A cutting positioning method and device for laser cutting metal based on big data

Abstract

The application discloses a kind of cutting positioning method and device based on laser cutting metal of big data, it is related to laser cutting technical field.This kind of cutting positioning method and device based on laser cutting metal of big data, including S1, preprocessing to thermal working condition data, process path space characteristic data and historical working condition data;S2, based on time series thermal fluctuation analysis, the initial calibration of the space offset of the coupling of thermal expansion, thermal fluctuation and disturbance before cutting multiple points, and enter detailed compensation process;S3, by thermal coupling compensation theory analysis to path node space fine tuning and high-precision trajectory correction;S4, based on time series fluctuation and mutation criterion, the time series response fluctuation and abnormal risk of compensation process are monitored and graded intervention in real time.Solve the complex coupling drift generated by multi-source high-temperature instantaneous thermal expansion and deformation of laser head and workpiece under extreme cutting conditions, existing positioning method is difficult to compensate accurately in real time, seriously affect high-precision cutting stability problem.

Description

Technical Field

[0001] This invention relates to the field of laser cutting technology, specifically to a cutting positioning method and apparatus for laser cutting of metal based on big data. Background Technology

[0002] With the rapid development of intelligent manufacturing and high-end equipment processing technologies, laser cutting has become an important means of forming metal materials and processing complex structural parts due to its advantages such as high precision, high efficiency, and flexible processing. Modern laser cutting systems are widely used in various industries such as aerospace, automotive manufacturing, rail transportation, engineering machinery, and electronics manufacturing, and the requirements for cutting quality and positioning accuracy are increasing. To achieve high-precision cutting of complex curves and irregular contours, laser cutting equipment is usually equipped with multi-axis CNC platforms, high-performance lasers, and various types of sensors, which can acquire process parameters such as temperature, position, speed, and acceleration in real time. These technologies have promoted the continuous improvement of the intelligence and automation level of laser cutting processes, and facilitated the widespread application of high-precision and high-efficiency metal processing.

[0003] For example, the invention patent with publication number CN119609395A discloses a laser cutting device for metal pipes with positioning and cutting functions, including a support frame, a conveyor plate, a mounting base, a mounting plate, a fixture base, and a laser cutting machine. The mounting plate is equipped with a second servo motor and a third servo motor, which drive the first rotating rod and the second gear respectively to achieve multi-axis linkage and fine adjustment. The fixture base is equipped with components such as a bidirectional screw, a clamping block, an elastic telescopic head, and a cylinder. Through gear meshing and screw transmission, combined with the fixture limiting slide groove and the L-shaped fixing plate, it can achieve clamping and positioning of metal pipes and efficient laser cutting, which is suitable for automated precision processing of metal pipes of different sizes.

[0004] For example, utility model patent CN222359467U discloses a metal processing cutting positioning structure, including a worktable, support legs, and an adjustment assembly. The adjustment assembly includes a drive shaft, a pinion, a ring gear, a placement disc, a drive pulley, a driven pulley, a reciprocating screw, a lifting frame, an adjustment slide rail, an adjustment cylinder, and a laser cutter. This structure uses the meshing of the pinion and ring gear to rotate the placement disc. The drive and driven pulleys drive the reciprocating screw and lifting frame via belt transmission to achieve lifting and adjustment. The adjustment slide rail is equipped with an adjustment cylinder, which drives the laser cutter to achieve cutting positioning adjustment. This structural design realizes multi-dimensional positioning of the workpiece and precise movement of the laser cutter, improving the cutting positioning accuracy and efficiency of metal processing.

[0005] While existing metal processing cutting positioning and laser cutting devices can achieve multi-dimensional workpiece positioning and a certain degree of automated cutting operations, they still suffer from problems in practical applications, such as insufficient precision in responding to spatial offsets and deformations under complex working conditions, limited dynamic compensation capabilities, and insufficient depth of multi-source process data fusion. These limitations make it difficult to consistently guarantee cutting accuracy and stability under extreme environments such as high temperatures, high speeds, and dynamic disturbances. Furthermore, existing technologies have relatively simple response strategies in areas such as risk warning, process anomaly monitoring, and intelligent intervention, failing to meet the higher requirements of high-precision, intelligent, and closed-loop control throughout the entire process in the high-end equipment manufacturing field.

[0006] Therefore, in order to address the above problems, there is an urgent need for a cutting positioning method and device for laser cutting of metal based on big data. Summary of the Invention

[0007] Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a cutting positioning method and device for laser cutting of metal based on big data. It solves the problem that existing positioning methods cannot accurately compensate for the complex coupling drift caused by the instantaneous thermal expansion and deformation of the laser head and workpiece under extreme cutting conditions, which seriously affects the stability of high-precision cutting.

[0008] Technical solution To achieve the above objectives, the present invention is implemented through the following technical solution: a cutting positioning method for laser cutting of metal based on big data, comprising: S1, collecting structural thermal condition data and process path spatial feature data, acquiring historical disturbance condition data, and preprocessing the thermal condition data, process path spatial feature data, and historical condition data; S2, based on time-series thermal fluctuation analysis, predicting and initially calibrating the trend of multi-point thermal expansion, thermal fluctuation, and disturbance coupled spatial offset before cutting, and entering a detailed compensation process; S3, for dynamic thermal drift and disturbance during the cutting process, fine-tuning and high-precision trajectory correction of the path node space through thermal coupling compensation theory analysis; S4, based on time-series fluctuation and mutation criteria, real-time monitoring and graded intervention of time-series response fluctuations and abnormal risks in the compensation process.

[0009] Furthermore, the specific process for collecting structural thermodynamic condition data and process path spatial characteristic data to obtain historical disturbance condition data is as follows: Structural thermodynamic condition data is collected, including: the number of compensation structural components, temperature-length variation data, geometric dimension data, real-time temperature data, initial temperature variables, surface temperature sequence data, current time, triaxial acceleration data, and angle variation data; process path spatial characteristic data is collected, including: main path spatial coordinate point data, compensation vector direction data, length variation data, current spatial coordinates, monitoring start time, and monitoring end time; historical disturbance condition data is obtained and a historical disturbance condition database is established, including: historical triaxial acceleration data and historical angle variation data.

[0010] Furthermore, the specific preprocessing steps for thermal condition data, process path spatial feature data, and historical condition data are as follows: Outlier removal and multi-source time alignment algorithms are used to complete missing values ​​and synchronize data in the structural thermal condition data, ensuring that all types of thermal parameters correspond at the same time. Spatial coordinate normalization and node spacing balancing algorithms are used to resample and spatially partition the process path spatial feature data, extracting key node features of the main path. Window mean filtering and disturbance amplitude normalization algorithms are used to denoise and smooth the disturbance historical condition data across multiple channels, enhancing the effective expression of disturbance change trends. Distribution standardization and linear normalization algorithms are used to standardize and normalize the structural thermal condition data, process path spatial feature data, and disturbance historical condition data.

[0011] Furthermore, the specific process based on time-series thermal fluctuation analysis is as follows: Acquire the number of compensation structural components, temperature length variation data, geometric dimension data, real-time temperature data, initial temperature variable, surface temperature sequence data, current moment, triaxial acceleration data, and angle variation data; obtain the thermal expansion coefficient from the temperature length variation data using a least-squares linear regression algorithm; obtain the effective feature length from the geometric dimension data using a three-dimensional distance calculation algorithm; obtain the real-time temperature variable from the real-time temperature data using a temperature mean algorithm; obtain the thermal fluctuation amplitude factor from the surface temperature sequence data using range calculation; analyze the dominant frequency from the surface temperature sequence data using a fast Fourier transform, and calculate the thermal fluctuation angular frequency by multiplying 2π by the dominant frequency; and obtain the triaxial acceleration data... The number of disturbance terms is obtained by Euclidean norm fusion and variance threshold detection algorithms for triaxial acceleration data and angle change data; the spatial offset value of process disturbance is obtained by Euclidean norm fusion, vector projection and absolute value operation for triaxial acceleration data and angle change data; the thermal expansion offset term of the structural component is obtained by calculating the product of the coefficient of thermal expansion, effective feature length and the difference between surface temperature and static temperature; the thermal fluctuation offset term of the structural component is obtained by calculating the product of thermal fluctuation amplitude factor, thermal fluctuation angular frequency and sine value at the current time; the thermal expansion offset term and the thermal fluctuation offset term of the structural component are added together to obtain the predicted spatial offset value of the structural component; the predicted spatial offset values ​​of all structural components are accumulated, and the spatial offset values ​​of all process disturbances are accumulated and then added together to obtain the predicted spatial offset value.

[0012] Furthermore, the specific process for predicting and initially calibrating the trend of multi-point thermal expansion, thermal fluctuation, and disturbance coupled spatial offset before cutting, and then entering the detailed compensation process, is as follows: Real-time comparison of the spatial offset prediction value and the spatial offset prediction threshold: When the spatial offset prediction value is less than the spatial offset prediction threshold, it is determined to be a normal initial offset state; Through three-dimensional geometric projection and vector superposition algorithm, the thermally sensitive spatial offset prediction value is projected onto the three-dimensional coordinate axis, and converted into a preliminary three-dimensional coordinate compensation to obtain the thermally sensitive spatial offset prediction three-dimensional value, which is applied to CNC zero point and path offset compensation, and enters the detailed compensation process; When the spatial offset prediction value is greater than or equal to the spatial offset prediction threshold, it is determined to be an initial abnormal offset state; Extend the preheating time, repeat the zero return and edge finding, and execute the workpiece static waiting process until the spatial offset prediction value is less than the spatial offset prediction value threshold.

[0013] Furthermore, regarding the dynamic thermal drift and disturbance during the cutting process, the specific process of analysis using thermo-mechanical coupling compensation theory is as follows: Obtain the predicted three-dimensional value of thermal spatial offset, main path spatial coordinate point data, compensation vector direction data, real-time temperature data, length change data, current spatial coordinates, triaxial acceleration data, angle change data, historical triaxial acceleration data, and historical angle change data; calculate the angle between the compensation vector and the main cutting path using vector angle calculation on the main path spatial coordinate point data and compensation vector direction data; obtain the thermal expansion coefficient of the workpiece's main material using the least squares linear regression algorithm on the real-time temperature data and length change data; obtain the remaining feature length using the three-dimensional Euclidean distance calculation algorithm on the main path spatial coordinate point data and current spatial coordinates; obtain the disturbance response amplification factor using the extremum algorithm and sensitivity calibration method on the triaxial acceleration data and angle change data; and obtain the triaxial acceleration data and angle change data using the weighted Euclidean norm fusion algorithm and window accumulation operation. The overall spatial drift increment is calculated as follows: Historical spatial drift increments are obtained by weighted Euclidean norm fusion and window accumulation operations on historical triaxial acceleration data and historical angle change data. The disturbance smoothing coefficient is obtained by a sliding window averaging algorithm on the historical spatial drift increments. The spatial compensation term is obtained by multiplying the predicted three-dimensional value of the thermally sensitive spatial offset and the cosine of the angle between the compensation vector and the main cutting path. The thermal expansion offset term is obtained by multiplying the workpiece's main material thermal expansion coefficient, remaining feature length, and the difference between the real-time temperature variable and the initial temperature variable. The path offset term is added to the thermal expansion offset term to obtain the path offset term. The disturbance drift term is obtained by multiplying the disturbance response amplification factor and the overall spatial drift increment. The absolute value of the path offset term and the disturbance drift term is added and used as the numerator. The sum of the safety drift amount, the absolute value of the difference between the real-time temperature variable and the initial temperature variable, and the disturbance smoothing coefficient is used as the denominator. The numerator is divided by the denominator, and the results for all structural components are summed to obtain the thermal drift response value.

[0014] Furthermore, the specific process of fine-tuning and high-precision trajectory correction in the path node space is as follows: Real-time comparison of thermal drift response values ​​and thermal drift response thresholds, including primary and secondary response thresholds, is performed. A detailed compensation process is then executed: when the thermal drift response value is greater than or equal to the primary response threshold, the execution of subsequent uncut paths is paused, the current partition undergoes regional inspection, compensation parameters are verified, and an abnormal area report is output; when the thermal drift response value is greater than or equal to the secondary response threshold but less than the primary response threshold, dynamic analysis is used to detect whether the fluctuation in the thermal drift response value is due to temperature rise or disturbance. If the issue stems from temperature rise and disturbance, the feed rate, laser pulse width, and energy density are reduced to mitigate local drift. A path node spatial compensation algorithm based on vector projection is used to calculate the compensation amount for each component when the thermal drift response value exceeds the risk threshold. This compensation amount is then superimposed onto the original path node coordinates along the path direction, dynamically fine-tuning the cutting trajectory and generating corresponding corrective CNC commands in real time. These commands are then synchronously sent to the motion controller, ensuring the path is updated online in real time. When the thermal drift response value is less than the secondary response threshold, the cutting path continues to advance according to the current state, with continuous monitoring of the thermal drift response value along the remaining path.

[0015] Further, the specific process based on the time series fluctuation and mutation criteria is as follows: Obtain the current time, the compensated thermal drift response value sequence data, the monitoring start time, and the monitoring end time; obtain the response sliding window mean using the sliding window mean algorithm on the compensated thermal drift response value sequence data; obtain the response sliding window standard deviation using the sliding window standard deviation algorithm on the compensated thermal drift response value sequence data; obtain the response value time derivative using the numerical differentiation algorithm on the compensated thermal drift response value sequence data; obtain the response sliding window complexity using the sample entropy algorithm on the compensated thermal drift response value sequence data; calculate the absolute value of the difference between the compensated thermal drift response value and the response sliding window mean, multiply it by the response sliding window standard deviation to obtain the response deviation fluctuation term; calculate the absolute value of the derivative of the compensated thermal drift response value with respect to time and add one to obtain the response change rate term; divide the response deviation fluctuation term by the response change rate term to obtain the response fluctuation value; calculate the response sliding window complexity and add one to obtain the response complexity adjustment term; multiply the response fluctuation value by the response complexity adjustment term, and take the maximum value within the monitoring period to obtain the compensated risk value.

[0016] Furthermore, the specific process for real-time monitoring and graded intervention of temporal response fluctuations and abnormal risks in the compensation process is as follows: Real-time comparison of compensation risk values ​​and compensation risk thresholds, including primary risk thresholds and secondary risk thresholds: When the compensation risk value is less than the secondary risk threshold, it is determined to be a process health zone, a process health database is established, and structural thermal condition data and process path spatial characteristic data are dynamically updated and archived into the process health database without special action; When the compensation risk value is greater than or equal to the secondary risk threshold but less than the primary risk threshold, it is determined to be a process concern zone, the sampling period is increased, and the nodes of the concern zone are digitally marked; When the compensation risk value is greater than or equal to the primary risk threshold, it is determined to be a process early warning zone, and physical interventions such as spray cooling and vibration isolation are implemented for the zone. If the compensation risk value is still greater than or equal to the primary risk threshold after physical intervention, the process is temporarily suspended, multi-dimensional abnormal parameters are pushed to generate a process abnormality report, and maintenance personnel are advised to conduct inspections.

[0017] The second aspect of this invention provides a cutting positioning device for laser cutting metal based on big data, including an acquisition and preprocessing module for acquiring structural thermal condition data and process path spatial feature data, obtaining historical disturbance condition data, and preprocessing the thermal condition data, process path spatial feature data, and historical condition data; an initial thermal drift spatial compensation and verification module for predicting and initially calibrating the trend of multi-point thermal expansion, thermal fluctuation, and disturbance coupled spatial offset before cutting based on time-series thermal fluctuation analysis, and entering a detailed compensation process; a thermal drift compensation strategy module for fine-tuning and high-precision trajectory correction of path node space through thermal coupling compensation theory analysis for dynamic thermal drift and disturbance during the cutting process; and a compensation risk monitoring and intervention module for real-time monitoring and graded intervention of time-series response fluctuations and abnormal risks in the compensation process based on time-series fluctuation and mutation criteria.

[0018] Beneficial effects The present invention has the following beneficial effects: (1) This invention integrates multi-source structural thermodynamic conditions data and process path multi-dimensional features to achieve real-time high-precision identification and adaptive correction of multiple physical field factors such as thermal expansion and structural deformation during laser cutting, thereby achieving accurate compensation effect for spatial dynamic displacement under complex thermo-mechanical coupling conditions, effectively solving the technical bottleneck of high-temperature transient deformation that is difficult to correct in time in the prior art.

[0019] (2) This invention, relying on multi-point heterogeneous sensor network and historical disturbance big data analysis, constructs a dynamic risk monitoring and intelligent early warning system that runs through the entire process, thereby realizing rapid perception and zoned adaptive intervention of cutting path anomalies and process disturbances, effectively solving the problems of delayed risk identification and insensitive response in the prior art.

[0020] (3) In this invention, partition space correction and trajectory dynamic optimization are implemented on the main cutting path and key nodes, thereby achieving the adaptive dynamic control effect of high-precision cutting trajectory, effectively solving the problems of coarse path offset correction granularity and insufficient control precision in the prior art.

[0021] (4) This invention realizes multi-module collaborative closed-loop scheduling such as cutting process space compensation, risk monitoring and path correction, thereby achieving a system-level improvement effect of high precision and high reliability intelligent cutting process, effectively solving the core problems of the disconnect between compensation and control links and limited response capability in the prior art.

[0022] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0023] Figure 1 This is a flowchart of a laser cutting positioning method for metal based on big data according to the present invention; Figure 2 This is a structural diagram of a laser metal cutting positioning device based on big data according to the present invention. Figure 3 This is a distribution diagram of the thermal drift response components of the structural components of the present invention. Figure 4 This is a flowchart of the dynamic risk control process for risk value partitioning in this invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Please see Figures 1-4 This invention provides a technical solution: a cutting positioning method for laser cutting metal based on big data, comprising: S1, collecting structural thermal condition data and process path spatial feature data, acquiring historical disturbance condition data, and preprocessing the thermal condition data, process path spatial feature data, and historical condition data; S2, based on time-series thermal fluctuation analysis, predicting and initially calibrating the trend of multi-point thermal expansion, thermal fluctuation, and disturbance coupled spatial offset before cutting, and entering a detailed compensation process; S3, for dynamic thermal drift and disturbance during the cutting process, fine-tuning and high-precision trajectory correction of the path node space through thermal coupling compensation theory analysis; S4, based on time-series fluctuation and mutation criteria, real-time monitoring and graded intervention of time-series response fluctuations and abnormal risks in the compensation process.

[0026] Specifically, the process of collecting structural thermal condition data and process path spatial characteristic data to obtain historical disturbance condition data is as follows: Structural thermal condition data is collected, including: the number of compensation structural components, temperature-length variation data, geometric dimension data, real-time temperature data, initial temperature variables, surface temperature sequence data, current time, triaxial acceleration data, and angle variation data. The number of compensation structural components refers to the number of key structural units requiring focused monitoring and dynamic compensation determined through CAD partitioning and process analysis. Heat-affected and easily deformable parts of the laser-cut workpiece are selected, and corresponding sensors are distributed accordingly. Temperature-length variation data is collected from sensors installed at key nodes of the workpiece with a precision of micrometers. The measurement is achieved by linking a grating ruler with a sampling rate of ≥50Hz and a K-type thermocouple with an accuracy of ±0.5℃; geometric dimension data is acquired in one go by a 3D laser scanner with an accuracy of 0.01mm before workpiece processing; real-time temperature data and initial temperature variables are acquired by a distributed thermocouple array fixed on the workpiece surface and the heat-affected zone of the cut, with high-frequency sampling for real-time temperature and the initial temperature as the starting baseline; surface temperature sequence data is continuously obtained by an infrared thermal imager with a frame rate of ≥25Hz above the workstation; the current time is recorded by the clock module of the main control system; three-axis acceleration data and angle change data are acquired in real time by IMUs installed on the laser head and clamping device to reflect dynamic disturbances.

[0027] The process path spatial feature data is collected, including: main path spatial coordinate point data, compensation vector direction data, length change data, current spatial coordinates, monitoring start time, and monitoring end time. The main path spatial coordinate point data is output by the CNC system and corrected by the laser rangefinder at the end of the laser head. The compensation vector direction data and length change data are generated by the linkage between path planning and sensors. The current spatial coordinates are output in real time through the positioning sensors on the laser head and fixture. The monitoring start time and monitoring end time are recorded and archived by the control system.

[0028] Acquire historical disturbance data and establish a historical disturbance data database. The historical disturbance data includes historical triaxial acceleration data and historical angle change data. Both historical triaxial acceleration data and historical angle change data are acquired over a long period by the IMU module and archived into the historical data database, which supports high-frequency sampling and querying by process stage and partition.

[0029] In this implementation plan, the integration of multi-source high-precision sensors and an intelligent control system enables comprehensive collection and standardized management of key data throughout the laser cutting process. By collecting and archiving structural thermal condition data, process path spatial characteristic data, and historical disturbance data, it is possible not only to obtain real-time and accurate multi-dimensional information on workpiece heating, deformation, motion, and path, but also to provide a solid data foundation for subsequent spatial compensation, dynamic control, and risk assessment. This step effectively improves the completeness, accuracy, and traceability of process data, laying a solid foundation for achieving high-precision and intelligent laser cutting process control.

[0030] Specifically, the preprocessing steps for thermal condition data, process path spatial feature data, and historical condition data are as follows: Outlier removal and multi-source time alignment algorithms are used to complete missing values ​​and synchronize data in the structural thermal condition data. High-precision clocks and sensor self-checking functions are utilized to improve data accuracy, ensuring that all thermal parameters correspond at the same time. Spatial coordinate normalization and node spacing balancing algorithms are used to resample and spatially partition the process path spatial feature data, optimizing node distribution density for subsequent fine-grained trajectory compensation and extracting key node features of the main path. Window mean filtering and disturbance amplitude normalization algorithms are used to denoise and smooth the historical disturbance condition data through multiple channels, improving the smoothness and discriminative stability of the disturbance data and strengthening the effective expression of disturbance change trends. Distribution standardization and linear normalization algorithms are used to standardize and normalize the structural thermal condition data, process path spatial feature data, and historical disturbance condition data, ensuring that all types of data have unified dimensions and comparability in subsequent analysis.

[0031] In this implementation scheme, data cleaning, alignment, and normalization effectively improve the completeness, accuracy, and temporal consistency of multi-source process data. By removing outliers, filling in missing information, and utilizing multi-source time alignment and spatial resampling algorithms, efficient synchronization and standardized representation of data such as structural thermal conditions, path characteristics, and historical disturbances are achieved. After processing with unified dimensions and scales, the data can better reflect the characteristics of key nodes and the trends of disturbance changes, providing a reliable data foundation for subsequent spatial compensation, trajectory optimization, and risk analysis. This significantly enhances the scientific rigor of data fusion, feature extraction, and intelligent analysis, and strongly supports high-precision decision-making and dynamic control capabilities.

[0032] Specifically, the process based on time-series thermal fluctuation analysis is as follows: Acquire the number of compensation structural components, temperature-length variation data, geometric dimension data, real-time temperature data, initial temperature variables, surface temperature sequence data, current time, triaxial acceleration data, and angle variation data to lay the initial data foundation for subsequent thermo-mechanical coupling analysis and dynamic compensation; obtain the thermal expansion coefficient from the temperature-length variation data using a least-squares linear regression algorithm. This algorithm has strong noise resistance and can effectively fit multiple sets of linear relationships between temperature and length variations; obtain the effective feature length from the geometric dimension data using a three-dimensional distance calculation algorithm, and quickly extract key dimensions of structural components using a spatial distance matrix to improve efficiency; obtain the real-time temperature variable from the real-time temperature data using a temperature averaging algorithm, and further reduce the impact of measurement fluctuations on the analysis results using a sliding window averaging method; and obtain the surface temperature sequence data. Temperature sequence data is used to calculate the thermal fluctuation amplitude factor through range calculation. The range algorithm can keenly capture the maximum dynamic range of temperature changes and is used to characterize the intensity of thermal disturbances. For surface temperature sequence data, the dominant frequency is analyzed through Fast Fourier Transform (FFT), and the thermal fluctuation angular frequency is obtained by multiplying 2π by the dominant frequency. The FFT algorithm can effectively extract the dominant frequency of complex thermal field changes and support periodic fluctuation modeling. For triaxial acceleration data and angle change data, the number of disturbance terms is obtained through Euclidean norm fusion and variance threshold detection algorithms. Euclidean norm fusion improves the ability to express multidimensional signals, and variance threshold detection enhances the accuracy of identifying abnormal disturbances. For triaxial acceleration data and angle change data, the spatial offset value of process disturbance is obtained through Euclidean norm fusion, vector projection, and absolute value operation. The comprehensive spatial vector algorithm can realize the normalization and spatial quantitative analysis of multi-source disturbances.

[0033] The thermal expansion offset term of the structural component is obtained by calculating the product of the coefficient of thermal expansion, effective feature length, and the difference between surface temperature and static temperature. Normalization of multiple parameters enables quantitative decomposition of the thermal expansion effect. The thermal fluctuation offset term of the structural component is obtained by calculating the product of the thermal fluctuation amplitude factor, the thermal fluctuation angular frequency, and the sine value at the current moment. Time-series sine modeling reflects the periodic characteristics of thermal fluctuations. The thermal expansion offset term and the thermal fluctuation offset term of the structural component are added to obtain the predicted spatial offset value. The superposition of multiple factors improves the overall accuracy of the spatial offset prediction. The predicted spatial offset values ​​of all structural components are summed, and the spatial offset values ​​of all process disturbances are summed again to obtain the predicted spatial offset value. Finally, a quantitative output of the comprehensive spatial offset under global thermal-mechanical-disturbance coordination is achieved. The specific calculation formula is as follows: ; In the formula, This represents the predicted spatial offset, which is the global spatial offset output after combining multi-source thermal conditions and disturbances. This indicates the number of compensating structural components, which is the total number of structural units involved in the compensation calculation. It represents the coefficient of thermal expansion, reflecting the expansion characteristics of a material after it is heated; Indicates the effective feature length, which is the initial length parameter of the structural component; This represents the real-time temperature variable, indicating the actual temperature of the structural component at the current moment. This represents the initial temperature variable, indicating the reference temperature of the structural component at the initial moment; This represents the thermal fluctuation amplitude factor, used to measure the magnitude of temperature changes; It represents the angular frequency of thermal fluctuations and is used to characterize the periodicity of temperature changes; This indicates the current moment, the point in time at which the judgment and calculation are performed; This indicates the number of perturbation terms, representing the number of perturbations that need to be considered during the cutting process; This represents the spatial offset value of the process disturbance, used to reflect the degree of spatial offset of each disturbance.

[0034] In this implementation scheme, quantitative analysis and collaborative modeling of various thermal and disturbance factors in the laser cutting process are achieved through data fusion and hierarchical feature extraction. A multi-algorithm collaborative processing approach, employing linear regression, spatial distance calculation, moving average, range, and fast Fourier transform, not only accurately decomposes and expresses various offset components of the structural components under thermal expansion, thermal fluctuations, and dynamic disturbances, but also realizes spatial offset prediction under the superposition of multiple factors. The final output global spatial offset comprehensively reflects the coupling effects of complex physical fields in the cutting process, providing a scientific and reliable criterion and data foundation for subsequent high-precision compensation, dynamic path correction, and risk warning, significantly improving the intelligent and refined control level of the laser cutting process.

[0035] Specifically, the process of predicting and initially calibrating the trend of spatial displacement coupled with multi-point thermal expansion, thermal fluctuations, and disturbances before cutting, and then proceeding to a detailed compensation process, involves: real-time comparison of the predicted spatial displacement value and the predicted spatial displacement threshold to ensure that the compensation decision can be dynamically adjusted according to the current actual working conditions. When the predicted spatial offset is less than the predicted spatial offset threshold, it is determined to be in a normal initial offset state. The preliminary compensation process will be initiated based on the real-time working conditions to ensure data closure. Through three-dimensional geometric projection and vector superposition algorithms, the predicted thermal spatial offset is projected onto the three-dimensional coordinate axis in parts, and converted into the preliminary three-dimensional coordinate compensation to obtain the predicted three-dimensional value of thermal spatial offset. The compensation value will be directly synchronized to the CNC zero point correction and path offset adjustment to ensure the dynamic accuracy of the spatial reference and enter the detailed compensation process.

[0036] When the predicted spatial offset is greater than or equal to the predicted spatial offset threshold, it is determined to be an initial abnormal offset state. The preheating time is extended, and the preheating time is dynamically adjusted according to the extent of the spatial offset prediction exceeding the limit, ranging from 5 to 20 minutes, to ensure the thermal stability of the structure. The zero-finding process is repeated 2 to 3 times, and each time the zero point is calibrated by high-precision position detection. The workpiece is then placed in a static waiting process for no less than 10 minutes until the predicted spatial offset is less than the predicted spatial offset threshold.

[0037] In this implementation scheme, by predicting and calibrating the trends of multi-point thermal expansion, thermal fluctuations, and spatial displacement caused by disturbances before cutting, accurate spatial position identification and intelligent compensation of the workpiece in its initial state are achieved. By comparing the predicted spatial displacement value with the threshold in real time, fine-grained operations such as preliminary compensation, 3D spatial projection adjustment, CNC zero-point correction, and path fine-tuning are triggered according to different working conditions to ensure that the cutting reference point and the compensation amount always maintain dynamic consistency. When an initial abnormal displacement is detected, the preheating time is extended, the workpiece is controlled to return to zero and perform high-precision calibration multiple times, and a static waiting strategy is used to continuously monitor until the spatial displacement recovers to within the set threshold. Through this series of dynamic identification and multi-stage compensation processes, the accuracy of spatial compensation under thermal-mechanical coupling and disturbance conditions is significantly improved, and the adaptability to complex start-up conditions and environmental fluctuations is enhanced, laying a solid foundation for high-precision start-up and subsequent stable operation of the laser cutting process.

[0038] Specifically, the analysis process using thermo-coupling compensation theory to address dynamic thermal drift and disturbance during the cutting process involves: acquiring three-dimensional values ​​of the predicted thermal spatial offset, main path spatial coordinate point data, compensation vector direction data, real-time temperature data, length change data, current spatial coordinates, triaxial acceleration data, angle change data, historical triaxial acceleration data, and historical angle change data. This provides comprehensive raw data support for dynamic compensation and multi-parameter response modeling, ensuring high data synchronization across all stages. The angle between the compensation vector and the main cutting path is calculated using the vector angle calculation from the main path spatial coordinate point data and the compensation vector direction data. This angle quantitatively describes the spatial relationship between the compensation direction and the actual cutting path, providing a key criterion for compensation accuracy. Finally, the workpiece's main material thermal expansion coefficient is obtained from the real-time temperature data and length change data using a least-squares linear regression algorithm, enhancing the dynamic identification capability of material characteristic parameters and enabling the compensation parameters to adapt to real-time conditions. Under actual working conditions, the remaining feature length is obtained by calculating the spatial coordinates of the main path and the current spatial coordinates using a three-dimensional Euclidean distance algorithm, reflecting the spatial distance changes of the path to be cut in real time and assisting in the dynamic adjustment of the compensation range. For triaxial acceleration data and angle change data, the disturbance response amplification factor is obtained through an extreme value algorithm and sensitivity calibration method, significantly improving the rapid response capability to sudden disturbances and vibration changes. For triaxial acceleration data and angle change data, the overall spatial drift increment is obtained through a weighted Euclidean norm fusion algorithm and window accumulation operation, taking into account both disturbance intensity and duration, and achieving a cumulative expression of the drift trend. For historical triaxial acceleration data and historical angle change data, the historical spatial drift increment is obtained through a weighted Euclidean norm fusion algorithm and window accumulation operation. The disturbance smoothing coefficient is obtained from the historical spatial drift increment through a sliding window averaging algorithm, enabling the compensation algorithm to fully utilize historical disturbance trends and improve the smoothness and stability of the overall compensation.

[0039] The spatial compensation term is obtained by multiplying the predicted three-dimensional value of the thermal spatial offset and the cosine of the angle between the compensation vector and the main cutting path. This directly incorporates the spatial offset directionality into the compensation calculation, effectively preventing misjudgment of the offset. The thermal expansion offset term is obtained by multiplying the workpiece's main material thermal expansion coefficient, remaining feature length, and the difference between the real-time temperature variable and the initial temperature variable. This comprehensively considers thermal expansion characteristics and changes in process path length, improving the accuracy of thermal compensation. The spatial compensation term and the thermal expansion offset term are added together to obtain the path offset term, which fully integrates the influence of directionality and thermal deformation. The disturbance response amplification factor and the overall spatial offset are then calculated. The product of drift increments yields the disturbance drift term, further amplifying the weight of strong disturbances on overall compensation and improving risk response capabilities. The path offset term is added to the disturbance drift term, and the absolute value is used as the numerator to ensure the consistency of offset compensation direction and enhance the robustness of the compensation criterion. The sum of the safe drift amount, the absolute value of the difference between the real-time temperature variable and the initial temperature variable, and the disturbance smoothing coefficient is calculated and used as the denominator, fully incorporating multiple factors such as safety margin, process temperature difference, and disturbance smoothing to enhance compensation stability and safety. The numerator is divided by the denominator, and the results for all structural components are summed to obtain the thermal drift response value. The specific calculation formula is as follows: ; In the formula, It represents the thermal drift response value, which comprehensively reflects the current thermal coupling and disturbance compensation effect of the system, and is an important indicator for dynamic adjustment and risk judgment; This represents the three-dimensional value predicted by thermal spatial migration, reflecting the overall dynamic changes in space; It represents the angle between the compensation vector and the main cutting path, and quantitatively describes the spatial relationship between the direction of compensation and the main cutting process path; It represents the coefficient of thermal expansion of the main material of the workpiece, and characterizes the volume and length expansion characteristics of the material under the action of temperature change; This represents the remaining feature length, which is the path length to be processed during the cutting process; This represents the real-time temperature variable, indicating the actual temperature of the structural component at the current moment. This represents the initial temperature variable, indicating the reference temperature of the structural component at the initial moment; This represents the disturbance response amplification factor, which weights and amplifies the responses to sudden disturbances and unsteady motion, increasing the weight of the disturbance signal on the compensation results. It represents the overall spatial drift increment and is used to describe the cumulative spatial drift under each disturbance source; It represents the perturbation smoothing coefficient, reflecting the smoothing and trend characteristics of historical perturbation data; The safe drift amount is a compensation constant obtained by using the denominator minimum robustness constraint algorithm on the absolute value of the difference between the initial temperature variable and the real-time temperature variable. The value ranges from 0.005 to 0.01.

[0040] In this embodiment, Table 1 is a data table of thermal drift influence of structural components. It records in detail the thermal drift response values ​​of eight structural components under the core components of spatial compensation, thermal expansion offset, and disturbance drift, which is used to quantify the multidimensional response performance of different structural components in the laser cutting positioning compensation process. Among them, the spatial compensation term of structural component 1 is 0.14, the thermal expansion offset term is 0.06, the disturbance drift term is 0.08, and the thermal drift response value component is 0.28; the spatial compensation term of structural component 3 is 0.18, the thermal expansion offset term is 0.08, the disturbance drift term is 0.12, and the thermal drift response value component reaches 0.38, which is the highest in the group, indicating that its comprehensive drift effect is the most significant; the three components of structural component 4 are 0.07, 0.06, and 0.05, respectively, and the thermal drift response value component is only 0.14, which is the lowest in the group, reflecting that its compensation pressure is the smallest. The denominator parameter not involved in the formula is uniformly fixed at 1.0 to eliminate the influence of other variables on the response value and ensure the horizontal comparability between the components.

[0041] Table 1 Data on the Influence of Thermal Drift on Structural Components like Figure 3 The diagram shows the distribution of thermal drift response components of structural components, illustrating the differences in their individual contributions to spatial compensation, thermal expansion offset, and disturbance drift, as well as the differences in their total thermal drift response. The diagram reveals that structural components 3 and 5 have the highest thermal drift response components, with spatial compensation and disturbance drift contributing particularly significantly to their overall drift. In contrast, structural components 2, 4, and 8 have lower response components, indicating a more balanced impact from each component. Overall, the comparison of the individual responses among structural components not only reveals the risk level of each component in the actual compensation process but also provides a quantitative basis for subsequent optimization of compensation strategies and dynamic zoning management.

[0042] In this implementation scheme, through real-time acquisition and dynamic fusion of multi-dimensional data, precise modeling and response compensation of dynamic thermal drift and multi-source disturbances during the cutting process are achieved using thermo-mechanical coupling compensation theory. Parameters are organically integrated after multi-layer algorithm processing, ensuring that the compensation criteria can take into account the complex effects of thermal, mechanical, and disturbance physical quantities. By comprehensively considering factors such as directionality, thermal deformation, disturbance dynamics, historical trends, and safety margins, sub-item compensation calculations can be performed on each structural unit, ultimately outputting the overall thermal drift response value. The response value not only reflects the spatial offset intensity under complex thermo-mechanical-disturbance conditions but also provides a quantitative basis for real-time path fine-tuning, compensation decisions, and risk classification and control during the cutting process. Overall, this process significantly improves the adaptive compensation and intelligent control capabilities under highly dynamic and complex process environments, providing theoretical support for achieving high-precision, stable, and safe laser cutting process control.

[0043] Specifically, the process of fine-tuning and high-precision trajectory correction of the path node space is as follows: real-time comparison of thermal drift response values ​​and thermal drift response thresholds, including primary and secondary response thresholds; execution of a detailed compensation process; continuous monitoring of the thermal drift response status of each partition; and ensuring that the compensation strategy is dynamically switched according to real-time operating conditions. When the thermal drift response value is greater than or equal to the first-level response threshold, the execution of subsequent uncut paths is suspended, the abnormal partition is immediately locked, the current partition is subjected to regional inspection, the compensation parameters are reviewed, and an abnormal area report is output to provide decision support for subsequent manual intervention and parameter optimization.

[0044] When the thermal drift response value is greater than or equal to the second-level response threshold but less than the first-level response threshold, the system analyzes the thermal drift response value fluctuations based on sub-item dynamic analysis to determine whether they are due to temperature rise and disturbance. It then identifies the fluctuation type based on historical trends and dynamically identifies the causes of temperature rise and disturbance. If they are due to temperature rise and disturbance, it reduces the feed rate, laser pulse width, and energy density, adjusts process parameters to reduce system load, and slows down local drift development. Based on a vector projection-based path node spatial compensation algorithm, it calculates the sub-item compensation amount for thermal drift response values ​​greater than the risk threshold in real time. The sub-item compensation amount is then superimposed onto the original path node coordinates according to the path direction component, and the cutting trajectory is dynamically vector-fine-tuned to ensure that the compensation amount accurately acts on each key node according to the actual spatial direction. Corresponding correction CNC commands are generated in real time, and these commands are synchronously sent to the motion controller. The path is updated online in real time, improving the intelligence level of compensation and CNC linkage.

[0045] When the thermal drift response value is less than the secondary response threshold, real-time monitoring of the remaining path is maintained, the trend of response value change is dynamically recorded, the cutting path execution continues to proceed according to the current state, and the thermal drift response value is continuously monitored along the remaining path to ensure the continuity and safety of process operation.

[0046] In this implementation scheme, high-precision fine-tuning and dynamic trajectory correction of path node space are achieved by real-time monitoring of thermal drift response values ​​and their relationship with multi-level response thresholds. The hierarchical response strategy ensures that when anomalies occur, the abnormal zone can be immediately identified and a detailed report output, quickly assisting in manual intervention and parameter optimization. In the critical fluctuation range, the causes of thermal drift are identified based on dynamic analysis and historical trend judgment, process parameters are intelligently adjusted to slow down the development of local thermal drift, and a vector projection compensation algorithm is used to perform fine spatial fine-tuning of key nodes in the cutting path, generating and issuing correction CNC commands in real time to ensure a high degree of coordination between trajectory and compensation. When the thermal drift response is within the normal range, high-frequency monitoring and data recording are continuously performed on the remaining path to dynamically ensure the stability and safety of the cutting process. This process effectively improves the trajectory self-adaptation and high-precision control capabilities of laser cutting in complex thermal disturbance environments, greatly enhancing the intelligence, stability, and risk controllability of the process.

[0047] Specifically, the process based on time series fluctuation and mutation criteria is as follows: First, acquire the current time, the compensated thermal drift response value sequence data, the monitoring start time, and the monitoring end time, ensuring that all key data are collected strictly according to a unified time sequence, laying a foundation for time consistency in subsequent dynamic risk assessment. Second, obtain the mean of the response sliding window using a sliding window mean algorithm on the compensated thermal drift response value sequence data, employing an adaptive window length to balance real-time performance and trend smoothing in the mean calculation. Third, obtain the standard deviation of the response sliding window using a sliding window standard deviation algorithm on the compensated thermal drift response value sequence data, improving the sensitivity to changes in local fluctuation intensity. Fourth, obtain the time derivative of the response value using a numerical differentiation algorithm on the compensated thermal drift response value sequence data, reflecting the rate of change and dynamic mutation of the response value in real time. Fifth, obtain the complexity of the response sliding window using a sample entropy algorithm on the compensated thermal drift response value sequence data. Sample entropy quantifies the complexity and anomaly of the response sequence within a local time period, improving the accuracy of risk assessment.

[0048] The absolute value of the difference between the compensated thermal drift response value and the mean of the response sliding window is multiplied by the standard deviation of the response sliding window to obtain the response deviation fluctuation term, which sensitively characterizes the deviation and fluctuation intensity of the current response value relative to the historical mean. The absolute value of the derivative of the compensated thermal drift response value with respect to time is calculated and then incremented by one to obtain the response change rate term, ensuring stability and discriminative power during periods of drastic change and abrupt change. The response deviation fluctuation term is divided by the response change rate term to obtain the response fluctuation value, dynamically balancing the offset amplitude and change rate to achieve a more robust quantitative risk criterion. The response sliding window complexity is calculated and incremented by one to obtain the response complexity adjustment term, which adaptively amplifies the impact of high-complexity anomaly intervals on risk. The response fluctuation value is multiplied by the response complexity adjustment term, and the maximum value within the monitoring period is taken to obtain the compensated risk value. The specific calculation formula is as follows: ; In the formula, This represents the compensation risk value, which quantifies the overall risk level of the thermal drift response after compensation within the monitoring period. It represents the current moment and serves as a unified time index for sampling, discrimination, and calculation; It represents the thermal drift response value after compensation, dynamically reflecting the compensation effect and spatial state at time t, and intuitively demonstrating the compensation accuracy under the current working conditions; This represents the mean of the sliding window response. It characterizes the stationary trend of the response value within a local time period. It represents the standard deviation of the sliding window of the response, measuring the fluctuation of the response value within the same window; It represents the time derivative of the response value, reflecting the rate and trend of change of the response value over time; It represents the sliding window complexity of the response, quantitatively describes the local complexity and uncertainty of the response sequence, and is used to highlight the risk impact of high-entropy anomaly intervals; This indicates the start time of monitoring, serving as the starting point of the monitoring interval to ensure that the risk calculation interval strictly corresponds to the actual process flow; Indicates the monitoring cutoff time and defines the end point of the monitoring interval, which helps with process archiving and periodic risk assessment.

[0049] This implementation plan comprehensively realizes the quantitative analysis and risk assessment of the dynamic fluctuations and abrupt changes in the compensated thermal drift response sequence. A sliding window mean and standard deviation algorithm is used to achieve real-time tracking of the local trends and fluctuation intensity of the response sequence. Through numerical differentiation and sample entropy calculation, multi-dimensional features of the response sequence, such as rate of change, abrupt changes, and complexity, are effectively extracted. The calculation results collectively construct a hierarchical criterion, which can sensitively capture instantaneous deviations and local anomalies, and dynamically adjust the risk amplification for high-complexity intervals. This achieves comprehensive monitoring of potential anomalies and trend abrupt changes under different operating conditions, and quantitative assessment of extreme risks. The final output compensated risk value can comprehensively quantify the risk level after compensation throughout the entire monitoring period, providing a scientific and reliable decision-making basis for graded early warning, dynamic intervention, and continuous optimization in the process, thereby improving the safety, stability, and intelligent management capabilities of the laser cutting process.

[0050] Specifically, the process of real-time monitoring and tiered intervention for time-series response fluctuations and abnormal risks in the compensation process is as follows: real-time comparison of compensation risk values ​​and compensation risk thresholds, including primary risk thresholds and secondary risk thresholds, to achieve dynamic prevention and control by region. like Figure 4 The diagram shows the dynamic prevention and control process for risk value zoning. When the risk value is less than the secondary risk threshold, it is determined to be in the process health zone and enters the regular data archiving process. The process health database is established and the structural thermal condition data and process path spatial characteristic data are dynamically updated and archived into the process health database. No special actions are required, which continuously ensures the accumulation of historical data and the ability to trace the process.

[0051] When the compensation risk value is greater than or equal to the secondary risk threshold but less than the primary risk threshold, it is identified as a process concern area. The sampling cycle is then encrypted and shortened to half of the current cycle. The nodes of the concern area are digitally marked to facilitate subsequent dynamic backtracking and precise monitoring.

[0052] When the compensated risk value is greater than or equal to the first-level risk threshold, it is identified as a process warning zone. Physical interventions such as spray cooling and vibration isolation are implemented in the zone. During the spray cooling process, the cooling flux and spray duration are adjusted according to the actual degree of thermal drift. By adjusting the cooling intensity and time per unit area, the temperature of the zone is effectively reduced rapidly, thereby suppressing structural warping and thermal deformation. If the compensated risk value is still greater than or equal to the first-level risk threshold after physical intervention, the process is temporarily suspended, and a process anomaly report is generated by pushing multi-dimensional abnormal parameters. It is recommended that maintenance personnel conduct inspections to ensure that the risk is controllable and the abnormal event is handled in a closed loop.

[0053] This implementation plan achieves intelligent hierarchical control of temporal response fluctuations and abnormal risks during the compensation process. In the process health zone, various process and structural thermal data are continuously archived to ensure the integrity and traceability of process data. In the process concern zone, the sampling frequency is increased and key areas are digitally marked, enabling more precise dynamic monitoring and process recording of risk nodes. When entering the process warning zone, physical intervention measures such as spray cooling and vibration isolation are intelligently activated, and the cooling intensity and duration are dynamically adjusted according to the actual thermal drift of each zone, significantly improving the local temperature drop rate and structural stability, and preventing thermal deformation and warping. When physical intervention fails to eliminate high-risk conditions, the process is promptly paused, abnormal parameters are pushed, and a report is generated, recommending manual inspection to ensure timely closed-loop handling of high-risk events. This greatly enhances the timeliness and precision of compensation risk identification, hierarchical response, and intelligent intervention, providing strong support for the safety, stability, and intelligent control of the laser cutting process.

[0054] like Figure 2 As shown, the second aspect of this invention provides a cutting positioning device for laser cutting metal based on big data, including an acquisition and preprocessing module for acquiring structural thermal condition data and process path spatial feature data, obtaining historical disturbance condition data, and preprocessing the thermal condition data, process path spatial feature data, and historical condition data; an initial thermal drift spatial compensation and verification module for predicting and initially calibrating the trend of multi-point thermal expansion, thermal fluctuation, and disturbance coupled spatial offset before cutting based on time-series thermal fluctuation analysis, and entering a detailed compensation process; a thermal drift compensation strategy module for fine-tuning and high-precision trajectory correction of path node space through thermal coupling compensation theory analysis for dynamic thermal drift and disturbance during the cutting process; and a compensation risk monitoring and intervention module for real-time monitoring and graded intervention of time-series response fluctuations and abnormal risks in the compensation process based on time-series fluctuation and mutation criteria.

[0055] This implementation scheme achieves multi-dimensional data fusion and dynamic management of structural thermal conditions, process path spatial characteristics, and disturbance history through full-process big data collection and efficient preprocessing. It can accurately predict and calibrate spatial offsets caused by multi-point thermal expansion, thermal fluctuations, and disturbance coupling before cutting, and continuously perform adaptive high-precision corrections on dynamic thermal drift and path node fine-tuning during cutting. Real-time compensation across the entire process, combined with temporal fluctuation and abrupt change criteria, identifies abnormal risks and dynamically intervenes according to a tiered strategy, ensuring the stability of cutting positioning and the safety and reliability of the process. The device significantly improves the intelligence level and dynamic risk response capability of metal laser cutting positioning, effectively guaranteeing cutting accuracy and process stability in high-end manufacturing scenarios.

[0056] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0057] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A cutting positioning method for laser cutting of metal based on big data, characterized in that, Includes the following steps: S1. Collect structural thermodynamic condition data and process path spatial characteristic data, obtain historical disturbance condition data, and preprocess the thermodynamic condition data, process path spatial characteristic data, and historical condition data. S2, based on time-series thermal fluctuation analysis, predicts and initially calibrates the trend of multi-point thermal expansion, thermal fluctuation and disturbance coupled spatial displacement before cutting, and then enters a detailed compensation process; S3, targeting the dynamic thermal drift and disturbance during the cutting process, uses thermal coupling compensation theory to fine-tune the path node space and correct the trajectory with high precision. S4, based on time series fluctuation and mutation criteria, performs real-time monitoring and graded intervention on the time series response fluctuation and abnormal risk of the compensation process; The specific process based on time-series thermal fluctuation analysis is as follows: The system acquires data on the number of compensation structural components, temperature length variation, geometric dimensions, real-time temperature, initial temperature variables, surface temperature sequence data, current moment, triaxial acceleration, and angular variation. It then uses a least-squares linear regression algorithm to obtain the coefficient of thermal expansion from the temperature length variation data; a three-dimensional distance calculation algorithm to obtain the effective feature length from the geometric dimensions; a temperature mean algorithm to obtain the real-time temperature variable from the real-time temperature data; a range calculation to obtain the thermal fluctuation amplitude factor from the surface temperature sequence data; and a fast Fourier transform analysis of the surface temperature sequence data to determine the dominant frequency, multiplying 2π by the dominant frequency to obtain the thermal fluctuation angular frequency. The number of disturbance terms is obtained from the triaxial acceleration data and angle change data through Euclidean norm fusion and variance threshold detection algorithms; the spatial offset value of process disturbance is obtained from the triaxial acceleration data and angle change data through Euclidean norm fusion, vector projection and absolute value operation. The thermal expansion offset term of the structural component is obtained by calculating the product of the coefficient of thermal expansion, the effective feature length, and the difference between the surface temperature and the static temperature. The thermal fluctuation offset term of the structural component is obtained by multiplying the thermal fluctuation amplitude factor by the thermal fluctuation angular frequency and the sine value at the current moment. The thermal expansion offset term and the thermal fluctuation offset term of the structural component are added together to obtain the predicted spatial offset value of the structural component. The predicted spatial offset values ​​of all structural components are summed together, and then summed together with the predicted spatial offset values ​​of all process disturbances to obtain the predicted spatial offset value. The specific calculation formula is as follows: ； In the formula, This represents the predicted spatial offset, which is the global spatial offset output after combining multi-source thermal conditions and disturbances. This indicates the number of compensating structural components, which is the total number of structural units involved in the compensation calculation. It represents the coefficient of thermal expansion, reflecting the expansion characteristics of a material after it is heated; Indicates the effective feature length, which is the initial length parameter of the structural component; This represents the real-time temperature variable, indicating the actual temperature of the structural component at the current moment. This represents the initial temperature variable, indicating the reference temperature of the structural component at the initial moment; This represents the thermal fluctuation amplitude factor, used to measure the magnitude of temperature changes. It represents the angular frequency of thermal fluctuations, used to characterize the periodicity of temperature changes; This indicates the current moment, the point in time at which the judgment and calculation are performed; This indicates the number of perturbation terms, representing the number of perturbations that need to be considered during the cutting process; This represents the spatial offset value of the process disturbance, used to reflect the degree of spatial offset of each disturbance; The specific process of predicting and initially calibrating the trend of multi-point thermal expansion, thermal fluctuation and disturbance coupled spatial offset before cutting, and then entering the detailed compensation process is as follows: Real-time comparison of spatial offset prediction values ​​and spatial offset prediction thresholds: When the predicted spatial offset is less than the predicted spatial offset threshold, it is determined to be a normal initial offset state. Through three-dimensional geometric projection and vector superposition algorithm, the predicted thermal spatial offset is projected onto the three-dimensional coordinate axis, which is converted into the preliminary three-dimensional coordinate compensation to obtain the predicted three-dimensional value of thermal spatial offset. This value is then applied to the CNC zero point and path offset compensation, and the detailed compensation process begins. When the predicted spatial offset value is greater than or equal to the predicted spatial offset threshold, it is determined to be an initial abnormal offset state. Extend the preheating time, repeat the zero-finding process, and execute the workpiece static waiting process until the spatial offset prediction value is less than the spatial offset prediction value threshold. The specific process based on the time series fluctuation and mutation criterion is as follows: Acquire the current time, the compensated thermal drift response value sequence data, the monitoring start time, and the monitoring end time; obtain the sliding window mean of the compensated thermal drift response value sequence data using the sliding window mean algorithm; obtain the sliding window standard deviation of the compensated thermal drift response value sequence data using the sliding window standard deviation algorithm; obtain the time derivative of the response value using the numerical differentiation algorithm. The sliding window complexity of the response is obtained by using the sample entropy algorithm on the compensated thermal drift response value sequence data. Calculate the absolute value of the difference between the compensated thermal drift response value and the mean of the response sliding window, multiply it by the standard deviation of the response sliding window to obtain the response deviation fluctuation term; calculate the absolute value of the derivative of the compensated thermal drift response value with respect to time and add one to obtain the response change rate term. Divide the response deviation fluctuation term by the response change rate term to obtain the response fluctuation value; Calculate the response sliding window complexity plus one to obtain the response complexity adjustment term; Multiply the response volatility value by the response complexity adjustment term, and take the maximum value within the monitoring period to obtain the compensated risk value; The specific process of real-time monitoring and graded intervention of the time-series response fluctuations and abnormal risks in the compensation process is as follows: Real-time comparison of compensation risk value and compensation risk threshold, including primary risk threshold and secondary risk threshold: When the compensation risk value is less than the secondary risk threshold, it is determined to be a process health zone. A process health database is established and the structural thermal condition data and process path spatial characteristic data are dynamically updated and archived into the process health database without any special action. When the compensation risk value is greater than or equal to the secondary risk threshold and less than the primary risk threshold, it is determined to be a process concern area, the sampling cycle is increased, and the concern area nodes are digitally marked. When the compensated risk value is greater than or equal to the first-level risk threshold, it is determined to be a process warning zone. Physical interventions such as spray cooling and vibration isolation are implemented in the zone. If the compensated risk value is still greater than or equal to the first-level risk threshold after physical intervention, the process is temporarily suspended, multi-dimensional abnormal parameters are pushed to generate a process abnormality report, and maintenance personnel are advised to conduct inspections.

2. The cutting positioning method for laser cutting metal based on big data according to claim 1, characterized in that: The specific process for acquiring historical disturbance data by collecting structural thermal condition data and process path spatial characteristic data is as follows: Collect structural thermal condition data, which includes: the number of compensating structural components, temperature length variation data, geometric dimension data, real-time temperature data, initial temperature variables, surface temperature sequence data, current time, triaxial acceleration data, and angle variation data; Collect process path spatial feature data, which includes: main path spatial coordinate point data, compensation vector direction data, length change data, current spatial coordinates, monitoring start time, and monitoring end time; Acquire historical disturbance data and establish a historical disturbance data database. The historical disturbance data includes: historical triaxial acceleration data and historical angle change data. The number of compensation structural components is the total number of structural units involved in the compensation calculation; The temperature length change data is measured by the linkage of a grating ruler and a thermocouple.

3. The cutting positioning method for laser cutting metal based on big data according to claim 1, characterized in that: The specific process for preprocessing thermal condition data, process path spatial feature data, and historical condition data is as follows: By using outlier removal and multi-source time alignment algorithms, missing values ​​are filled in and data is synchronized for structural thermal condition data to ensure that various thermal parameters correspond at the same time. By using spatial coordinate normalization and node spacing balancing algorithms, path resampling and spatial partitioning are performed on the spatial feature data of the process path to extract the key node features of the main path. By using window mean filtering and disturbance amplitude normalization algorithms, multi-channel data denoising and smoothing are performed on historical disturbance data to enhance the effective expression of disturbance change trends. By using distribution standardization and linear normalization algorithms, structural thermal condition data, process path spatial feature data, and historical disturbance data are standardized and normalized.

4. The cutting positioning method for laser cutting metal based on big data according to claim 1, characterized in that: The specific process of analyzing the dynamic thermal drift and disturbance during the cutting process using thermo-mechanical coupling compensation theory is as follows: The system acquires the following data: thermal spatial offset prediction 3D value, main path spatial coordinate point data, compensation vector direction data, real-time temperature data, length change data, current spatial coordinates, triaxial acceleration data, angle change data, historical triaxial acceleration data, and historical angle change data. It calculates the angle between the compensation vector and the main cutting path using the vector angle calculation based on the main path spatial coordinate point data and compensation vector direction data. It obtains the workpiece's main material thermal expansion coefficient using the least squares linear regression algorithm based on the real-time temperature data and length change data. It obtains the remaining feature length using a 3D Euclidean distance calculation algorithm based on the main path spatial coordinate point data and current spatial coordinates. Finally, it obtains the disturbance response amplification factor using an extremum algorithm and sensitivity calibration method based on the triaxial acceleration data and angle change data. The overall spatial drift increment is obtained by using a weighted Euclidean norm fusion algorithm and window accumulation operation on triaxial acceleration data and angle change data; the historical spatial drift increment is obtained by using a weighted Euclidean norm fusion algorithm and window accumulation operation on historical triaxial acceleration data and historical angle change data; and the disturbance smoothing coefficient is obtained by using a sliding window averaging algorithm on the historical spatial drift increment. The spatial compensation term is obtained by multiplying the predicted three-dimensional value of the thermal spatial offset and the compensation vector with the cosine of the angle between them and the main cutting path. The thermal expansion offset term is obtained by multiplying the workpiece's main material thermal expansion coefficient, remaining characteristic length, and the difference between the real-time temperature variable and the initial temperature variable. The spatial compensation term is added to the thermal expansion offset term to obtain the path offset term. The disturbance response amplification factor is calculated and the overall spatial drift increment is calculated to obtain the disturbance drift term. The path offset term and the disturbance drift term are added together, and the absolute value is taken as the numerator. The sum of the safety drift amount, the absolute value of the difference between the real-time temperature variable and the initial temperature variable, and the disturbance smoothing coefficient is calculated as the denominator. The numerator is divided by the denominator, and the results for all structural components are summed to obtain the thermal drift response value. The safety drift amount represents the compensation constant obtained by using a denominator minimization robustness constraint algorithm to calculate the absolute value of the difference between the initial temperature variable and the real-time temperature variable.

5. The cutting positioning method for laser cutting metal based on big data according to claim 1, characterized in that: The specific process of fine-tuning the path node space and correcting the high-precision trajectory is as follows: Real-time comparison of thermal drift response values ​​and thermal drift response thresholds, including primary and secondary response thresholds, and execution of a detailed compensation process: When the thermal drift response value is greater than or equal to the first-level response threshold, the execution of subsequent uncut paths is paused, the current partition performs a regional inspection, verifies the compensation parameters, and outputs an abnormal area report. When the thermal drift response value is greater than or equal to the second-level response threshold and less than the first-level response threshold, the thermal drift response value fluctuation is detected based on sub-item dynamic analysis to determine whether it belongs to temperature rise and disturbance. If it belongs to temperature rise and disturbance, the feed rate, laser pulse width and energy density are reduced to slow down the development of local drift. Based on the path node space compensation algorithm of vector projection, the sub-item compensation amount for thermal drift response values ​​greater than the risk threshold is calculated in real time. The sub-item compensation amount is superimposed on the original path node coordinates according to the path direction component, and the cutting trajectory is dynamically vector fine-tuned. The corresponding correction CNC command is generated in real time, and the CNC command is synchronously sent to the motion controller, and the path is updated online in real time. When the thermal drift response value is less than the secondary response threshold, the cutting path execution continues to proceed according to the current state, and the thermal drift response value is continuously monitored along the remaining path.

6. A cutting positioning device for laser cutting metal based on big data, employing the cutting positioning method for laser cutting metal based on big data as described in any one of claims 1-5, comprising: The acquisition and preprocessing module is used to acquire structural thermodynamic condition data and process path spatial characteristic data, obtain historical disturbance condition data, and preprocess the thermodynamic condition data, process path spatial characteristic data, and historical condition data. The initial thermal drift spatial compensation and verification module is used to predict and initially calibrate the trend of multi-point thermal expansion, thermal fluctuation and disturbance coupled spatial offset before cutting based on time-series thermal fluctuation analysis, and then enter the detailed compensation process. The thermal drift compensation strategy module is used to fine-tune the path node space and correct the trajectory with high precision by analyzing the dynamic thermal drift and disturbance during the cutting process through thermal coupling compensation theory. The compensation risk monitoring and intervention module is used to monitor and intervene in the time-series response fluctuations and abnormal risks of the compensation process in real time based on time-series fluctuation and mutation criteria.

Images