range is obtained by calculating the difference between the maximum and minimum values ​​of the harmonic distortion values ​​in the harmonic distortion sequence. The ratio of the difference between each harmonic distortion value and the minimum value in the harmonic distortion sequence to the second range is calculated to obtain the dimension-unified harmonic distortion value for each harmonic distortion value. Based on the correspondence, the dimensionally unified temperature rise gradient value and the corresponding dimensionally unified harmonic distortion value are combined into a two-dimensional vector, and all the two-dimensional vectors are arranged in chronological order according to the timestamps to form a first vector sequence.

[0008] Optionally, constructing a first distribution mapping relationship for the temperature rise gradient and a second distribution mapping relationship for the current harmonics based on the first vector sequence includes: Determine the number of two-dimensional vectors in the first vector sequence; Extract the first and second components of each two-dimensional vector from the first vector sequence to obtain the temperature rise gradient value sequence and harmonic distortion value sequence. Arrange the values ​​in the temperature rise gradient value sequence in ascending order, and calculate the ratio of the frequency of each value to the number of values ​​after arrangement to obtain the first distribution mapping relationship; The values ​​in the harmonic distortion value sequence are arranged in ascending order, and the ratio of the frequency of each value to the total number of values ​​is calculated to obtain the second distribution mapping relationship.

[0009] Optionally, determining the coupling evaluation factor between thermal stress and mechanical resistance by calculating the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship using the Archimedes Copula function includes: Substitute each value in the temperature rise gradient value sequence into the first distribution mapping relationship to obtain the first probability value corresponding to each temperature rise gradient value; Substitute each value in the harmonic distortion value sequence into the second distribution mapping relationship to obtain the second probability value corresponding to each harmonic distortion value; The first probability value and the corresponding second probability value are combined to form a probability pair, and the joint probability density value of all the probability pairs is calculated by the Archimedes Copula function. The relevant parameter in the Archimedes Copula function that maximizes the sum of all the joint probability density values ​​is determined as the coupling evaluation factor.

[0010] Optionally, the step of determining the target damping coefficient based on the coupling evaluation factor and according to the preset mapping relationship between the evaluation factor and the damping coefficient, and performing a differential operation between the target damping coefficient and the preset reference damping value to obtain the impedance deviation includes: The evaluation factor interval in which the coupled evaluation factor is located is found in the mapping relationship table between the evaluation factor and the damping coefficient, and the target damping coefficient corresponding to the evaluation factor interval is determined. The mapping relationship table is constructed in advance by cutting steel plates of various thicknesses. The difference between the target damping coefficient and the preset reference damping value is calculated to obtain the impedance deviation. The preset reference damping value is the target damping coefficient obtained during the cutting process under ideal processing conditions.

[0011] Optionally, the step of calculating a power correction based on the impedance deviation and a preset power gain coefficient to adjust the output duty cycle, and calculating a speed scaling factor based on the impedance deviation and a preset speed adjustment parameter to adjust the feed rate, to obtain target power parameters and target speed parameters, includes: The impedance deviation is multiplied by a preset power gain coefficient to obtain a power correction amount, and the current output duty cycle is added to the power correction amount to obtain the adjusted output duty cycle. The impedance deviation is multiplied by the preset speed adjustment parameter to obtain the speed scaling factor, and the current feed speed value is added to the speed scaling factor to obtain the adjusted feed speed value. Determine whether the adjusted output duty cycle is greater than a preset power upper limit or less than a preset power lower limit. If the adjusted output duty cycle is greater than the preset power upper limit, determine the preset power upper limit as the target power parameter. If the adjusted output duty cycle is less than the preset power lower limit, determine the preset power lower limit as the target power parameter. Otherwise, determine the adjusted output duty cycle as the target power parameter. Determine whether the adjusted feed rate value is greater than a preset upper speed limit or less than a preset lower speed limit. If the adjusted feed rate value is greater than the preset upper speed limit, determine the preset upper speed limit as the target speed parameter. If the adjusted feed rate value is less than the preset lower speed limit, determine the preset lower speed limit as the target speed parameter. Otherwise, determine the adjusted feed rate value as the target speed parameter.

[0012] Optionally, before calculating the power correction amount based on the impedance deviation and a preset power gain coefficient to adjust the output duty cycle, the method further includes: Extract the set of pixels in the radiance image sequence whose luminance scores exceed a preset luminance threshold to obtain the bright areas; The area percentage is obtained by calculating the ratio of the number of pixels in all the highlighted areas to the total number of pixels in a single frame of the radiance image sequence. Calculate the geometric distance between the mean coordinates of the center points of each of the highlighted areas and the real-time physical coordinates of the cutting head to obtain the centroid offset distance, and calculate the spatter density by multiplying the area ratio by the centroid offset distance. The target adjustment factor corresponding to the splash density is determined by a preset mapping table between splash density and adjustment factor, and the power gain coefficient is adjusted according to the target adjustment factor.

[0013] Secondly, this application provides an intelligent optimization system for laser cutting process parameters, comprising: The acquisition module is used to acquire the radiance image sequence of the cutting area and the torque current signal of the servo driver, extract the temperature rise gradient sequence from the radiance image sequence, and perform a fast Fourier transform on the torque current signal to obtain the harmonic distortion sequence of the current in each sampling period. The combination module is used to obtain a first vector sequence by performing time-series alignment and vectorization combination of the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence; The construction module is used to construct a first distribution mapping relationship of temperature rise gradient and a second distribution mapping relationship of current harmonics based on the first vector sequence, and to determine the coupling evaluation factor between thermal stress and mechanical resistance by calculating the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship using the Archimedes Copula function. The determination module is used to determine the target damping coefficient based on the coupling evaluation factor and according to the preset mapping relationship between the evaluation factor and the damping coefficient, and to perform a differential operation between the target damping coefficient and the preset reference damping value to obtain the impedance deviation. The calculation module is used to calculate the power correction amount based on the impedance deviation and the preset power gain coefficient to adjust the output duty cycle, and to calculate the speed scaling factor based on the impedance deviation and the preset speed adjustment parameter to adjust the feed speed value, so as to obtain the target power parameter and the target speed parameter.

[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to execute the computer program to implement the steps of the intelligent optimization method for laser cutting process parameters as described in the first aspect above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the intelligent optimization method for laser cutting process parameters as described in the first aspect above.

[0016] The intelligent optimization method for laser cutting process parameters provided in this application has the following beneficial effects: First, this application acquires a radiance image sequence and a torque current signal, and extracts a temperature rise gradient sequence and harmonic distortion sequence from them to provide basic data for subsequent thermo-mechanical coupling analysis. Then, the two sequences after dimension unification are time-aligned and vectorized to obtain a first vector sequence, thereby eliminating the dimensional differences and time asynchrony between the thermal field and force field data. Then, based on the first vector sequence, a distribution mapping relationship between the temperature rise gradient and current harmonics is constructed, and the joint probability density value between the two is calculated using the Archimedes Copula function to determine the coupling evaluation factor, thereby quantifying the degree of nonlinear correlation between thermal stress and mechanical resistance. Then, the target damping coefficient is obtained by mapping according to the coupling evaluation factor and differentially calculated with the reference damping value to obtain the impedance deviation. The statistical characteristics of thermo-mechanical coupling are then transformed into quantifiable physical deviation indicators. Finally, based on the impedance deviation, the power correction and speed scaling factors are calculated respectively, and the output duty cycle and feed rate values ​​are adjusted to output the optimal power parameters and speed parameters, thereby realizing closed-loop adaptive adjustment of process parameters.

[0017] Furthermore, this application substitutes each value in the temperature rise gradient value sequence into the first distribution mapping relationship to obtain a first probability value, and substitutes each value in the harmonic distortion value sequence into the second distribution mapping relationship to obtain a second probability value. Then, the first probability value and the corresponding second probability value are combined to form a probability pair, and the joint probability density value of all probability pairs is calculated using the Archimedes Copula function. The correlation parameter that maximizes the sum of all joint probability density values ​​is determined as the coupling evaluation factor. Subsequently, by mapping the thermal field and force field data to the probability space and fitting their joint distribution using the Copula function, the tail correlation characteristics between thermal stress and mechanical resistance can be accurately captured, thereby obtaining an evaluation index that quantitatively describes the degree of asymmetric coupling between the two. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of 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, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating an intelligent optimization method for laser cutting process parameters provided in this application embodiment; Figure 2 A schematic diagram illustrating a specific implementation of an intelligent optimization method for laser cutting process parameters provided in this application embodiment; Figure 3 A schematic diagram of a laser cutting process parameter intelligent optimization system provided in this application embodiment; Figure 4 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0020] In the process of high-power continuous laser cutting of extra-thick carbon steel plates, long-term heat accumulation will cause changes in the thermophysical properties of the material. These changes form a complex nonlinear coupling relationship with the mechanical load of the feed drive system. However, existing optimization methods usually treat thermal field parameters and force field parameters as independent factors, making it difficult to effectively capture the dynamic correlation between them. This leads to quality defects such as slag on the cross-section, overheating, or incomplete cutting in the later stages of processing, and restricts the stability and processing quality of thick plate cutting.

[0021] To address the aforementioned issues, this application proposes an intelligent optimization method for laser cutting process parameters. This method first acquires the radiation brightness image of the cutting area and the torque current signal of the servo driver, extracting the temperature rise gradient information reflecting thermal field changes and the current harmonic information reflecting force field changes, respectively. The two types of data are then time-aligned and fused. Based on this, a statistical analysis model is used to establish a correlation mapping between thermal stress and mechanical resistance, quantitatively assessing their coupling degree. Furthermore, the coupling degree is converted into impedance deviation. Based on this, the output duty cycle of the laser power and the cutting feed speed are dynamically adjusted so that the process parameters can adaptively match the current processing state. In this way, this application achieves deep fusion and closed-loop adjustment of thermo-mechanical coupling characteristics, effectively solving the nonlinear coupling problem that existing methods struggle to address, thereby improving the stability and cross-sectional quality of the thick plate cutting process.

[0022] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] The core of this application is to provide an intelligent optimization method for laser cutting process parameters, and a flowchart of one specific implementation is shown below. Figure 1 As shown, the method includes: S101. Obtain the radiance image sequence of the cutting area and the torque current signal of the servo driver, extract the temperature rise gradient sequence from the radiance image sequence, and perform a fast Fourier transform on the torque current signal to obtain the harmonic distortion sequence of the current in each sampling period. In this sub-step, the radiance image sequence refers to the set of image data arranged in chronological order of the radiance values ​​of the cut area captured in real time by the infrared thermal imaging acquisition unit. The temperature rise gradient sequence refers to the set of data extracted from the radiance image sequence that reflects the rate of temperature change within the heat-affected zone of the cut area, used to characterize the thermal field evolution characteristics. The torque current signal refers to the time-series data of the torque current output by the servo driver, read in real time through the CNC system bus. The harmonic distortion sequence refers to the set of data on the degree of distortion of the current waveform in each sampling period after performing a fast Fourier transform on the torque current signal, used to characterize the severity of mechanical load fluctuations.

[0024] In this embodiment of the application, in the scenario of high-power continuous laser cutting of extra-thick carbon steel plates, the radiance image sequence of the cutting area is first captured in real time by an infrared thermal imaging acquisition unit arranged on the side of the cutting head. Each frame of the radiance image sequence records the radiance value of each pixel in the cutting area. Then, the boundary of the heat-affected zone is identified from each frame of the radiance image, and multiple feature points within the heat-affected zone are selected. The change in temperature of the feature points between adjacent frames is calculated and divided by the time interval to obtain the temperature rise gradient value. The feature point temperature is obtained by converting the pixel brightness value in the radiance image through a pre-calibrated brightness and temperature model. Then, the temperature rise gradient values ​​calculated from all frames are arranged in chronological order to form a temperature rise gradient sequence.

[0025] The torque current signal of the servo drive is read in real time through the CNC system bus. The torque current signal is a continuous time series data collected at a fixed sampling frequency. Then, the torque current data collected in each sampling period is taken as a data segment, and a fast Fourier transform is performed on each data segment to obtain the frequency domain distribution of the current signal in that sampling period. The ratio of each harmonic amplitude to the fundamental amplitude is calculated and accumulated to obtain the harmonic distortion value of that sampling period. The harmonic distortion values ​​calculated in all sampling periods are arranged in chronological order to form a harmonic distortion sequence.

[0026] This application, by simultaneously acquiring raw data of the thermal and force fields, can provide basic information for subsequent thermo-mechanical coupling analysis.

[0027] S102. By performing time-series alignment and vectorization combination on the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence, a first vector sequence is obtained. In one specific implementation, step S102 includes: Step 1021: By calculating the difference between the first timestamp sequence of the radiance image sequence and the second timestamp sequence of the torque current signal, time point matching is performed to establish the correspondence between the temperature rise gradient value in the temperature rise gradient sequence and the harmonic distortion value in the harmonic distortion sequence. In this sub-step, the first timestamp sequence refers to the data set arranged sequentially according to the acquisition time points corresponding to each frame of the radiance image sequence. The second timestamp sequence refers to the data set arranged sequentially according to the acquisition time points corresponding to each sampling point in the torque current signal. The correspondence refers to the one-to-one association mapping established between the temperature rise gradient values ​​in the temperature rise gradient sequence and the harmonic distortion values ​​in the harmonic distortion sequence through time matching.

[0028] In this embodiment, firstly, a first timestamp sequence is obtained for each frame of the radiance image sequence, and simultaneously, a second timestamp sequence is obtained for each sampling point of the torque current signal. Since the sampling frequency of the infrared thermal imaging acquisition unit differs from that of the servo driver, the sampling frequencies of the two sets of data differ, and the time points cannot be directly aligned. Therefore, for each time point in the second timestamp sequence, the absolute value of the difference between that time point and each time point in the first timestamp sequence is calculated, and the first timestamp with the smallest absolute difference is selected as the matching time point. Then, a correspondence is established between the temperature rise gradient value corresponding to the matching time point and the harmonic distortion value corresponding to the current time point. For example, the k-th time point in the second timestamp sequence is... The first timestamp sequence contains time points. and ,like Then The corresponding temperature gradient value and Establish a correspondence between the corresponding harmonic distortion values.

[0029] Step 1022: Calculate the difference between the maximum and minimum values ​​of the temperature rise gradient in the temperature rise gradient sequence to obtain the first range. Calculate the ratio of the difference between each temperature rise gradient value and the minimum value in the temperature rise gradient sequence to the first range to obtain the temperature rise gradient value after dimension unification. In this sub-step, the first range refers to the difference between the maximum and minimum values ​​in the temperature rise gradient sequence, reflecting the overall numerical range of the temperature rise gradient values. The dimensionally unified temperature rise gradient value refers to the value obtained by mapping the original temperature rise gradient values ​​to a preset numerical range using the range standardization method.

[0030] In this embodiment of the application, after obtaining the temperature rise gradient sequence, all temperature rise gradient values ​​in the sequence are traversed, and the maximum value is found by comparison. and minimum value and calculate The first range was obtained For each temperature gradient value in the temperature gradient sequence... Through the formula: Calculate the corresponding dimensionless temperature rise gradient value.

[0031] Step 1023: Calculate the difference between the maximum and minimum values ​​of the harmonic distortion values ​​in the harmonic distortion sequence to obtain the second range. Calculate the ratio of the difference between each harmonic distortion value and the minimum value in the harmonic distortion sequence to the second range to obtain the dimensionally unified harmonic distortion value for each harmonic distortion value. In this sub-step, the second range refers to the difference between the maximum and minimum values ​​in the harmonic distortion sequence, reflecting the overall numerical range of the harmonic distortion values. The dimensionally unified harmonic distortion value refers to the value obtained by mapping the original harmonic distortion values ​​to a preset numerical range using the range standardization method.

[0032] In this embodiment of the application, after obtaining the harmonic distortion sequence, all harmonic distortion values ​​in the sequence are traversed, and the maximum value is found by comparison. and minimum value and calculate The first range was obtained Furthermore, for each harmonic distortion value in the harmonic distortion sequence... Through the formula: Calculate the corresponding dimensionally unified harmonic distortion value.

[0033] Step 1024: Based on the correspondence, combine the dimensionally unified temperature rise gradient value and the corresponding dimensionally unified harmonic distortion value into a two-dimensional vector, and arrange all the two-dimensional vectors in chronological order according to the timestamps to form a first vector sequence.

[0034] In this sub-step, a two-dimensional vector refers to a data unit consisting of two numerical values ​​arranged in a fixed order. The first vector sequence refers to a set of vectors formed by arranging multiple two-dimensional vectors in chronological order.

[0035] In this embodiment, based on the correspondence established in step 1021, for each successfully matched time point, the dimensionally unified temperature rise gradient value and the dimensionally unified harmonic distortion value corresponding to that time point are obtained, and the two values ​​are combined into a two-dimensional vector, where the first component is the dimensionally unified temperature rise gradient value and the second component is the dimensionally unified harmonic distortion value. All two-dimensional vectors are then arranged according to the order of the matching timestamps to form a first vector sequence. For example, at three consecutive time points, the dimensionally unified temperature rise gradient values ​​are as follows: The corresponding dimensionlessly unified harmonic distortion values ​​are as follows: The resulting two-dimensional vectors are as follows: Each two-dimensional vector is treated as an independent data unit, and these two-dimensional vectors are arranged in chronological order to obtain the first vector sequence.

[0036] This application eliminates the time discrepancy and dimensional difference between thermal field and force field data by aligning the time sequence and unifying the dimensions, thereby making the two types of data comparable.

[0037] S103. Based on the first vector sequence, construct a first distribution mapping relationship of temperature rise gradient and a second distribution mapping relationship of current harmonics, and determine the coupling evaluation factor between thermal stress and mechanical resistance by using the Archimedes Copula function to calculate the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship. In one specific implementation, such as Figure 2 As shown, step S103 includes: Step 1031: Determine the number of two-dimensional vectors in the first vector sequence; In this embodiment, the first vector sequence generated in step S102 is obtained, all elements in the sequence are traversed, and the total number of two-dimensional vectors in the sequence is counted. For example, the first vector sequence is represented as follows: , of which each If a vector is a two-dimensional vector, then the number of two-dimensional vectors is n.

[0038] Step 1032: Extract the first and second components of each two-dimensional vector from the first vector sequence to obtain the temperature rise gradient value sequence and harmonic distortion value sequence; In this embodiment of the application, each two-dimensional vector in the first vector sequence is traversed, and for the i-th two-dimensional vector... Extract the first component Input the temperature gradient value sequence and extract the second component. After inputting the harmonic distortion value sequence and traversing it, the temperature rise gradient value sequence is obtained. and , where n is the number of two-dimensional vectors determined in step 1031.

[0039] Step 1033: Arrange the values ​​in the temperature rise gradient value sequence in ascending order, and calculate the ratio of the frequency of each value to the number of values ​​after arrangement to obtain the first distribution mapping relationship; In this embodiment of the application, the temperature rise gradient value sequence obtained in step 1032 is... Sort the data in ascending order to obtain an ordered sequence. For each unique value u in the ordered sequence, count the number of times that value appears in the original sequence. And calculate the ratio of the frequency of this value to the number of two-dimensional vectors n. And associate each unique value u with its corresponding probability value The corresponding data is stored to form the first distribution mapping relationship. This mapping relationship can be represented in the form of a lookup table, for example... correspond , correspond .

[0040] Step 1034: Arrange the values ​​in the harmonic distortion value sequence in ascending order, and calculate the ratio of the frequency of each value to the number of values ​​after arrangement to obtain the second distribution mapping relationship.

[0041] In this embodiment of the application, the harmonic distortion value sequence obtained in step 1032 is... Sort the data in ascending order to obtain an ordered sequence. For each unique value v in the ordered sequence, count the number of times that value appears in the original sequence. And calculate the ratio of the frequency of this value to the number of two-dimensional vectors n. And associate each unique value v with its corresponding probability value. The corresponding data is stored to form a second distribution mapping relationship. This mapping relationship can be represented in the form of a lookup table, for example... correspond , correspond .

[0042] Step 1035: Substitute each value in the temperature rise gradient value sequence into the first distribution mapping relationship to obtain the first probability value corresponding to each temperature rise gradient value; In this embodiment of the application, the temperature rise gradient value sequence obtained in step 1032 is traversed. For each temperature gradient value in the sequence The probability value corresponding to the given value is found in the first distribution mapping relationship constructed in step 1033. Since the first distribution mapping relationship is stored in the form of a lookup table, the corresponding probability value can be obtained directly through numerical matching. Therefore, the found probability value is used as the first probability value corresponding to the temperature rise gradient value, and finally the first probability value sequence is obtained. .

[0043] Step 1036: Substitute each value in the harmonic distortion value sequence into the second distribution mapping relationship to obtain the second probability value corresponding to each harmonic distortion value; In this embodiment of the application, the harmonic distortion value sequence obtained in step 1032 is traversed. For each harmonic distortion value v in the sequence i In the second distribution mapping relationship constructed in step 1033, find the probability value corresponding to the value. Since the second distribution mapping relationship is stored in the form of a lookup table, the corresponding probability value can be obtained directly through numerical matching. Therefore, the probability value found is used as the second probability value corresponding to the harmonic distortion value, and finally the second probability value sequence is obtained. .

[0044] Step 1037: Form a probability pair by combining the first probability value with the corresponding second probability value, and calculate the joint probability density value of all the probability pairs using the Archimedes Copula function; In this embodiment of the application, the first probability value sequence obtained in step 1035 is paired with the values ​​of the same position indices in the second probability value sequence obtained in step 1036. For the i-th position, the values ​​of the same position indices are paired. and Forming probability pairs ( , ), to obtain the probability pair of the set Then choose one from the Archimedes Copula family of functions, such as the Gumbel Copula function, whose function form is: ,in, For the relevant parameters, u and v are the first probability value and the second probability value, respectively. Then, for each probability pair ( , Substitute this probability pair into the Copula function to calculate the joint probability density value. The joint probability density value sequence is obtained. .

[0045] Step 1038: Determine the relevant parameter in the Archimedes Copula function that maximizes the sum of all the joint probability density values ​​as the coupling evaluation factor.

[0046] In this sub-step, the correlation parameter refers to the parameter in the Archimedes Copula function used to control the degree of correlation between variables, denoted as The coupling evaluation factor is a relevant parameter that maximizes the sum of all joint probability density values, and is used to quantitatively describe the degree of asymmetric coupling between thermal stress and mechanical resistance.

[0047] In this embodiment, the calculation result of the joint probability density value of the Archimedes Copula function depends on relevant parameters. The value of is determined, so a search range is set, for example... Within this interval, traverse different... Value, and then for each The joint probability density values ​​of all probability pairs are calculated according to the method in step 1037, and then the sum of these joint probability density values ​​is calculated. And after the traversal is complete, find the one that makes Reaching the maximum value The value, will The value is used as a relevant parameter and determined as the output of the coupled evaluation factor.

[0048] This application establishes a thermodynamic joint distribution model using the Copula function to quantify the degree of nonlinear correlation between thermal stress and mechanical resistance, thereby generating a coupling evaluation factor.

[0049] S104. Based on the coupling evaluation factor, determine the target damping coefficient according to the preset mapping relationship between the evaluation factor and the damping coefficient, and perform a differential operation between the target damping coefficient and the preset reference damping value to obtain the impedance deviation. In one specific implementation, step S104 includes: Step 1041: Find the evaluation factor interval where the coupled evaluation factor is located in the mapping relationship table between the evaluation factor and the damping coefficient, and determine the target damping coefficient corresponding to the evaluation factor interval. The mapping relationship table is constructed in advance by cutting steel plates of various thicknesses. In this sub-step, the mapping table between evaluation factors and damping coefficients refers to a lookup table constructed by pre-testing extra-thick carbon steel plates of various thicknesses, recording the coupled evaluation factors and corresponding processing damping coefficients generated at different cutting stages, dividing the coupled evaluation factors into multiple continuous intervals, and assigning a damping coefficient to each interval. An evaluation factor interval refers to multiple continuous sub-intervals into which the value range of the coupled evaluation factor is divided, with each sub-interval corresponding to a damping coefficient. The target damping coefficient refers to the damping coefficient value read from the mapping table based on the evaluation factor interval in which the current coupled evaluation factor is located.

[0050] In this embodiment of the application, the coupling evaluation factor output in step S103 is obtained. The coupling evaluation factor is a value within a preset range, such as between 1 and 8. Then, the coupling evaluation factor is compared with the preset evaluation factor intervals in the mapping relationship table between the evaluation factor and the damping coefficient, and it is determined which evaluation factor interval the coupling evaluation factor falls into. The mapping table between evaluation factors and damping coefficients is pre-constructed. The construction process is as follows: Select extra-thick carbon steel plates of various thicknesses, such as 25 mm, 30 mm, 35 mm, and 40 mm, and conduct cutting tests under various processing conditions. Record the coupled evaluation factors and corresponding processing damping coefficients for each test. Then, divide the value range of the coupled evaluation factors into multiple continuous intervals, such as 1 to 2, 2 to 3, 3 to 4, 4 to 5, 5 to 6, 6 to 7, and 7 to 8. Calculate the average value of the damping coefficients in all test data falling into each interval, and then use this average value as the damping coefficient corresponding to that interval. After determining the evaluation factor interval in which the coupled evaluation factor lies, the damping coefficient corresponding to that interval is read from the mapping table, and this damping coefficient is used as the target damping coefficient. For example, the mapping table can be shown in Table 1.

[0051] Table 1: Mapping Relationship between Evaluation Factors and Damping Coefficients

[0052] Step 1042: Calculate the difference between the target damping coefficient and the preset reference damping value to obtain the impedance deviation. The preset reference damping value is the target damping coefficient obtained during the cutting process under ideal processing conditions.

[0053] In this sub-step, the reference damping value refers to the damping coefficient value obtained by the same method and determined according to step 1041 during the cutting process under ideal processing conditions, i.e., without thermo-coupling deviation. The impedance deviation is the difference between the target damping coefficient and the reference damping value, used to characterize the comprehensive physical index of the deviation of the system's equivalent impedance from the ideal state after the material properties are changed due to thermal field evolution.

[0054] In this embodiment of the application, the target damping coefficient determined in step 1041 is obtained and denoted as... Next, a pre-calibrated reference damping value is obtained. The calibration process for this reference damping value is as follows: Under ideal processing conditions, i.e., the plate is at room temperature, the cutting head is in good condition, and there is no heat accumulation, a complete cutting process is performed. Then, the coupling evaluation factor under this ideal condition is calculated according to the methods in steps S101 to S104. Then, the corresponding damping coefficient is determined according to the method in step 1041, and this damping coefficient is used as the reference damping value, denoted as . Then calculate the difference between the target damping coefficient and the reference damping value, i.e. The difference Output as impedance deviation.

[0055] This application enables the quantitative mapping of thermo-coupling statistical characteristics to physical control deviations by converting coupling evaluation factors into impedance deviations.

[0056] S105. Calculate the power correction amount based on the impedance deviation and the preset power gain coefficient to adjust the output duty cycle, and calculate the speed scaling factor based on the impedance deviation and the preset speed adjustment parameter to adjust the feed speed value, thereby obtaining the target power parameter and the target speed parameter.

[0057] In one specific implementation, step S105 includes: Step 1051: Multiply the impedance deviation by the preset power gain coefficient to obtain the power correction amount, and add the current output duty cycle to the power correction amount to obtain the adjusted output duty cycle; In this sub-step, the power gain coefficient is a pre-set scaling factor used to convert the impedance deviation into a power correction value, controlling the adjustment range of the output duty cycle by the impedance deviation. The power correction value is the value obtained by multiplying the impedance deviation by the power gain coefficient, used to adjust the current output duty cycle. The adjusted output duty cycle is the value obtained by adding the current output duty cycle to the power correction value, used to determine the target laser power.

[0058] In this embodiment of the application, the impedance deviation output in step S104 is obtained. The impedance deviation can be positive, negative, or zero. Then, a preset power gain coefficient Kp is obtained. This coefficient is a constant pre-calibrated based on the laser characteristics and cutting process requirements, for example, 0.5. Finally, the impedance deviation is multiplied by the power gain coefficient to obtain the power correction amount. Then, the current output duty cycle Dc is obtained, which represents the percentage of the current laser power to the rated power. The current output duty cycle is then added to the power correction amount to obtain the adjusted output duty cycle. .

[0059] Step 1052: Multiply the impedance deviation by the preset speed adjustment parameter to obtain the speed scaling factor, and add the current feed speed value to the speed scaling factor to obtain the adjusted feed speed value. In this sub-step, the speed adjustment parameter refers to a pre-set scaling factor used to convert the impedance deviation into a speed scaling factor, controlling the adjustment range of the impedance deviation on the feed rate. The speed scaling factor is a value obtained by multiplying the impedance deviation by the speed adjustment parameter, used to adjust the current feed rate. The adjusted feed rate value is the value obtained by adding the current feed rate value to the speed scaling factor, used to determine the target cutting speed.

[0060] In this embodiment of the application, the impedance deviation output in step S104 is obtained. Next, obtain the preset speed adjustment parameter Kv. This coefficient is a constant pre-calibrated according to the cutting process requirements and equipment characteristics, such as 0.3. Then, multiply the impedance deviation by the speed adjustment parameter to obtain the speed scaling factor. Then, the current feed rate value Vc is obtained, which represents the current movement speed of the cutting head. The current feed rate value is added to the speed scaling factor to obtain the adjusted feed rate value. .

[0061] Step 1053: Determine whether the adjusted output duty cycle is greater than the preset power upper limit or less than the preset power lower limit. When the adjusted output duty cycle is greater than the preset power upper limit, determine the preset power upper limit as the target power parameter. When the adjusted output duty cycle is less than the preset power lower limit, determine the preset power lower limit as the target power parameter. Otherwise, determine the adjusted output duty cycle as the target power parameter; In this sub-step, the preset power upper limit refers to the maximum output duty cycle preset according to the rated power of the laser and the process safety requirements. The preset power lower limit refers to the minimum output duty cycle preset according to the stable light output requirements of the laser and the process feasibility. The target power parameter refers to the output duty cycle finally determined after boundary clipping, which is used to control the actual output power of the laser.

[0062] In the embodiment of the present application, obtain the adjusted output duty cycle Da obtained in step 1052, and obtain the preset power upper limit Dmax and the preset power lower limit Dmin; then determine whether the adjusted output duty cycle is greater than the preset power upper limit. If Da > Dmax, then use the preset power upper limit Dmax as the target power parameter; then determine whether the adjusted output duty cycle is less than the preset power lower limit. If Da < Dmin, then use the preset power lower limit Dmin as the target power parameter. If the adjusted output duty cycle is between the upper and lower limits, directly use the adjusted output duty cycle Da as the target power parameter.

[0063] Step 1054: Determine whether the adjusted feed speed value is greater than the preset speed upper limit or less than the preset speed lower limit. When the adjusted feed speed value is greater than the preset speed upper limit, determine the preset speed upper limit as the target speed parameter. When the adjusted feed speed value is less than the preset speed lower limit, determine the preset speed lower limit as the target speed parameter. Otherwise, determine the adjusted feed speed value as the target speed parameter.

[0064] In this sub-step, the preset speed upper limit refers to the maximum feed speed preset according to the performance of the cutting equipment and the process safety requirements. The preset speed lower limit refers to the minimum feed speed preset according to the cutting quality requirements and the stable operation requirements of the equipment. The target speed parameter refers to the feed speed value finally determined after boundary clipping, which is used to control the actual moving speed of the cutting head.

[0065] In an embodiment of the present application, obtain the adjusted feed rate value Va obtained in step 1051, and obtain a preset speed upper limit Vmax and a preset speed lower limit Vmin; then determine whether the adjusted feed rate value is greater than the preset speed upper limit. If Va > Vmax, then use the preset speed upper limit Vmax as the target speed parameter; then determine whether the adjusted feed rate value is less than the preset speed lower limit. If Va < Vmin, then use the preset speed lower limit Vmin as the target speed parameter. If the adjusted feed rate value is between the upper and lower limits, then directly use the adjusted feed rate value Va as the target speed parameter.

[0066] The present application dynamically adjusts the laser power and cutting speed based on the impedance deviation amount, and can achieve closed-loop adaptive adjustment of process parameters, so as to output optimal control parameters.

[0067] S106. Before calculating the power correction amount based on the impedance deviation amount and the preset power gain coefficient to adjust the output duty ratio, the method further includes: Step 1061. Extract a pixel set in the radiation luminance image sequence whose luminance score exceeds a preset luminance threshold to obtain a highlighted area; In this sub-step, the highlighted area refers to an image area formed by clustering the pixel points in the radiation luminance image sequence whose luminance values exceed the preset luminance threshold through connectivity clustering, and each highlighted area corresponds to an independent bright spot of the splash sparks during the cutting process.

[0068] In an embodiment of the present application, obtain a single-frame image in the radiation luminance image sequence collected in step S101. Each pixel point in this image records the radiation luminance value at that position, and then set a preset luminance threshold, which is calibrated according to the cutting test. For example, take 80% of the upper limit of the radiation luminance value range; then traverse all pixel points in the image, and mark the pixel points whose luminance values exceed the preset luminance threshold as candidate points. Then, for the marked candidate points, use the eight-neighborhood connected region labeling algorithm to cluster the adjacent candidate points into connected regions, and each connected region is a highlighted area.

[0069] Step 1062. Calculate the ratio of the number of pixels of all the highlighted areas to the total number of pixels of a single-frame image in the radiation luminance image sequence to obtain the area ratio; In this sub-step, the area ratio refers to the ratio of the total number of pixels contained in all highlighted areas to the total number of pixels of a single-frame image, and is used to quantify the overall degree of spark splash during the cutting process.

[0070] In this embodiment of the application, all the highlighted areas identified in step 1061 are obtained, then each highlighted area is traversed and the number of pixels contained in the area is counted. The number of pixels in all highlighted areas is added together to obtain the total number of splashed pixels Ns. Then the total number of pixels Nt of the current frame image is obtained. This value is determined by the image resolution. For example, when the image resolution is 320×256, the total number of pixels is 81920. Then the area ratio Ra=Ns / Nt is calculated.

[0071] Step 1063: Calculate the geometric distance between the mean coordinates of the center points of each of the bright areas and the real-time physical coordinates of the cutting head to obtain the centroid offset distance, and calculate the spatter density by multiplying the area ratio and the centroid offset distance. In this sub-step, the centroid offset distance refers to the Euclidean distance between the average coordinates of the center points of all highlighted areas and the real-time physical coordinates of the cutting head, used to quantify the degree to which the spatter area deviates from the cutting head position. Spatter density is the product of the area percentage and the centroid offset distance, used to comprehensively characterize the severity of the spatter and the degree of positional deviation.

[0072] In this embodiment of the application, all highlighted regions identified in step 1061 are obtained. If the total number of highlighted regions is denoted as M, and for the j-th highlighted region, the number of pixels contained in that region is denoted as... The sum of the x-coordinates of all pixels in this region is denoted as . The sum of the vertical axes is denoted as Then the coordinates of the center point of the highlighted area are ( , )for: , Then, sum the x-coordinates of the center points of all highlighted areas and divide by the number of highlighted areas M to obtain the average x-coordinate of the center points. Similarly, calculate the mean of the ordinates of the center point. Then, using preset camera intrinsic parameter calibration coefficients, the average coordinates of the center point of the highlighted area in the image space are calculated. , The position is converted into the real-time physical space position, and then compared with the real-time physical coordinates of the cutting head. , Calculate the centroid offset distance Then obtain the area percentage Ra calculated in step 1062, and calculate the splash density F as F = Ra × .

[0073] Step 1064: Determine the target adjustment factor corresponding to the splash density through a preset mapping table of splash density and adjustment factor, and adjust the power gain coefficient according to the target adjustment factor.

[0074] In this sub-step, the mapping table between spatter density and adjustment factor refers to a lookup table pre-constructed through cutting experiments. This table divides the spatter density into multiple continuous intervals and assigns an adjustment factor to each interval, used to adjust the power gain coefficient based on the spatter density. The target adjustment factor is the adjustment factor value obtained by looking up the value in the mapping table based on the current spatter density. The real-time power gain coefficient is the value obtained by multiplying the target adjustment factor by the preset power gain coefficient, used to replace the original power gain coefficient in the calculation of the power correction amount.

[0075] In this embodiment of the application, the spatter density F calculated in step 1063 and the preset mapping table of spatter density and adjustment factor are obtained. The construction process of the mapping table is as follows: select extra-thick carbon steel plates of various thicknesses and conduct experiments under various cutting conditions. Record the spatter density and the power gain coefficient that can make the cutting quality optimal in each experiment. Then divide the value range of spatter density into multiple continuous intervals, such as 0 to 0.2, 0.2 to 0.4, 0.4 to 0.6, 0.6 to 0.8, and 0.8 to 1.0. Calculate the ratio of the optimal power gain coefficient to the reference power gain coefficient corresponding to the spatter density falling into each interval. Use this ratio as the adjustment factor corresponding to the interval. The mapping table of spatter density and adjustment factor is shown in Table 2. Table 2: Mapping Table of Splash Density and Adjustment Factor

[0076] Then, obtain the preset power gain coefficient and calculate the product of the target adjustment factor and the preset power gain coefficient to adjust the power gain coefficient.

[0077] This application improves control stability by real-time correction of the power gain coefficient through splash characteristics and reducing the interference of spark noise on process parameter adjustment.

[0078] Figure 3 This is a schematic diagram of a specific implementation of an intelligent optimization system for laser cutting process parameters provided in this application, referring to... Figure 3 The system may include: The acquisition module 31 is used to acquire the radiance image sequence of the cutting area and the torque current signal of the servo driver, extract the temperature rise gradient sequence from the radiance image sequence, and perform a fast Fourier transform on the torque current signal to obtain the harmonic distortion sequence of the current in each sampling period. Combination module 32 is used to obtain a first vector sequence by performing time-series alignment and vectorization combination of the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence; Construction module 33 is used to construct a first distribution mapping relationship of temperature rise gradient and a second distribution mapping relationship of current harmonics based on the first vector sequence, and to determine the coupling evaluation factor between thermal stress and mechanical resistance by calculating the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship using the Archimedes Copula function. The determination module 34 is used to determine the target damping coefficient based on the coupling evaluation factor and according to the preset mapping relationship between the evaluation factor and the damping coefficient, and to perform a differential operation between the target damping coefficient and the preset reference damping value to obtain the impedance deviation. The calculation module 35 is used to calculate the power correction amount based on the impedance deviation and the preset power gain coefficient to adjust the output duty cycle, and to calculate the speed scaling factor based on the impedance deviation and the preset speed adjustment parameter to adjust the feed speed value, so as to obtain the target power parameter and the target speed parameter.

[0079] The intelligent optimization system for laser cutting process parameters in this application is used to implement the aforementioned intelligent optimization method for laser cutting process parameters. Therefore, the specific implementation of the intelligent optimization system for laser cutting process parameters can be found in the embodiment section of the intelligent optimization method for laser cutting process parameters above. The specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0080] like Figure 4 As shown, this application also provides an electronic device, including: a memory 41 for storing a computer program; and a processor 42 for executing the computer program to implement the steps of the intelligent optimization method for laser cutting process parameters described above.

[0081] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described intelligent optimization methods for laser cutting process parameters.

[0082] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0083] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the intelligent optimization method for laser cutting process parameters.

[0084] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0085] The above provides a detailed description of the intelligent optimization method system for laser cutting process parameters provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for intelligent optimization of laser cutting process parameters, characterized in that, include: The radiometric image sequence of the cut area and the torque current signal of the servo driver are acquired. The temperature rise gradient sequence is extracted from the radiometric image sequence. The torque current signal is subjected to fast Fourier transform to obtain the harmonic distortion sequence of the current in each sampling period. By performing time-series alignment and vectorization combination on the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence, a first vector sequence is obtained. Based on the first vector sequence, a first distribution mapping relationship of temperature rise gradient and a second distribution mapping relationship of current harmonics are constructed, and the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship is calculated by using the Archimedes Copula function to determine the coupling evaluation factor between thermal stress and mechanical resistance. Based on the coupling evaluation factor, the target damping coefficient is determined according to the preset mapping relationship between the evaluation factor and the damping coefficient. The target damping coefficient is then differentially calculated with the preset reference damping value to obtain the impedance deviation. The power correction is calculated based on the impedance deviation and the preset power gain coefficient to adjust the output duty cycle, and the speed scaling factor is calculated based on the impedance deviation and the preset speed adjustment parameter to adjust the feed speed value, thus obtaining the target power parameter and the target speed parameter.

2. The method according to claim 1, characterized in that, The first vector sequence is obtained by temporally aligning and vectorizing the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence, including: By calculating the difference between the first timestamp sequence of the radiance image sequence and the second timestamp sequence of the torque current signal, time point matching is performed to establish the correspondence between the temperature rise gradient value in the temperature rise gradient sequence and the harmonic distortion value in the harmonic distortion sequence. The first range is obtained by calculating the difference between the maximum and minimum values ​​of the temperature rise gradient in the temperature rise gradient sequence. The ratio of the difference between each temperature rise gradient value and the minimum value in the temperature rise gradient sequence to the first range is calculated to obtain the temperature rise gradient value after dimension unification. The second range is obtained by calculating the difference between the maximum and minimum values ​​of the harmonic distortion values ​​in the harmonic distortion sequence. The ratio of the difference between each harmonic distortion value and the minimum value in the harmonic distortion sequence to the second range is calculated to obtain the dimension-unified harmonic distortion value for each harmonic distortion value. Based on the correspondence, the dimensionally unified temperature rise gradient value and the corresponding dimensionally unified harmonic distortion value are combined into a two-dimensional vector, and all the two-dimensional vectors are arranged in chronological order according to the timestamps to form a first vector sequence.

3. The method according to claim 1, characterized in that, The construction of a first distribution mapping relationship for the temperature rise gradient and a second distribution mapping relationship for the current harmonics based on the first vector sequence includes: Determine the number of two-dimensional vectors in the first vector sequence; Extract the first and second components of each two-dimensional vector from the first vector sequence to obtain the temperature rise gradient value sequence and harmonic distortion value sequence. Arrange the values ​​in the temperature rise gradient value sequence in ascending order, and calculate the ratio of the frequency of each value to the number of values ​​after arrangement to obtain the first distribution mapping relationship; The values ​​in the harmonic distortion value sequence are arranged in ascending order, and the ratio of the frequency of each value to the total number of values ​​is calculated to obtain the second distribution mapping relationship.

4. The method according to claim 1, characterized in that, The step of determining the coupling evaluation factor between thermal stress and mechanical resistance by calculating the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship using the Archimedes Copula function includes: Substitute each value in the temperature rise gradient value sequence into the first distribution mapping relationship to obtain the first probability value corresponding to each temperature rise gradient value; Substitute each value in the harmonic distortion value sequence into the second distribution mapping relationship to obtain the second probability value corresponding to each harmonic distortion value; The first probability value and the corresponding second probability value are combined to form a probability pair, and the joint probability density value of all the probability pairs is calculated by the Archimedes Copula function. The relevant parameter in the Archimedes Copula function that maximizes the sum of all the joint probability density values ​​is determined as the coupling evaluation factor.

5. The method according to claim 1, characterized in that, Based on the coupling evaluation factor, the target damping coefficient is determined according to the preset mapping relationship between the evaluation factor and the damping coefficient. The target damping coefficient is then differentially calculated with a preset reference damping value to obtain the impedance deviation, including: The evaluation factor interval in which the coupled evaluation factor is located is found in the mapping relationship table between the evaluation factor and the damping coefficient, and the target damping coefficient corresponding to the evaluation factor interval is determined. The mapping relationship table is constructed in advance by cutting steel plates of various thicknesses. The difference between the target damping coefficient and the preset reference damping value is calculated to obtain the impedance deviation. The preset reference damping value is the target damping coefficient obtained during the cutting process under ideal processing conditions.

6. The method according to claim 1, characterized in that, The process of calculating a power correction based on the impedance deviation and a preset power gain coefficient to adjust the output duty cycle, and calculating a speed scaling factor based on the impedance deviation and a preset speed adjustment parameter to adjust the feed rate, to obtain target power parameters and target speed parameters, includes: The impedance deviation is multiplied by a preset power gain coefficient to obtain a power correction amount, and the current output duty cycle is added to the power correction amount to obtain the adjusted output duty cycle. The impedance deviation is multiplied by the preset speed adjustment parameter to obtain the speed scaling factor, and the current feed speed value is added to the speed scaling factor to obtain the adjusted feed speed value. Determine whether the adjusted output duty cycle is greater than a preset power upper limit or less than a preset power lower limit. If the adjusted output duty cycle is greater than the preset power upper limit, determine the preset power upper limit as the target power parameter. If the adjusted output duty cycle is less than the preset power lower limit, determine the preset power lower limit as the target power parameter. Otherwise, determine the adjusted output duty cycle as the target power parameter. Determine whether the adjusted feed rate value is greater than a preset upper speed limit or less than a preset lower speed limit. If the adjusted feed rate value is greater than the preset upper speed limit, determine the preset upper speed limit as the target speed parameter. If the adjusted feed rate value is less than the preset lower speed limit, determine the preset lower speed limit as the target speed parameter. Otherwise, determine the adjusted feed rate value as the target speed parameter.

7. The method according to claim 1, characterized in that, Before calculating the power correction amount based on the impedance deviation and a preset power gain coefficient to adjust the output duty cycle, the method further includes: Extract the set of pixels in the radiance image sequence whose luminance scores exceed a preset luminance threshold to obtain the bright areas; The area percentage is obtained by calculating the ratio of the number of pixels in all the highlighted areas to the total number of pixels in a single frame of the radiance image sequence. Calculate the geometric distance between the mean coordinates of the center points of each of the highlighted areas and the real-time physical coordinates of the cutting head to obtain the centroid offset distance, and calculate the spatter density by multiplying the area ratio by the centroid offset distance. The target adjustment factor corresponding to the splash density is determined by a preset mapping table between splash density and adjustment factor, and the power gain coefficient is adjusted according to the target adjustment factor.

8. A laser cutting process parameter intelligent optimization system, characterized in that, include: The acquisition module is used to acquire the radiance image sequence of the cutting area and the torque current signal of the servo driver, extract the temperature rise gradient sequence from the radiance image sequence, and perform a fast Fourier transform on the torque current signal to obtain the harmonic distortion sequence of the current in each sampling period. The combination module is used to obtain a first vector sequence by performing time-series alignment and vectorization combination of the dimensionally unified temperature rise gradient sequence and the harmonic distortion sequence; The construction module is used to construct a first distribution mapping relationship of temperature rise gradient and a second distribution mapping relationship of current harmonics based on the first vector sequence, and to determine the coupling evaluation factor between thermal stress and mechanical resistance by calculating the joint probability density value between the first distribution mapping relationship and the second distribution mapping relationship using the Archimedes Copula function. The determination module is used to determine the target damping coefficient based on the coupling evaluation factor and according to the preset mapping relationship between the evaluation factor and the damping coefficient, and to perform a differential operation between the target damping coefficient and the preset reference damping value to obtain the impedance deviation. The calculation module is used to calculate the power correction amount based on the impedance deviation and the preset power gain coefficient to adjust the output duty cycle, and to calculate the speed scaling factor based on the impedance deviation and the preset speed adjustment parameter to adjust the feed speed value, so as to obtain the target power parameter and the target speed parameter.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the intelligent optimization method for laser cutting process parameters as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the intelligent optimization method for laser cutting process parameters as described in any one of claims 1 to 7.

Images