Hardware part size compensation method and platform for machining errors

Abstract

The invention discloses a machining error-oriented hardware part size compensation method and platform, and relates to the technical field of industrial control, in particular to the technical field of hardware part machining. The method comprises the steps that based on a to-be-machined part, structural basic characteristics are obtained, and a three-dimensional size compensation map is drawn up with the elasticity modulus of a material as a constraint; matching each tool path target compensation amount sequence based on the map, and synchronously predicting a size deviation value; introducing an environmental interference factor, and inputting the pre-training model to generate a dynamic correction coefficient; and on the basis of the data, cutter compensation, the feeding rate and the rotating speed are dynamically adjusted through a self-adaptive strategy till the machining error is stabilized within the tolerance range. The technical problem that in traditional hardware part machining, due to the fact that a static compensation strategy is difficult to adapt to a dynamic machining environment, the machining size precision is unstable is solved, and the technical effects that self-adaptive high-precision size compensation based on the real-time machining state and multi-source interference sensing is achieved, and the part machining consistency and the percent of pass are improved are achieved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of industrial control, in particular to the technical field of hardware part machining, and especially to a hardware part size compensation method and platform for machining errors. BACKGROUND

[0002] With the development of intelligent manufacturing and precision machining technology, the dimensional precision requirements of hardware parts in the industries of mold manufacturing, consumer electronics, automobile parts and high-end equipment are continuously improved. Although modern numerical control machining has a high automation level, it is still affected by multiple interference factors in the actual machining process, such as elastic deformation caused by insufficient rigidity of the machine tool structure, dimensional drift caused by machining thermal effects, cutting trajectory deviation caused by tool wear, and transient errors caused by cutting force fluctuations. These factors jointly act on the hardware parts, which fail to fully utilize the structural characteristics of the parts and the surrounding environmental factors in the machining process, resulting in dimensional deviation from the design value, thereby increasing the rework rate, reducing the machining efficiency, and even affecting the quality of the whole machine assembly. SUMMARY

[0003] The application provides a hardware part size compensation method and platform for machining errors, which solves the technical problem of unstable machining dimensional precision caused by the difficulty of traditional static compensation strategies in adapting to dynamic machining environments in the machining of hardware parts.

[0004] In a first aspect, the application provides a hardware part size compensation method for machining errors, which comprises: Based on a hardware part to be machined, structural basic features including surface roughness distribution and geometric dimensional deviation are obtained, and a three-dimensional dimensional compensation map is drafted based on material elastic modulus data as a constraint condition. Based on the three-dimensional dimensional compensation map, a target compensation amount sequence of each tool path in a numerical control machining assembly is matched, and dimensional deviation values in the machining process are synchronously predicted. Machining environment interference factors including a machine tool bed temperature drift curve, tool wear rate and cutting force fluctuation spectrum are introduced and input into a pre-trained error compensation model to generate a dynamic correction coefficient. Based on the target compensation amount sequence, the dimensional deviation values and the dynamic correction coefficient, a tool radius compensation value, a feed rate and a spindle speed of the numerical control machining assembly are dynamically adjusted by an adaptive control strategy until the machining dimensional error is stabilized within a dimensional tolerance threshold.

[0005] In a second aspect, the application provides a hardware part size compensation platform for machining errors, which comprises: The atlas mapping module: based on the hardware parts to be processed, the structural basic features including surface roughness distribution, geometric size deviation are obtained, and the three-dimensional size compensation atlas is mapped based on the material elastic modulus data as the constraint condition; the deviation prediction module: based on the three-dimensional size compensation atlas, the target compensation quantity sequence of each tool path in the numerical control machining assembly is matched, and the size deviation value in the machining process is synchronously predicted; the environment compensation module: the machining environment interference factors including the machine tool bed body temperature drift curve, the tool wear rate and the cutting force fluctuation spectrum are introduced and input into the pre-trained error compensation model to generate a dynamic correction coefficient; the dynamic adjustment module: based on the target compensation quantity sequence, the size deviation value and the dynamic correction coefficient, the tool radius compensation value, the feed rate and the spindle speed of the numerical control machining assembly are dynamically adjusted by the adaptive control strategy until the machining size error is stable within the size tolerance threshold.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages: First, the overall feature analysis is performed on the parts to be processed, and the three-dimensional size compensation atlas is established through the information such as surface roughness, size deviation and material elastic modulus. Then, the corresponding compensation quantity is assigned to each tool path in numerical control machining according to the atlas, and the size error is predicted in real time during the machining process. After that, the machine tool temperature drift, tool wear and cutting force fluctuation and other environmental interference factors are input into the pre-trained compensation model to obtain a dynamic correction coefficient. Finally, the compensation quantity, the predicted error and the dynamic correction parameter are comprehensively considered, and the adaptive control strategy is used to adjust the tool compensation value, the feed speed and the spindle speed in real time, so that the machining process is continuously optimized until the part size error is kept within the specified tolerance range, the size compensation in the complex machining environment is realized, and the stability and the finished product quality in the machining process are improved. BRIEF DESCRIPTION OF DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0008] Figure 1 The hardware part size compensation method flowchart for machining error provided by the embodiment of the present application.

[0009] Figure 2 The hardware part size compensation platform structure schematic diagram for machining error provided by the embodiment of the present application.

[0010] Marked explanation: atlas mapping module 11, deviation prediction module 12, environment compensation module 13, dynamic adjustment module 14. DETAILED DESCRIPTION

[0011] To further clarify the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the specific embodiments, structures, features and effects thereof according to the present application are described in detail below in conjunction with the drawings and preferred embodiments.

[0012] In an embodiment, as shown in the accompanying drawings, the present application provides a hardware part size compensation method for machining errors, which comprises the following steps: Figure 1 Based on a hardware part to be machined, structural basic features including surface roughness distribution and geometric size deviation are obtained, and a three-dimensional size compensation atlas is prepared with material elastic modulus data as a constraint condition. Based on a hardware part to be machined, structural basic features including surface roughness distribution and geometric size deviation are obtained, and a three-dimensional size compensation atlas is prepared with material elastic modulus data as a constraint condition.

[0013] In the embodiment of the present application, before machining, multi-dimensional feature acquisition is performed on the hardware part to be machined. First, the topographic data of the part surface is obtained by a three-dimensional optical scanner, a laser displacement sensor or a white light interferometer, etc., from which the surface roughness distribution parameters (such as Ra, Rz, Rq, etc.) are extracted, and then the geometric size of the part is measured by a high-precision three-coordinate measuring machine (CMM) and compared with the design model to obtain geometric size deviation data for describing the geometric error distribution characteristics of the part in different regions. Subsequently, the obtained surface roughness distribution data and geometric size deviation data are added to a set to establish the structural basic features of the part. Then, the material attribute data of the part to be machined, such as the material elastic modulus data, is retrieved from the material database, and the material elastic modulus data is used as a constraint condition for multi-dimensional fusion processing with the structural basic features to generate a three-dimensional size compensation atlas, which is used to guide the subsequent optimization configuration of the tool path and the allocation of the target compensation amount, and is the core data basis of the entire machining error compensation process. Through the atlas, directional compensation control of different machining regions, different material properties and complex curved surface parts can be realized, providing a basic support for subsequent dynamic correction and adaptive control.

[0014] Further, the three-dimensional size compensation atlas is prepared with the material elastic modulus data as a constraint condition, and the method comprises the following steps: The structural basic features of the hardware part to be machined are obtained, and the material elastic modulus data is determined by using the material database; the structural basic features and the material elastic modulus data are fused to construct the three-dimensional size compensation atlas, and the three-dimensional size compensation atlas contains at least three compensation gradient partitions.

[0015] Preferably, after obtaining the structural basic features of the hardware part to be processed, the physical properties of the metal material used by the part are queried through a material performance database, such as a metal material performance table, a factory material database or a third-party material library, the material parameters such as the elastic modulus of the material are obtained, and the elastic modulus is taken as a constraint condition to reflect the sensitivity of the part material to processing force and thermal effect, and to provide a material characteristic basis for the construction of the compensation map. After obtaining the structural basic features and the elastic modulus of the material, data fusion processing is performed on the two, which can specifically adopt a multi-dimensional weighted modeling method, wherein the geometric size deviation distribution is taken as a main variable, the surface roughness distribution is taken as a local correction amount, and the material elastic modulus is taken as a constraint coefficient. After normalization by dividing by the maximum deviation, the maximum roughness and the minimum elastic modulus, the three-dimensional space coordinates of the part are weighted and calculated. For the material region with high rigidity (high elastic modulus), the compensation gradient can be appropriately reduced; for the region with weak rigidity and prone to deformation, the compensation gradient is increased to form a differentiated compensation strategy. After data fusion, a three-dimensional size compensation map is constructed. The three-dimensional size compensation map takes the three-dimensional model coordinates of the part as the basis, and represents the compensation direction and compensation amount of each space position in the form of layered gradients. The layered gradients are divided into at least three layers, corresponding to the first compensation gradient partition, the second compensation gradient partition and the third compensation gradient partition, respectively. The first compensation gradient partition corresponds to the region with the most obvious size deviation or stress sensitivity, and the compensation amount is the largest. The second compensation gradient partition corresponds to the general machining region, and the compensation amount is moderate. The third compensation gradient partition corresponds to the region with smaller size deviation and stable machining, and the compensation amount is the smallest. The above compensation gradient partitions can be automatically generated according to the size deviation field threshold, the local curvature change rate and the material elastic modulus, thereby forming a continuous and executable three-dimensional compensation distribution map for guiding the compensation amount configuration of the subsequent tool path, and laying a data foundation for improving the machining precision.

[0016] Based on the three-dimensional size compensation map, the target compensation amount sequence of each tool path in the numerical control machining assembly is matched, and the size deviation value in the machining process is predicted synchronously.

[0017] In one embodiment, after obtaining the three-dimensional size compensation map, the three-dimensional size compensation map is first data-matched with tool path information in a numerical control machining program (NC program). By analyzing the G code or tool trajectory file of the numerical control machining assembly, the spatial coordinate point sequence, machining direction, feed rate, and machining layer depth of each tool path are extracted. Subsequently, according to the coordinate system of the three-dimensional model of the part, the tool path is coordinate-aligned and spatially mapped with the spatial compensation partition in the compensation map, so as to determine the compensation gradient partition to which each path belongs. After the path matching is completed, according to the compensation gradient density and local curvature change rate in the corresponding partition, the target compensation amount of each tool path segment is calculated, and then these target compensation amounts are projected onto the tool path coordinate points to generate a target compensation amount sequence arranged in path order. At the same time, in the process of tool path matching and compensation amount generation, a prediction model based on long short-term memory network (LSTM) is also used to simultaneously predict the size deviation value in the machining process. This prediction model estimates the real-time size deviation in the machining process due to cutting force, thermal deformation, and tool wear based on historical machining data and current tool parameters. The predicted size deviation value corresponds to the target compensation amount sequence, providing a reference basis for subsequent dynamic correction and adaptive control. The prediction model has previously performed iteration training steps such as forward propagation, loss calculation, back propagation, and parameter optimization. Through the above process, a complete path-compensation amount-deviation value correspondence can be established before machining begins, realizing feedforward planning of the compensation strategy and real-time deviation prediction, and laying a foundation for subsequent dynamic adjustment of tool compensation, feed rate, and spindle speed.

[0018] Further, based on the three-dimensional size compensation map, the target compensation amount sequence of each tool path in the numerical control machining assembly is matched, and the method comprises: According to each compensation gradient partition of the three-dimensional size compensation map, the first tool path is divided into a plurality of machining segments corresponding to each compensation gradient partition; for the plurality of machining segments, the target compensation amount of the plurality of machining segments is configured in combination with the spatial curvature and the compensation density requirement of the compensation gradient partition, to obtain the target compensation amount sequence.

[0019] Optionally, after obtaining the three-dimensional size compensation map, the first tool path trajectory information in the numerical control machining program is read, including the spatial continuous coordinate points of the path, the machining direction, the path length, and the local curvature and other parameters. Subsequently, taking the coordinate system of the three-dimensional model of the part as the reference, the first tool path is mapped point by point to the three-dimensional size compensation map, so that any point on the path can be corresponded to the compensation gradient partition of the map. After completing the coordinate mapping, according to each compensation gradient partition delineated in the three-dimensional size compensation map, the first tool path is automatically divided into multiple machining sections, each machining section corresponding to one compensation gradient partition. When the path coordinate point is located in the first compensation gradient partition, it is divided into a high-compensation machining section; when it is located in the second compensation gradient partition, it is divided into a medium-compensation machining section; and when it is located in the third compensation gradient partition, it is divided into a low-compensation machining section. The division process can adopt a threshold-based partition rule to ensure accurate switching and no overlap of the tool path between different compensation regions. After the path is successfully segmented, the compensation amount of each machining section is calculated. In this process, the spatial curvature information in the machining section is first obtained. The area with large curvature usually corresponds to uneven machining stress, complex surface topography, or insufficient local stiffness of the material, so the compensation density needs to be increased. Conversely, the area with small curvature can use a lower compensation density. The spatial curvature value is used as a weighting factor to fuse and calculate the compensation density requirement of the corresponding compensation gradient partition in the compensation map, so as to determine the compensation amount distribution mode of each machining section. Then, using the fused compensation density and the compensation direction, the compensation amount of each coordinate point in the machining section is distributed. The compensation amount can be generated by interpolation fitting, weighted average, or gradient smoothing algorithm to ensure that the compensation amount is continuous, executable, and adaptive to the tool path in the machining section. Taking interpolation fitting as an example, according to the determined starting point and end point of the target compensation amount in the machining section, as well as the actual coordinate point sequence along the tool path, the compensation amount of these coordinate points is fitted by a cubic spline curve, so that the compensation amount changes continuously along the tool path in the entire machining section, thereby avoiding sudden changes in the compensation amount and ensuring smooth and executable path. Finally, the compensation amounts of each machining section are combined in sequence according to the tool path to form a complete target compensation amount sequence. Through the above process, accurate matching of the tool path and the three-dimensional compensation map can be achieved, so that the compensation strategy can be configured differently according to the structural characteristics, curvature variation, and machining difficulty of different areas of the part, thereby providing a basis for subsequent dynamic adjustment of machining parameters.

[0020] Furthermore, with the optimization objectives of machining accuracy uniformity, tool load balance, and energy efficiency, the local curvature of the surface roughness distribution is introduced as a weight correction factor to optimize the configuration of the target compensation amount. At the same time, based on historical machining data and in comparison with the three-dimensional dimensional compensation map, a thermo-mechanical coupling analysis model is constructed. The thermo-mechanical coupling analysis model is used to simulate the temperature field and stress field distribution under the machining state, quantify the dimensional deviation value, and determine whether the compensation amount enhancement mechanism is triggered. The compensation amount enhancement mechanism is used to dynamically correct the target compensation amount sequence.

[0021] Optionally, after configuring the initial toolpath compensation, to further improve machining quality and the rationality of machining parameter distribution, machining accuracy uniformity, tool load balance, and energy efficiency are introduced as comprehensive optimization objectives to optimize and adjust the initial target compensation sequence. Specifically, firstly, a detailed analysis of the surface roughness distribution of the workpiece is performed, extracting the curvature change rate of each local region. Regions with larger local curvature usually correspond to drastic changes in the unevenness of the machined surface or obvious surface transitions. These regions are more susceptible to tool deflection, vibration, or thermal effects during cutting. Therefore, local curvature is used as a weighting correction factor, and it is weighted and fused with the compensation gradient and path features in the three-dimensional dimensional compensation map to optimize the target compensation amount. This enhances the compensation amount in critical regions and appropriately reduces it in smooth regions to meet the optimization objective of machining accuracy uniformity. While optimizing the compensation amount, a thermo-mechanical coupling analysis model is constructed using long-term recorded historical machining data, including spindle load changes, cutting force curves, temperature sensor data, tool wear trajectories, and post-machining dimensional measurements. This model is compared with a three-dimensional dimensional compensation map to establish a unified model of thermal input and mechanical load. This model uses material parameters (such as elastic modulus and coefficient of thermal expansion), tool-workpiece interface heat conduction parameters, and a cutting force model to establish predictive equations for the temperature and stress fields under machining conditions. Subsequently, using the constructed thermo-mechanical coupling model, potential heat sources (including tool cutting heat, spindle drive heat, and frictional heat) and mechanical stress sources (including stress concentration caused by cutting force and clamping force) during machining are simulated to obtain the temperature and stress field distributions during the machining process. Based on simulation results, the dimensional deviations of the machining area in real-time are quantified, including thermal deformation error and mechanical stress error. The thermal deformation error is caused by the thermal expansion of the workpiece, tool elongation, and thermal drift of the machine tool structure due to the temperature field. The mechanical stress error is caused by the elastic or plastic displacement resulting from cutting force, clamping force, and local deformation of the workpiece. After obtaining the dimensional deviation values, a compensation enhancement mechanism is triggered based on a set deviation threshold. If the dimensional deviation of a certain machining segment exceeds the corresponding preset threshold, the enhancement mechanism is automatically activated to dynamically increase the target compensation amount for the corresponding machining segment. This includes increasing the compensation gradient, adjusting the compensation direction, or increasing the compensation density to preemptively offset the error accumulation caused by thermal-mechanical coupling.For example, first, the specific spatial location where the deviation exceeds the limit is determined. Based on the local curvature and compensation gradient partition at that location, the magnitude of the compensation vector is preferentially amplified, increasing the compensation gradient by 10% to 30% of its original proportion. Then, according to the dominant direction of the deviation, the direction of the compensation vector is adjusted to align it with the error direction in the opposite direction, thereby more effectively offsetting the dimensional offset. For processing segments that continuously exceed the threshold, the compensation sampling density of that segment is further increased, densifying the distribution of compensation points to 1.5 to 2 times the original density to improve the spatial resolution of the compensation, making the compensation effect more refined and continuous. The optimized compensation amount is updated to the target compensation amount sequence in real time to achieve dynamic correction and real-time control. Through the above steps, the compensation strategy can be optimized in two layers before and during processing, so that the target compensation amount not only conforms to the regional structural characteristics but also responds in real time to the dimensional deviation changes caused by heat and stress, thereby significantly improving the consistency and stability of the processed dimensions.

[0022] Machining environment interference factors, including machine tool bed temperature drift curve, tool wear rate, and cutting force fluctuation spectrum, are introduced and input into a pre-trained error compensation model to generate dynamic correction coefficients.

[0023] In one embodiment, in actual CNC machining, in addition to part structure and tool path factors, the machine tool operating environment also has a significant impact on machining accuracy. Therefore, environmental interference factors such as machine tool bed temperature drift curve, tool wear rate, and cutting force fluctuation spectrum are introduced to compensate for the dynamic errors of the machining system in real time. First, the temperature drift characteristics of the machine tool bed are collected. This involves using temperature sensors placed at key machine tool nodes, such as the spindle seat, worktable, and guide rail ends, to monitor the temperature change curve over time in real time and obtain the bed temperature drift curve data. This data reflects the spatial thermal expansion characteristics of the machine tool structure caused by changes in ambient temperature, cutting heat, and heat dissipation conditions, which is a significant source of machining dimensional deviations. Then, an online tool monitoring system or cutting force sensing module is used to obtain the tool wear rate. The tool wear rate can be calculated based on changes in spindle current, cutting acoustic emission signals, or tool displacement errors. Furthermore, dynamic force sensors at the spindle or tool holder are used to obtain the amplitude distribution of cutting force at different frequencies. Fast Fourier Transform (FFT) is then used to perform spectral analysis on the time-domain signal to extract the main fluctuation frequencies and amplitudes, forming a cutting force fluctuation spectrum. This spectrum reflects the dynamic interaction characteristics between the tool and the workpiece and can be used to determine the system vibration state and transient cutting instability. Subsequently, the collected three types of environmental interference factors are subjected to data denoising preprocessing, including signal denoising and normalization. The preprocessed feature data is then input into the pre-trained error compensation model to generate dynamic correction coefficients, which are used to correct the target compensation quantity sequence in real time. These dynamic correction coefficients can adaptively change according to the current operating status of the machine tool, realizing online error prediction and compensation adjustment in the machining process.

[0024] Furthermore, the input to the pre-trained error compensation model generates dynamic correction coefficients, including: Data noise reduction preprocessing is performed on the machine tool bed temperature drift curve, tool wear rate, and cutting force fluctuation spectrum among the interference factors in the machining environment to obtain preprocessed data; based on the preprocessed data, an error compensation model under the Transformer architecture is used to generate dynamic correction coefficients.

[0025] Optionally, after obtaining the machine tool bed temperature drift curve, tool wear rate, and cutting force fluctuation spectrum, and other machining environment interference factors, for the machine tool bed temperature drift curve, the time-series data collected by the temperature sensor is subjected to moving mean filtering, Savitzky-Golay smoothing, and first-order difference analysis to remove noise caused by environmental interference, sensor jitter, or transient temperature spikes, so that the temperature drift curve reflects the true trend of thermal deformation of the machine tool structure. Then, the temperature drift curve is sliced ​​according to the time window to extract characteristic parameters such as temperature gradient, temperature change rate, and steady-state drift amplitude. For tool wear rate data, a wear index sequence is constructed by spindle current, cutting acoustic emission signal, or force sensor signal. Wavelet denoising and median filtering methods are used to remove high-frequency noise interference generated during machining. Then, exponential regression or polynomial fitting is used to recover the wear trend curve, and key features such as wear rate, wear stage category (initial wear period, stable wear period, rapid wear period), and current wear rate are extracted. For the cutting force fluctuation spectrum data, the cutting force time-domain signal is converted into frequency-domain data using Fast Fourier Transform (FFT), and invalid frequency noise is removed by a bandpass filter. Then, features such as the main vibration frequency, frequency energy distribution, peak width, and peak offset are extracted based on the amplitude spectrum. After denoising, normalization, and multi-dimensional feature extraction of the above three types of data, a preprocessed data vector of the machining environment is obtained. Subsequently, this preprocessed data vector is input into an error compensation model based on the Transformer architecture. This error compensation model includes a self-attention encoding layer, a multi-head attention mechanism, and a feedforward neural network structure, which can automatically learn the coupling relationship between temperature drift, wear, and cutting force fluctuations. The model first calculates the correlation weights between different environmental features through the self-attention mechanism to identify the dominant factors causing dimensional errors in the machining environment. Then, the weighted features are input into a multi-layer feedforward network to complete the error mapping calculation. Finally, dynamic correction coefficients are output to correct the target compensation amount sequence in real time, enabling the compensation strategy to automatically adjust according to changes in machine tool condition, tool condition, and cutting stability. Through the above process, dynamic compensation can be effectively utilized to utilize interference factors in the processing environment, thereby achieving real-time prediction and high-precision control of processing errors.

[0026] Based on the target compensation sequence, the dimensional deviation value, and the dynamic correction coefficient, the tool radius compensation value, feed rate, and spindle speed of the CNC machining component are dynamically adjusted using an adaptive control strategy until the machining dimensional error stabilizes within the dimensional tolerance threshold.

[0027] In one embodiment, after obtaining the target compensation sequence, dimensional deviation value, and dynamic correction coefficient, the machining control stage begins. This stage uses an adaptive control strategy to dynamically adjust the key process parameters of the CNC machining components in real time, ensuring stable dimensional accuracy under different environmental conditions and error states. First, based on the target compensation sequence, the theoretical dimensional correction target for the current machining segment is calculated. Simultaneously, the real-time dimensional deviation value and dynamic correction coefficient output from the thermo-mechanical coupling model and the environmental compensation model are read to form a machining error feedback signal. The adaptive control strategy uses error feedback as the core control variable and employs a fuzzy control and proportional-integral (PI) composite algorithm to achieve multi-parameter linkage adjustment. It compares the deviation between the current machining dimensional error and the set dimensional tolerance threshold in real time. When the deviation exceeds the threshold, the dynamic adjustment process is automatically triggered. During the adjustment process, the tool radius compensation adjustment is first calculated to correct the effective cutting radius of the tool in the machining coordinate system. Based on the changing trend of the tool radius compensation adjustment, the tool compensation value is dynamically adjusted to ensure that the actual cutting trajectory of the tool is consistent with the target machining contour. Simultaneously, based on real-time error feedback, the feed rate and spindle speed are adjusted in tandem. Specifically, when a trend of dimensional deviation exceeding the upper deviation threshold of 5–15 μm is detected, the current thermal load level is matched based on the absolute value and rate of change of the error, and the impact of each adjustment parameter is weighed according to a dynamic correction coefficient. Based on this, according to a preset adaptive adjustment rule, the feed rate is gradually reduced by 1%–8% of its original value. If continuous adjustment of the feed rate still cannot suppress thermal deformation, the spindle speed is further reduced synchronously by 1%–5% until the temperature increase trend in the cutting area weakens, causing the dimensional deviation to gradually fall back within the tolerance range. Conversely, when the machining dimension is detected to be too small or there are signs of undercut due to tool wear, such as a deviation below the lower deviation threshold of 3-10 μm, the machining power parameters are increased based on the current tool wear compensation requirements and the degree of deviation. The feed rate is increased by 1% to 10% of the original value, increasing the cutting amount per unit time. If the increase in feed rate is still insufficient to offset the undercut caused by wear, the spindle speed is further increased, gradually raising the cutting speed by 1% to 6% of the original value. This improves the depth of cut and cutting stability, restoring the machining dimension to within the target contour range. The entire adjustment process uses real-time closed-loop control, automatically calculating the adjustment range for the next cycle based on each change in dimension feedback. This avoids excessive adjustment that could lead to machining vibration or surface roughness deterioration, allowing the feed rate and spindle speed to gradually stabilize under the constraints of dynamic correction coefficients. Ultimately, this ensures that the machining dimension error remains stable within the set dimension tolerance threshold. Through multiple rounds of real-time iterative control, the machining error is gradually reduced, bringing the dimension error close to zero and ultimately stabilizing within the preset dimension tolerance threshold.This process can run continuously throughout the entire machining process, enabling online correction and dynamic compensation of machining accuracy. It not only improves the consistency and repeatability of hardware parts dimensions, but also effectively reduces rework rate and energy consumption, ensuring high-precision adaptive machining performance.

[0028] Furthermore, the method includes: Based on the difference between the target compensation amount sequence and the real-time measurement feedback value, the tool radius compensation adjustment amount ∆C is determined, and the compensation feed parameter is dynamically adjusted using a fuzzy control algorithm. According to the tool radius compensation adjustment amount ∆C, combined with the dimensional deviation value and the dynamic correction coefficient, the gradient change path of the tool radius compensation value is configured.

[0029] Optionally, during the machining process, to achieve real-time correction of the tool trajectory, the tool radius compensation adjustment amount ΔC is first calculated based on the deviation between the target compensation amount sequence and the real-time measurement feedback value. Specifically, during each segment of the tool machining path, the actual dimension feedback value of the current machining contour is obtained through an online measurement device, such as a laser probe, contact probe, or cutting displacement sensor. This actual dimension feedback value is compared with the target compensation amount of the corresponding machining segment to obtain the instantaneous error deviation. Subsequently, the tool radius compensation adjustment amount ΔC is calculated using a proportional-integral (PI) control algorithm or an equivalent incremental control method, reflecting the magnitude of the current tool compensation adjustment. The obtained tool radius compensation adjustment amount ΔC is then used as an input variable, and the compensation feed parameter is dynamically adjusted using a fuzzy control algorithm. The fuzzy controller sets the corresponding membership function based on the error amount (deviation magnitude) and the error change rate (deviation change trend), and outputs the adjustment level through a rule base to achieve smooth and gradual compensation control. For example, when the error suddenly increases, the fuzzy controller will output a larger feed correction level to make the compensation action respond quickly; while when the error change tends to be stable, a smaller adjustment level is output to avoid overcompensation. Through fuzzy control, complex noise and nonlinear machining behavior can be adaptively processed, making the compensation process more stable. Then, based on the calculated tool radius compensation adjustment amount ΔC, and combined with the previously obtained dimensional deviation value and dynamic correction coefficient, a gradient change path for the tool radius compensation value is constructed. The gradient change path uses the machining segment as a time series, employing a continuous smoothing method to adjust the compensation value segment by segment. The compensation change amplitude is determined based on the magnitude of the dimensional deviation value, and the compensation response speed is determined based on the dynamic correction coefficient. If the dimensional deviation value is large or the environmental interference is strong, the slope of the gradient change is increased, allowing the tool radius compensation value to approach the target more quickly; conversely, the slope of the compensation value change is decreased to improve machining stability. For example, spline interpolation or gradient smoothing algorithms are used to form a continuous curve for the compensation values ​​of each machining segment, preventing abrupt changes in the compensation path and avoiding instantaneous deviations in the tool trajectory that could lead to surface defects. The final gradient change path is used to control the tool motion module, ensuring that the tool radius compensation value changes smoothly during machining, thereby gradually eliminating dimensional errors and bringing the machining result closer to the design dimensions and maintaining it within the dimensional tolerance range.

[0030] Furthermore, the formula for calculating the tool radius compensation adjustment amount ∆C is as follows: ∆C= ,in, This is the proportionality coefficient. The integral coefficient is... For the target compensation amount, This represents the current actual compensation amount.

[0031] Optionally, to achieve precise control of the tool compensation parameters, a proportional-integral control algorithm is used to calculate the tool radius compensation adjustment amount ΔC, and the calculation formula is as follows: ΔC = ;in, This is a proportionality coefficient used to adjust the sensitivity of the compensation response; These are integral coefficients used to accumulate historical errors and eliminate steady-state deviations of the system. and Determined by experimental calibration or adaptive algorithms; The target compensation amount represents the desired tool compensation value calculated from the three-dimensional dimensional compensation map. This represents the current actual compensation amount, indicating the compensation value actually executed by the tool in the real-time machining feedback. When the error changes drastically or the response is sluggish during machining, k is automatically increased. p To enhance response speed, when machining errors tend to stabilize but residual deviations exist, kᵢ is gradually increased to eliminate steady-state errors. Through this control formula and dynamic adjustment mechanism, the tool compensation adjustment process achieves comprehensive control of rapid response, error accumulation correction, and steady-state accuracy maintenance. Ultimately, the actual tool trajectory gradually converges with the target machining trajectory, causing the machining dimensional error to continuously decrease and be stably controlled within the dimensional tolerance threshold range, thereby significantly improving the machining consistency and dimensional accuracy of hardware parts.

[0032] Furthermore, the circular hole positioning diameter of the CNC machine tool worktable is collected, and the circular hole positioning diameter is used for correlation mapping to extract the temperature drift field distribution characteristics of the machine tool; based on the temperature drift field distribution characteristics of the machine tool, the response threshold of the CNC machining component is dynamically adjusted in combination with machining stability, and the gradient change path under the tool radius compensation value is unidirectionally reduced and corrected.

[0033] Optionally, during the machining process, to accurately assess the thermal drift state of the machine tool structure, the circular hole features used for positioning on the CNC machine tool's worktable are first collected. The circular hole on the worktable is a fixed structure, and its actual diameter remains stable under normal conditions, thus serving as an important reference for judging the thermal expansion and thermal drift of the machine tool structure. The diameter of this circular hole is periodically sampled using a contact probe, laser probe, or vision measurement module to obtain the measured values ​​of the positioning hole diameter at different times. Subsequently, the collected actual hole diameter is compared with the theoretical nominal hole diameter to calculate the hole diameter deviation. Because temperature changes during long-term machine tool operation cause thermal expansion of the machine bed, worktable, and other structures, the actual hole diameter and coordinate position of the positioning hole change. This change is spatially correlated and mapped; that is, based on the absolute coordinates of the circular hole on the worktable, the hole diameter change is converted into the corresponding local drift characteristics of the machine tool's temperature field. A curve showing the change in hole diameter deviation over time is established using multiple sampling data, forming the machine tool's temperature drift field distribution characteristics, including the temperature drift direction, drift rate, and regional drift gradient. After obtaining the temperature drift field distribution characteristics of the machine tool, these characteristics are analyzed in conjunction with the current machining stability. If the temperature drift field changes drastically or the temperature drift gradient increases, it indicates that the machine tool structure is in an unstable thermal state. In this case, the response threshold of the CNC machining components is automatically increased to make the compensation more sensitive, so that compensation can be made in time at the initial stage of error generation. If the temperature drift field changes gently, the response threshold is decreased to make the compensation action more stable and avoid over-adjustment. Based on the updated response threshold, the gradient change path of the tool radius compensation value is unidirectionally reduced and corrected. That is, in the tool compensation path, when the machine tool temperature drift shows an expanding trend, the compensation value is gradually reduced by decreasing the compensation intensity or slowing down the compensation change slope to offset the dimensional increase error caused by the thermal expansion of the machine tool. If the temperature drift weakens, the compensation value is kept stable and does not increase, and machining oscillation is avoided through a decreasing adjustment. The specific correction method is to multiply the drift correction coefficient by the original compensation gradient path. This drift correction coefficient is obtained by weighted inversion of the temperature drift amplitude normalization index, the temperature drift gradient normalization index, and the temperature drift direction consistency index. When machine tool temperature drift increases, the drift correction coefficient gradually decreases, and the compensation path corrects in a unidirectional decreasing manner; when temperature drift moderates, the drift correction coefficient is slightly increased. By applying this unidirectional decreasing correction strategy, the tool compensation value can dynamically adapt to changes in machine tool temperature drift, allowing the tool to gradually offset the error accumulation caused by the thermal deformation of the machine bed during machining, thereby significantly improving the stability of machining dimensions and the thermal robustness of the machining process.

[0034] Furthermore, the parameters of the waist-shaped positioning groove of the CNC machine tool worktable are collected, and the angle between the major axis length, minor axis length, and major axis direction of the waist-shaped positioning groove parameters and the key dimension direction of the hardware part to be processed is decoupled by parameters to establish a dimension compensation distribution mapping relationship. Based on the dimension compensation distribution mapping relationship, the influence coefficient of the waist-shaped positioning groove parameters on the compensation amount distribution is fitted to obtain the segmented compensation coupling sequence of multiple processing segments corresponding to the CNC machining component.

[0035] Optionally, during the machining process, to further identify the thermal drift and mechanical offset characteristics of the machine tool structure in different directions, a directional waist-shaped positioning groove on the CNC machine tool table is used as a reference feature for detection and compensation mapping. The waist-shaped positioning groove is typically composed of a major axis and a minor axis, exhibiting significant directional variation characteristics, which can be used to assist in judging the deformation trend of the machine tool in different directions. First, using a 3D measurement probe, vision measurement device, or laser ranging module, the key geometric parameters of the waist-shaped positioning groove on the table are periodically collected, including the length of the major axis, the length of the minor axis, and the orientation angle of the major axis in the table coordinate system. Since the thermal expansion or stress of the machine tool structure during machining causes slight deformation on the table surface, these deformations directly affect the major axis, minor axis, and orientation angle of the waist-shaped groove. Therefore, this parameter sequence can characterize the dynamic error distribution of the machine tool in different directions. Subsequently, the collected major axis length and minor axis length are parametrically decoupled from the key dimensional directions of the metal part to be machined (such as hole orientation, principal normal direction of the curved surface, and key edge orientation). Specifically, the angle θ between the direction vector of the waist-shaped positioning groove and the direction vector of the critical dimension of the part is calculated. Then, the changes in the major axis ΔL1 and minor axis ΔL2 are projected onto this critical direction to form direction-dependent drift response quantities. Through this parameter decoupling, the correspondence between the machine tool's directional drift and the critical dimension error of the part can be obtained. After decoupling, a dimension compensation distribution mapping relationship is established. This mapping model uses multiple linear fitting, weighted least squares, or nonlinear fitting based on neural networks to establish a correlation between the changes in the waist-shaped positioning groove parameters (ΔL1, ΔL2, θ) and the changes in compensation quantities in the three-dimensional dimension compensation map. For example, when the angle between the major axis direction and the critical dimension direction of the part is small, the impact of the machine tool's directional drift on this machining section is more significant, and its weight is automatically increased in the mapping relationship; conversely, its weight is decreased. Through continuous sampling and model training, a complete positioning groove parameter-compensation quantity change mapping function is formed. Finally, the influence coefficients corresponding to the waist-shaped positioning groove parameters are calculated based on the positioning groove parameter-compensation amount change mapping function, and these influence coefficients are superimposed on the original compensation gradient path to redistribute the compensation amounts for each machining segment of the CNC machining component. By performing piecewise fitting on each machining segment, a piecewise compensation coupling sequence reflecting the variation law of machine tool directional error is generated. This piecewise compensation coupling sequence can describe the changes in compensation amount under the influence of machine tool offset in different directions, making the compensation strategy more directional, adaptive, and refined. Through the above process, not only can the waist-shaped positioning groove reflect the directional drift characteristics of the machine tool body, but these drift factors can also be mapped to the key machining directions of the parts, thereby improving the accuracy of the machining compensation strategy, further reducing machining dimensional errors, and significantly improving part dimensional consistency and process stability.

[0036] In summary, the embodiments of this application have at least the following technical effects: First, based on the hardware parts to be processed, the structural basic features, including surface roughness distribution and geometric dimensional deviations, are obtained. Using material elastic modulus data as constraints, a three-dimensional dimensional compensation map is formulated. Then, based on the three-dimensional dimensional compensation map, the target compensation sequence for each tool path in the CNC machining component is matched, and the dimensional deviation value during machining is predicted simultaneously. Next, machining environment interference factors, including machine tool bed temperature drift curve, tool wear rate, and cutting force fluctuation spectrum, are introduced and input into a pre-trained error compensation model to generate dynamic correction coefficients. Finally, based on the target compensation sequence, the dimensional deviation value, and the dynamic correction coefficients, an adaptive control strategy is used to dynamically adjust the tool radius compensation value, feed rate, and spindle speed of the CNC machining component until the machining dimensional error stabilizes within the dimensional tolerance threshold. This solves the technical problem in traditional hardware parts machining where static compensation strategies are difficult to adapt to dynamic machining environments, leading to unstable machining dimensional accuracy. It achieves the technical effect of generating optimal charging and discharging strategies through adaptive optimization coupled with a data model, improving the economy of energy storage utilization and equipment lifespan.

[0037] Example 2 is based on the same inventive concept as the method for compensating for machining errors in the hardware parts dimensions described in the previous examples, such as... Figure 2 As shown, this application provides a dimensional compensation platform for hardware parts to address machining errors. The platform includes: Atlas formulation module 11: Based on the hardware parts to be processed, obtain the structural basic features including surface roughness distribution and geometric dimension deviation, and formulate a three-dimensional dimension compensation atlas using material elastic modulus data as constraints; Deviation prediction module 12: Based on the three-dimensional dimension compensation atlas, match the target compensation amount sequence of each tool path in the CNC machining component, and simultaneously predict the dimension deviation value during the machining process; Environmental compensation module 13: Introduce machining environment interference factors including machine tool bed temperature drift curve, tool wear rate and cutting force fluctuation spectrum, and input them into the pre-trained error compensation model to generate dynamic correction coefficients; Dynamic adjustment module 14: Based on the target compensation amount sequence, the dimension deviation value, and the dynamic correction coefficients, dynamically adjust the tool radius compensation value, feed rate and spindle speed of the CNC machining component using an adaptive control strategy until the machining dimension error stabilizes within the dimension tolerance threshold.

[0038] Furthermore, the map estimation module 11 is used to perform the following method: Obtain the basic structural features of the hardware parts to be processed, and determine the elastic modulus data of the materials using a material database; fuse the basic structural features with the elastic modulus data of the materials to construct the three-dimensional size compensation map, which contains at least three compensation gradient partitions.

[0039] Furthermore, the deviation prediction module 12 is used to perform the following method: Based on the various compensation gradient partitions of the three-dimensional size compensation map, the first toolpath is divided into multiple machining segments corresponding to each compensation gradient partition; for the multiple machining segments, the target compensation amount is configured for the multiple machining segments in combination with the spatial curvature and the compensation density requirements of the compensation gradient partition, and the target compensation amount sequence is obtained.

[0040] Furthermore, the deviation prediction module 12 is used to perform the following method: With machining accuracy uniformity, tool load balance, and energy efficiency as optimization objectives, the local curvature of surface roughness distribution is introduced as a weight correction factor to optimize the configuration of the target compensation amount. Simultaneously, based on historical machining data and referring to the three-dimensional dimensional compensation map, a thermo-mechanical coupling analysis model is constructed. The thermo-mechanical coupling analysis model is used to simulate the temperature and stress field distribution under machining conditions, quantify the dimensional deviation value, and determine whether the compensation amount enhancement mechanism is triggered. The compensation amount enhancement mechanism is used to dynamically correct the target compensation amount sequence.

[0041] Furthermore, the environmental compensation module 13 is used to perform the following method: Data noise reduction preprocessing is performed on the machine tool bed temperature drift curve, tool wear rate, and cutting force fluctuation spectrum among the interference factors in the machining environment to obtain preprocessed data; based on the preprocessed data, an error compensation model under the Transformer architecture is used to generate dynamic correction coefficients.

[0042] Furthermore, the dynamic adjustment module 14 is used to perform the following method: Based on the difference between the target compensation amount sequence and the real-time measurement feedback value, the tool radius compensation adjustment amount ∆C is determined, and the compensation feed parameter is dynamically adjusted using a fuzzy control algorithm. According to the tool radius compensation adjustment amount ∆C, combined with the dimensional deviation value and the dynamic correction coefficient, the gradient change path of the tool radius compensation value is configured.

[0043] Furthermore, the dynamic adjustment module 14 is used to perform the following method: The formula for calculating the tool radius compensation adjustment amount ∆C is as follows: ∆C= ,in, This is the proportionality coefficient. The integral coefficient is... For the target compensation amount, This represents the current actual compensation amount.

[0044] Furthermore, the dynamic adjustment module 14 is used to perform the following method: The circular hole positioning diameter of the CNC machine tool worktable is collected, and the circular hole positioning diameter is used for correlation mapping to extract the temperature drift field distribution characteristics of the machine tool. Based on the temperature drift field distribution characteristics of the machine tool, the response threshold of the CNC machining component is dynamically adjusted in combination with machining stability, and the gradient change path under the tool radius compensation value is unidirectionally reduced and corrected.

[0045] Furthermore, the dynamic adjustment module 14 is used to perform the following method: The parameters of the waist-shaped positioning groove of the CNC machine tool worktable are collected. The angle between the major axis length, minor axis length and major axis direction of the waist-shaped positioning groove parameters and the key dimension direction of the hardware part to be processed is decoupled by parameters to establish a dimension compensation distribution mapping relationship. According to the dimension compensation distribution mapping relationship, the influence coefficient of the waist-shaped positioning groove parameters on the compensation amount distribution is fitted to obtain the segmented compensation coupling sequence of multiple processing segments corresponding to the CNC machining component.

[0046] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for compensating for machining errors in the dimensions of hardware parts, characterized in that, The method comprises: Based on the hardware parts to be processed, the structural basic features including surface roughness distribution and geometric size deviation are obtained, and a three-dimensional size compensation atlas is prepared with material elastic modulus data as a constraint condition; Based on the three-dimensional size compensation atlas, the target compensation amount sequence of each tool path in the numerical control machining assembly is matched, and the size deviation value in the machining process is predicted synchronously; The machining environment interference factors including the machine tool bed temperature drift curve, tool wear rate and cutting force fluctuation spectrum are introduced and input into a pre-trained error compensation model to generate a dynamic correction coefficient; Based on the target compensation amount sequence, the size deviation value and the dynamic correction coefficient, the tool radius compensation value, the feed rate and the spindle speed of the numerical control machining assembly are dynamically adjusted by using an adaptive control strategy until the machining size error is stabilized within the size tolerance threshold.

2. The method for machining error oriented hardware part size compensation as claimed in claim 1, wherein, With material elastic modulus data as a constraint condition, a three-dimensional size compensation atlas is prepared, and the method comprises: The structural basic features of the hardware parts to be processed are obtained, and the material elastic modulus data is determined by using a material database; The structural basic features and the material elastic modulus data are fused to construct the three-dimensional size compensation atlas, and the three-dimensional size compensation atlas contains at least three compensation gradient partitions.

3. The method for machining error oriented hardware part size compensation as claimed in claim 2, wherein, Based on the three-dimensional size compensation atlas, the target compensation amount sequence of each tool path in the numerical control machining assembly is matched, and the method comprises: According to each compensation gradient partition of the three-dimensional size compensation atlas, the first tool path is divided into a plurality of machining sections corresponding to each compensation gradient partition one by one; For the plurality of machining sections, the target compensation amount of the plurality of machining sections is configured in combination with the spatial curvature and the compensation density requirement of the compensation gradient partition to obtain the target compensation amount sequence.

4. The method for machining error oriented hardware part size compensation as claimed in claim 3, wherein, The local curvature of the surface roughness distribution is introduced as a weight correction factor for the optimization and configuration of the target compensation amount, with machining precision uniformity, tool load balance and energy consumption efficiency as optimization targets; At the same time, based on historical machining data, a thermal-mechanical coupling analysis model is constructed by comparing the three-dimensional size compensation atlas; The temperature field and stress field distribution under the machining state are simulated by using the thermal-mechanical coupling analysis model, the size deviation value is quantified, and it is judged whether the compensation amount enhancement mechanism is triggered or not, and the compensation amount enhancement mechanism is used for dynamically modifying the target compensation amount sequence.

5. The method for machining error oriented hardware part size compensation as claimed in claim 4, wherein, And input into a pre-trained error compensation model to generate a dynamic correction coefficient, the method comprises: The machine tool bed temperature drift curve, tool wear rate and cutting force fluctuation spectrum in the machining environment interference factors are subjected to data noise preprocessing to obtain preprocessed data; Based on the preprocessed data, an error compensation model under the Transformer architecture is used to generate a dynamic correction coefficient.

6. The method for machining error oriented hardware part size compensation as claimed in claim 5, wherein, The method comprises: Based on the difference between the target compensation amount sequence and the real-time measurement feedback value, the tool radius compensation adjustment amount ΔC is determined, and the compensation feed parameters are dynamically adjusted by using a fuzzy control algorithm; According to the tool radius compensation adjustment amount ΔC, the gradient change path of the tool radius compensation value is configured in combination with the size deviation value and the dynamic correction coefficient.

7. The method for machining error oriented hardware part size compensation as claimed in claim 6, wherein, The calculation formula of the tool radius compensation adjustment amount ΔC is: AC = Kp * e + Ki * ∑e wherein, Kp is a proportional coefficient, Ki is an integral coefficient, Ctarget is a target compensation amount, Ccurrent is a current actual compensation amount.

8. The method for machining error oriented hardware part size compensation as claimed in claim 6, wherein, Collect the round hole positioning aperture of the worktable of the numerical control machine tool, perform correlation mapping on the round hole positioning aperture, and extract the machine tool thermal drift field distribution characteristics; Based on the machine tool thermal drift field distribution characteristics, the response threshold of the numerical control machining assembly is dynamically adjusted in combination with the machining stability, and the gradient change path under the tool radius compensation value is unidirectionally and gradually decreased.

9. The method for machining error oriented hardware part size compensation as claimed in claim 8, wherein, Collect the waist-shaped positioning groove parameters of the worktable of the numerical control machine tool, parameterize decoupling the long axis length, short axis length and long axis direction of the waist-shaped positioning groove parameters and the included angle between the key dimension direction of the to-be-machined hardware part, and establish a size compensation distribution mapping relationship; According to the size compensation distribution mapping relationship, the influence coefficient of the waist-shaped positioning groove parameters on the compensation amount distribution is fitted, and a segmented compensation coupling sequence of a plurality of machining sections corresponding to the numerical control machining assembly is obtained.

10. A hardware part dimensional compensation platform oriented to machining errors, characterized in that, The platform comprises: A graph mapping module: based on the to-be-machined hardware part, structural basic characteristics including surface roughness distribution and geometric size deviation are obtained, and a three-dimensional size compensation graph is mapped based on the material elastic modulus data as a constraint condition; A deviation prediction module: based on the three-dimensional size compensation graph, the target compensation amount sequence of each tool path in the numerical control machining assembly is matched, and the size deviation value in the machining process is synchronously predicted; An environment compensation module: introduce the machining environment interference factors including the machine tool bed thermal drift curve, the tool wear rate and the cutting force fluctuation spectrum, and input them into the pre-trained error compensation model to generate a dynamic correction coefficient; A dynamic adjustment module: based on the target compensation amount sequence, the size deviation value and the dynamic correction coefficient, the tool radius compensation value, the feed rate and the spindle speed of the numerical control machining assembly are dynamically adjusted by using the adaptive control strategy until the machining size error is stable within the size tolerance threshold.

Images