Data processing method and system for predicting machining accuracy of gear hobbing machine tools

Abstract

The present invention discloses a method and system for processing data for predicting machining accuracy of gear hobbing machines, relating to the field of data processing technology. The method comprises: obtaining a target gear, obtaining a historical machining parameter group, obtaining historical accuracy data, and obtaining an error ratio; forming a parameter group sequence; obtaining a current machining parameter group, obtaining the position of the value of the i-th subdivision parameter of the current machining parameter group in the parameter group sequence corresponding to the i-th subdivision parameter, obtaining the historical machining parameter groups corresponding to the two adjacent values ​​at that position and using them as two reference parameter groups; and obtaining the subdivision error ratio corresponding to the i-th subdivision parameter in the current machining parameter group based on the error ratios corresponding to the two reference parameter groups of the i-th subdivision parameter; and obtaining the prediction error ratio corresponding to the current machining parameter group based on the subdivision error ratios corresponding to each subdivision parameter in the current machining parameter group. The present invention has the advantages of high precision of multiple prediction results, precise capture of sequence interpolation, and multi-parameter quantization.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for processing data for predicting machining accuracy of a gear hobbing machine tool. Background Art

[0002] In the field of gear hobbing, gear machining accuracy (tooth profile dimensional error) directly impacts the performance and life of transmission components. This accuracy is influenced by a combination of dynamic machining parameters (such as feed rate, cutting speed, and material hardness). Currently, the industry generally uses statistical methods based on historical data to predict the machining accuracy of the current batch, thereby optimizing parameter settings and reducing trial cutting costs. However, existing technologies have numerous drawbacks.

[0003] First, existing methods (such as simple linear regression or mean value calculation) fail to systematically address the sequential relationships and relative positions of machining parameters when utilizing historical data. This results in the prediction model being unable to capture the nonlinear effects of small changes in parameter values ​​on accuracy. For example, when feed rate or cutting speed undergoes gradual changes in the historical sequence, existing models simply use the arithmetic mean of the historical accuracy values, ignoring the trends implicit in the order of parameter values, resulting in prediction results that deviate from reality. Second, existing methods do not incorporate error ratios (i.e., the ratio of the absolute error between the actual machined dimension and the designed dimension to the maximum allowable error), but instead directly manipulate the raw error data. This makes it difficult to quantify the differences in the contributions of different parameters to the overall error. For example, a slight increase in cutting speed may have a greater impact on the error than a change in feed rate. However, existing methods do not distinguish these contributions through weighting, resulting in inaccurate predictions. Finally, existing technologies lack an interpolation mechanism based on adjacent parameter groups. That is, they do not select adjacent control parameter groups based on the current parameter's position in the historical sequence for error ratio interpolation calculation, resulting in significant deviations in prediction results at parameter boundaries. These problems lead to insufficient prediction accuracy, which in turn causes machining deviations, excessive tool wear or increased rework rates, seriously restricting the mass production efficiency and cost control of high-precision gears (such as automotive transmission gears). Summary of the Invention

[0004] In view of the defects in the prior art, the present invention provides a method and system for processing data of machining accuracy prediction of gear hobbing machine tools.

[0005] A method for processing machining accuracy prediction data of a gear hobbing machine tool comprises: obtaining a target gear, obtaining multiple historical machining parameter groups corresponding to the target gear, obtaining historical accuracy data of the target gear after machining according to each historical machining parameter group, and obtaining an error ratio corresponding to each historical machining parameter group based on the historical accuracy data corresponding to each historical machining parameter group, wherein each historical machining parameter group includes multiple subdivision parameters; arranging the multiple historical machining parameter groups in ascending order of the values ​​of each subdivision parameter in the historical machining parameter group to form a parameter group sequence corresponding to each subdivision parameter; obtaining a current machining parameter group corresponding to the target gear and including the multiple subdivision parameters, obtaining the position of the value of the i-th subdivision parameter of the current machining parameter group in the parameter group sequence corresponding to the i-th subdivision parameter, obtaining the historical machining parameter groups corresponding to two adjacent values ​​of the position and using them as two comparison parameter groups for the i-th subdivision parameter, and obtaining a subdivision error ratio corresponding to the i-th subdivision parameter in the current machining parameter group based on the error ratios corresponding to the two comparison parameter groups for the i-th subdivision parameter; and obtaining a predicted error ratio corresponding to the current machining parameter group based on the subdivision error ratios corresponding to each subdivision parameter in the current machining parameter group.

[0006] Optionally, obtaining the error ratio corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group includes: obtaining the gear design size of the target gear, and obtaining the historical processing size corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group; obtaining the error ratio based on the gear design size and the historical processing size.

[0007] Optionally, the error ratio is obtained based on the gear design size and historical processing size and expressed as: ;in, is the error ratio corresponding to the jth group of historical processing parameter groups, Design dimensions for the gears, is the historical processing size corresponding to the jth group of historical processing parameter groups, is the maximum error.

[0008] Optionally, obtaining the segmentation error ratio corresponding to the i-th segmentation parameter in the current processing parameter group based on the error ratio corresponding to the two control parameter groups of the i-th segmentation parameter includes: obtaining the difference between the error ratios corresponding to the two control parameter groups of the i-th segmentation parameter and using it as the total error variable, and obtaining the error ratio corresponding to the smaller value of the two control parameter groups of the i-th segmentation parameter; obtaining the difference between the values ​​of the i-th segmentation parameter in the two control parameter groups and using it as the control parameter variable, obtaining the difference between the value of the i-th segmentation parameter in the current processing parameter group and the value of the i-th segmentation parameter in the smaller value control parameter group and using it as the current parameter variable; obtaining the segmentation error ratio corresponding to the i-th segmentation parameter based on the current parameter variable, control parameter variable, total error variable and the error ratio corresponding to the smaller value of the two control parameter groups corresponding to the i-th segmentation parameter.

[0009] Optionally, the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group is obtained according to the error ratios corresponding to the two comparison parameter groups of the i-th subdivision parameter, and is expressed as: ;in, is the subdivision error ratio corresponding to the i-th subdivision parameter, is the error ratio of a control parameter group, is the error ratio corresponding to another control parameter group, is the larger value of the i-th subdivision parameter of the two control parameter groups, is the smaller value of the i-th subdivision parameter in the two control parameter groups, The value of the i-th subdivision parameter of the current processing parameter group.

[0010] Optionally, the method further includes: obtaining a gear design dimension of a target gear, and obtaining prediction accuracy data based on a prediction error ratio.

[0011] Optionally, the prediction error ratio corresponding to the current processing parameter group is obtained according to the subdivision error ratio corresponding to each subdivision parameter in the current processing parameter group, and is expressed as: , ;in, is the prediction error ratio corresponding to the current processing parameter group, The number of types of subdivided parameters in the current processing parameter group. is the contribution weight of the i-th subdivision parameter, is the subdivision error ratio corresponding to the i-th subdivision parameter.

[0012] A gear hobbing machine tool machining accuracy prediction data processing system is also provided, the system includes: a data acquisition module, used to acquire a target gear, and acquire a plurality of historical machining parameter groups corresponding to the target gear, and acquire historical accuracy data after the gear cutting process is completed according to each historical machining parameter group, and acquire the error ratio corresponding to each historical machining parameter group according to the historical accuracy data corresponding to each historical machining parameter group, wherein each historical machining parameter group includes a plurality of subdivided parameters; a sequence generation module, used to arrange the plurality of historical machining parameter groups in order from small to large according to the value of each subdivided parameter in the historical machining parameter group and form a parameter group sequence corresponding to each subdivided parameter; a data acquisition module, used to acquire a target gear, and acquire a plurality of historical machining parameter groups in order from small to large according to the value of each subdivided parameter in the historical machining parameter group and form a parameter group sequence corresponding to each subdivided parameter; a data acquisition module, used to acquire a target gear, and acquire a plurality of historical machining parameter groups A data processing module is used to obtain a current processing parameter group corresponding to a target gear and including multiple subdivision parameters, obtain the position of the value of the i-th subdivision parameter of the current processing parameter group in the parameter group sequence corresponding to the i-th subdivision parameter, obtain the historical processing parameter groups corresponding to the two adjacent values ​​at this position and use them as two comparison parameter groups for the i-th subdivision parameter, and obtain the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group based on the error ratios corresponding to the two comparison parameter groups of the i-th subdivision parameter; a data prediction module is used to obtain the predicted error ratio corresponding to the current processing parameter group based on the subdivision error ratios corresponding to each subdivision parameter in the current processing parameter group.

[0013] Optionally, the data acquisition module is also used to: obtain the gear design dimensions of the target gear, and obtain the historical processing dimensions corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group; obtain the error ratio based on the gear design dimensions and the historical processing dimensions.

[0014] Optionally, the data processing module is also used to: obtain the difference in error ratios corresponding to the two control parameter groups of the i-th subdivision parameter and use it as the total error variable, and obtain the error ratio corresponding to the smaller value of the two control parameter groups of the i-th subdivision parameter; obtain the difference in the values ​​of the i-th subdivision parameter in the two control parameter groups and use it as the control parameter variable, obtain the difference between the value of the i-th subdivision parameter in the current processing parameter group and the value of the i-th subdivision parameter in the smaller value control parameter group and use it as the current parameter variable; obtain the subdivision error ratio corresponding to the i-th subdivision parameter based on the current parameter variable, control parameter variable, total error variable and error ratio corresponding to the smaller value of the two control parameter groups corresponding to the i-th subdivision parameter.

[0015] The beneficial effects of the present invention are embodied in:

[0016] In the entire gear hobbing machine tool processing accuracy prediction data processing method, first, the error ratio mechanism converts the absolute size error into the relative tolerance occupancy rate, eliminates the data incomparability caused by the difference in tolerance bands, realizes the same-dimensional quantitative evaluation of historical and current data, and avoids the misjudgment caused by the confusion of the original error dimension in traditional methods; secondly, the parameter serialization and reorganization independently constructs an ascending historical trajectory for each processing parameter, and converts the implicit trend law (such as the nonlinear transition of the error caused by the increase in material hardness) into a computable topological relationship, overcoming the defect of the statistical averaging method that ignores the parameter gradual change process; further, based on the current The parameters are precisely positioned in the sequence, and only the two most adjacent historical parameter groups are selected as reference points. The proportion of subdivision errors caused by the independent influence of a single parameter is deduced through interpolation calculation, and the accuracy mutation characteristics of the parameter boundary area (such as the error steepening effect of the feed rate near the critical value) are captured at the micro level, so that the prediction model has the ability to handle nonlinear responses; finally, dynamic weight aggregation distinguishes the differential contributions of core parameters and secondary parameters to the error based on process knowledge, and amplifies the sensitivity of the core parameters by suppressing the interference of secondary parameters through unequal weighting, solving the prediction distortion problem caused by the coupling of multiple parameters at the system level. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.

[0018] Figure 1 Schematic diagram of the steps of a method for processing data for predicting machining accuracy of a gear hobbing machine tool according to one embodiment of the present invention;

[0019] Figure 2 Schematic diagram of part of the steps S1 in the data processing method for predicting machining accuracy of a gear hobbing machine tool according to the present invention;

[0020] Figure 3 Schematic diagram of a portion of step S3 in the method for processing data for predicting machining accuracy of a gear hobbing machine tool according to the present invention;

[0021] Figure 4 Schematic diagram of the steps of another embodiment of the method for processing data for predicting machining accuracy of a gear hobbing machine according to the present invention. DETAILED DESCRIPTION

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0023] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather merely represents selected embodiments of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are intended to fall within the scope of protection of the present invention.

[0024] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. In addition, the terms "first," "second," etc. are used only to distinguish the descriptions and are not to be understood as indicating or implying relative importance.

[0025] like Figure 1 As shown, a method for processing data of machining accuracy prediction of a gear hobbing machine tool is provided, comprising:

[0026] S1. Obtain a target gear, obtain multiple historical processing parameter groups corresponding to the target gear, obtain historical accuracy data after completing cutting processing of the target gear according to each historical processing parameter group, and obtain an error ratio corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group, wherein each historical processing parameter group includes multiple subdivided parameters;

[0027] S2. Arrange the multiple historical processing parameter groups in ascending order of the values ​​of each subdivided parameter in the historical processing parameter group to form a parameter group sequence corresponding to each subdivided parameter;

[0028] S3. Obtain a current processing parameter group corresponding to the target gear and including multiple subdivision parameters, obtain the position of the value of the i-th subdivision parameter of the current processing parameter group in the parameter group sequence corresponding to the i-th subdivision parameter, obtain the historical processing parameter groups corresponding to the two adjacent values ​​at the position and use them as two comparison parameter groups for the i-th subdivision parameter, and obtain the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group based on the error ratios corresponding to the two comparison parameter groups for the i-th subdivision parameter;

[0029] S4. Obtain the prediction error ratio corresponding to the current processing parameter group according to the subdivision error ratio corresponding to each subdivision parameter in the current processing parameter group.

[0030] In this embodiment, it should be noted that step S1 constitutes the data foundation phase of the entire prediction method, with the core objective being to organize and transform historical processing information related to the target gear. Specifically, it first requires identifying the type of "target gear" and collecting "historical processing parameter sets" representing different combinations of processing conditions accumulated during past production of that type of gear. Each of these parameter sets contains several key, dynamically controllable processing settings (also known as detailed parameters, such as feed rate, spindle cutting speed, and the measured hardness of the gear blank material). Furthermore, for each actual finished gear processed using these specific parameter combinations, it is necessary to obtain its key "historical accuracy data"—the absolute value of the dimensional error between the actual dimensions of the machined gear (e.g., tooth thickness, tooth profile) and the standard dimensions specified in the design drawings. To more scientifically and comparably characterize the significance and impact of these errors, S1 introduces an error ratio. This step calculates the corresponding error ratio for each set of historical machining parameters. This error ratio essentially divides the absolute value of the actual dimensional deviation generated during the historical machining by the maximum dimensional deviation tolerance (i.e., the width of the tolerance band) allowed by the design for the target gear type, resulting in a standardized, dimensionless ratio. This ratio clearly expresses "what proportion of the allowable deviation range the actual machining deviation occupies." The closer the ratio is to 1 or exceeds 1, the closer the error is to or has exceeded the tolerance limit.

[0031] For example, let's assume the target gear is a certain model of automotive transmission gear, with a critical tooth thickness tolerance of ±0.01mm (i.e., a maximum allowable error ΔS = 0.01mm). S1 retrieves the production records for the past 20 batches of this gear from the production database. Each record represents a historical machining parameter set (e.g., feed rate = 0.15mm / rev, cutting speed = 90m / min, material hardness = 220HBW), and is associated with the difference between the actual tooth thickness measured for that batch of gear samples and the standard tooth thickness (e.g., a deviation of +0.005mm results in an absolute error of 0.005mm). Next, for each historical parameter set, the corresponding error ratio is calculated: the "absolute historical dimensional error" (0.005mm) is divided by the "maximum allowable error" (0.01mm), resulting in a ratio (0.5). This 0.5 indicates that the dimensional error produced by that particular parameter combination accounted for 50% of the allowable error range, still within the safe range but with some margin. Through S1, all historical processing data (different parameter combinations and their results) are systematically converted into this "proportional" form that represents the severity of the error, which prepares for the subsequent steps to use this historical experience to predict the accuracy risk under new parameter combinations.

[0032] In S2, the core purpose of step S2 is to reorganize historical machining data according to parameter variation patterns, constructing a structured sequence that reflects the continuous parameter variation trends. For each independent sub-parameter (such as feed rate, cutting speed, material hardness, etc.), this step arranges all historical machining parameter groups compiled in step S1 in strictly ascending order according to their values, generating a parameter group sequence specific to that parameter. The specific operation is: select a sub-parameter (such as cutting speed), traverse all historical parameter groups, extract the parameter values, and sort them from small to large, while retaining the complete parameter combination corresponding to each parameter group and its associated error ratio. This forms multiple parallel longitudinal sequences (for example, in the "cutting speed sequence," parameter groups are arranged from low to high speed values; in the "feed rate sequence," parameter groups are arranged from small to large feed rates). The relative position of the parameter groups within each sequence directly reflects the numerical progression of the sub-parameter.

[0033] For example, consider material hardness parameters. Suppose there are five parameter groups in the historical database, with hardness values ​​of 200 HBW, 210 HBW, 220 HBW, 230 HBW, and 240 HBW, respectively. S2 generates independent sequences for these hardness parameters, sorting them in ascending order of hardness value: group with hardness = 200 HBW → group with hardness = 210 HBW → group with hardness = 220 HBW → group with hardness = 230 HBW → group with hardness = 240 HBW. Each row in the sequence still contains the original values ​​of other parameters, such as feed rate and cutting speed, along with their error ratios. By constructing this sequence for each sub-parameter, the gradual change of parameter values ​​(e.g., the gradual increase in hardness from 200 HBW to 240 HBW) is made explicit, and the relative positional relationships between parameters are quantified and recorded. This lays the structural foundation for the subsequent steps of trend interpolation based on parameter proximity (such as locating the influence of 215HBW between hardness 210HBW and 220HBW), thus overcoming the defect of traditional methods that ignore the implicit rules of parameter sorting.

[0034] In S3, the goal is to leverage the local relationships within the historical parameter sequence to infer the independent impact of each sub-parameter on precision within the current machining parameter combination. This step first obtains the current machining parameter group containing all sub-parameters to be optimized (such as feed rate and cutting speed). Then, for each sub-parameter, the following operations are performed: Within the parameter-specific ascending sequence established in S2 (e.g., the cutting speed sequence), the current parameter value is precisely located (e.g., the current speed value is between two adjacent historical speed values ​​in the sequence). Next, two adjacent historical parameter groups are extracted as reference parameter groups (e.g., historical group A with a slightly lower speed and historical group B with a slightly higher speed; the current speed value can overlap with the historical speed value. In this case, the overlapping historical group is first extracted as one reference parameter group, and then an adjacent historical group with a higher speed or a lower speed is extracted as the other reference parameter group. It should be noted that only when two adjacent values ​​(including overlaps) exist at the current parameter value's numerical position can two reference parameter groups be obtained). These two groups are adjacent only to the current parameter in terms of that parameter value; other parameters may be completely different. Finally, based on the historical values ​​of the error ratio of the two control groups (i.e., the proportion of standardized accuracy deviation calculated by S1), the accuracy impact value that the current parameter value may have when considering the subdivision parameter in isolation is derived through nonlinear interpolation calculation that reflects the parameter proximity relationship - this is called the subdivision error ratio.

[0035] Taking feed rate parameters as an example, let's assume the historical feed rate sequence contains three parameter groups in ascending order (corresponding to feed rates of 0.10mm / rev, 0.15mm / rev, and 0.20mm / rev), and the current machining feed rate is set to 0.18mm / rev. S3 first locates 0.18mm / rev in the sequence—between 0.15mm / rev and 0.20mm / rev—and selects these two groups as control parameter groups. Assume the historical error ratio for the 0.15mm / rev group is 0.4 (i.e., the error accounts for 40% of the tolerance), and the historical error ratio for the 0.20mm / rev group is 0.8 (80% of the tolerance). Based on the distance between 0.18mm / rev and the two control values ​​(closer to 0.20mm / rev) and the difference in historical error ratios (0.8 - 0.4 = 0.4), the system automatically calculates the subdivision error ratio for the current feed rate alone—a value higher than 0.4 but lower than 0.8. This calculation assumes a core logic: the accuracy shift caused by a single parameter change exhibits a continuous transition within the adjacent parameter range. This allows each sub-parameter to obtain an independent impact estimate based on its own sequence, free from interference from other parameters, laying the foundation for subsequent comprehensive weighting.

[0036] In S4, the core task is to comprehensively consider the differentiated process impacts of each machining parameter on precision and generate an overall precision risk prediction value for the current parameter combination. This step is based on the error ratios of all subdivisions obtained in S3 (i.e., the contribution of each parameter to dimensional deviation under "isolated ideal conditions") and introduces a weighted aggregation mechanism for process knowledge to derive the predicted error ratio for the entire parameter combination. Specifically, the system pre-assigns contribution weights to different subdivisions based on engineering experience or data analysis (for example, cutting speed is assigned a weight of 0.6 due to its sensitivity to thermal deformation, feed rate is assigned a weight of 0.3 due to its lesser impact on force deformation, and material hardness is assigned a weight of 0.1). The total weight is constant at 1. A weighted summation calculation is then performed: the subdivision error ratio of each parameter is multiplied by its weight, and the weighted results of all parameters are added together. The final output value, the predicted error ratio, represents the proportion of the expected dimensional error within the tolerance range for the current parameter combination and directly reflects the risk of machining out-of-tolerance.

[0037] Taking automotive transmission gear machining as an example, let's assume that S3 calculates the subdivision error ratio for the current parameters as 0.7 for cutting speed, 0.4 for feed rate, and 0.5 for material hardness. Based on the process characteristics of this gear model, the speed weighting is set at 60% (primary factor), 30% for feed rate (secondary factor), and 10% for hardness (minor factor). S4 is calculated as: (cutting speed effect 0.7 × 60% = 0.42) + (feed effect 0.4 × 30% = 0.12) + (hardness effect 0.5 × 10% = 0.05) = a predicted error ratio of 0.59. This value indicates that, under the current combination, the expected dimensional deviation will account for 59% of the tolerance, which is within tolerance but close to the safety margin (>0.6 requires a warning). This mechanism breaks through the defect of traditional methods that ignore the differences in parameter influence. For example, when the cutting speed increases from 90m / min to 95m / min, its weight amplification effect will make the predicted value more sensitively reflect the actual risk jump, rather than being diluted by secondary parameters such as feed rate.

[0038] In summary, in the entire gear hobbing machine tool processing accuracy prediction data processing method, first, the error ratio mechanism converts the absolute size error into the relative tolerance occupancy rate, eliminates the data incomparability caused by the difference in tolerance bands, realizes the same-dimensional quantitative evaluation of historical and current data, and avoids the misjudgment caused by the confusion of the original error dimension in the traditional method; secondly, the parameter serialization and reorganization independently constructs an ascending historical trajectory for each processing parameter, and converts the implicit trend law (such as the nonlinear transition of the error caused by the increase in material hardness) into a computable topological relationship, which overcomes the defect of the statistical averaging method that ignores the parameter gradual change process; further, based on In order to accurately position the current parameter in the sequence, only the two most adjacent historical parameter groups are selected as reference points. The subdivision error ratio of the independent influence of a single parameter is deduced through interpolation calculation, and the accuracy mutation characteristics of the parameter boundary area (such as the error steepening effect of the feed rate near the critical value) are captured at the micro level, so that the prediction model has the ability to handle nonlinear responses; finally, dynamic weight aggregation distinguishes the differential contributions of core parameters and secondary parameters to the error based on process knowledge, and amplifies the sensitivity of the core parameters by suppressing the interference of secondary parameters through unequal weighting, solving the prediction distortion problem caused by the coupling of multiple parameters at the system level.

[0039] like Figure 2 As shown, in one embodiment, obtaining the error ratio corresponding to each historical processing parameter group according to the historical accuracy data corresponding to each historical processing parameter group in S1 includes:

[0040] S11, obtaining the gear design dimensions of the target gear, and obtaining the historical processing dimensions corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group;

[0041] S12. Obtain an error ratio based on the gear design dimensions and historical processing dimensions.

[0042] In this embodiment, it should be noted that in S11, a precise mapping relationship between historical machining data and design standards is established, providing the original calculation basis for error quantification. This step first determines the theoretical design dimensions of the target gear (i.e., the standard values ​​for key dimensions such as tooth thickness and tooth height specified in the drawings). Then, for each record of a historical machining parameter set, the corresponding actual machining dimensions are extracted. This data is derived from the physical inspection report of the finished gear after the parameter combination is actually machined (e.g., the actual tooth thickness measured by a coordinate measuring machine). This step essentially establishes a physical correspondence between "parameter combination input" and "machining result output": it precisely associates an abstract machining parameter (e.g., a feed rate of 0.15 mm / rev) with a concrete dimensional result (e.g., a measured tooth thickness deviation of +0.005 mm from the standard value). For example, if a batch of gearbox gears machined using a specific cutting speed and feed rate is found to have an actual tooth thickness of a certain value during quality inspection, the absolute difference between this value and the designed dimension becomes the core output of this step, providing irreplaceable basic data for subsequent error standardization.

[0043] In S12, absolute dimensional errors are converted into process-interpretable risk indicators, enabling unified quantification of the impact of errors across parameter combinations. This step uses the design and actual machining dimensions obtained in S11 to calculate the absolute dimensional deviation (|actual value - design value|) for each historical parameter group. This deviation is then divided by the maximum allowable tolerance for the gear type (for example, ±0.01mm means a tolerance band width of 0.02mm) to generate a dimensionless error ratio. This ratio overcomes the limitations of traditional methods: firstly, it uniformly converts dimensional errors of different magnitudes (such as 0.005mm and 0.008mm) into tolerance range occupancy (e.g., 50% vs. 80%), eliminating dimensional interference; secondly, it directly reflects the distance between the machining result and the scrap boundary (a ratio ≥100% indicates an out-of-tolerance condition), thus enabling historical data to serve as a risk warning. For example, when there are two sets of processing records with the same material hardness parameters, one set has an error ratio of 0.3 (safe) and the other set has an error ratio of 0.9 (on the verge of out of tolerance), even if the absolute error amounts are similar (such as 0.006mm vs. 0.018mm), this indicator can clearly characterize the substantial risk differences under the current tolerance system, providing an intuitive basis for subsequent parameter optimization.

[0044] In one embodiment, the error ratio obtained in S12 based on the gear design dimensions and the historical processing dimensions is expressed as:

[0045] ;in,

[0046] is the error ratio corresponding to the jth group of historical processing parameter groups, Design dimensions for the gears, is the historical processing size corresponding to the jth group of historical processing parameter groups, is the maximum error.

[0047] In this embodiment, it should be noted that Absolute deviation calculation eliminates the directional influence of positive and negative deviations. Traditional methods directly use raw errors, where positive deviations (oversize) and negative deviations (undersize) cancel each other out (e.g., when calculating the average), leading to an underestimation of the true error. For example, if one set of parameters results in a tooth thickness that is 0.008mm larger (error +0.008mm) and another set that is 0.006mm smaller (error 0.006mm), the raw error average is 0.001mm—seriously masking the actual risk. Absolute value processing ensures that all offsets are cumulatively added in a positive direction, avoiding directional interference.

[0048] Further, For normalization, the physical size error is converted into a dimensionless risk ratio, so that the data of gears with different tolerance requirements can be compared horizontally. This solves the dimensional confusion. For example, the 0.005mm error caused by the feed rate change is mm gears account for 25%, and The risk levels of the gears are completely different. The quantified risk level is also realized. When the ratio value is ≥100%, an out-of-tolerance warning is triggered, achieving a unified decision-making standard across working conditions. (Tolerance zone width), The essence is the design tolerance range. For example, when the tolerance is ±0.01mm, the actual allowable total deviation range is 0.01mm (i.e. from 5mm to 4.990mm or 5.010mm). hour, Directly judge the deviation, when hour, Safety.

[0049] like Figure 3 As shown, in one embodiment, obtaining the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group according to the error ratios corresponding to the two comparison parameter groups of the i-th subdivision parameter in S3 includes:

[0050] S31, obtaining the difference in error ratios corresponding to the two comparison parameter groups of the i-th subdivision parameter and using it as the total error variable, and obtaining the error ratio corresponding to the smaller value of the two comparison parameter groups of the i-th subdivision parameter;

[0051] S32, obtaining the difference between the values ​​of the i-th subdivision parameter in the two comparison parameter groups and using it as the comparison parameter variable, obtaining the difference between the value of the i-th subdivision parameter in the current processing parameter group and the value of the i-th subdivision parameter in the smaller comparison parameter group and using it as the current parameter variable;

[0052] S33. Obtain the segmentation error ratio corresponding to the i-th segmentation parameter according to the current parameter variable, the control parameter variable, the total error variable, and the error ratio corresponding to the smaller value of the two control parameter groups corresponding to the i-th segmentation parameter.

[0053] In this embodiment, it should be noted that in S31, the accuracy fluctuation range of adjacent historical parameter groups is quantified to provide a benchmark for interpolation calculations. This step first selects two control parameter groups for the i-th subdivision parameter (the adjacent historical groups located by S3) and calculates the absolute difference in their error ratios as the total error variable. This value represents the maximum fluctuation range of accuracy within this parameter range. Simultaneously, the larger error ratio of the two control groups is identified, reflecting the upper limit of the accuracy risk for this parameter value range. For example, in a cutting speed sequence, if the error ratios of the control groups are 0.3 and 0.6, the total error variable |0.6-0.3|=0.3 indicates that the accuracy fluctuation range caused by speed changes is 30% of the tolerance, while the larger value of 0.6 indicates that the maximum risk in this range reaches 60% of the tolerance. This step quantifies the extreme range characteristics of the historical trend, providing a basis for dynamic adjustment for subsequent interpolation.

[0054] In S32, a coordinate framework for interpolation is constructed through a mathematical mapping of the relative positional relationships of the parameters. This step performs two difference calculations: the reference parameter variable: the spacing between the values ​​of the i-th parameter in the two reference parameter groups (e.g., the difference between cutting speeds of 90 m / min and 100 m / min is 10 m / min), which serves as the reference interval length; and the current parameter variable: the positional offset of the current parameter value relative to either reference group (e.g., the difference between the current speed of 95 m / min and the low-speed group of 90 m / min is 5 m / min), which serves as the interpolation coordinate point. For example, if the current feed rate of 0.18 mm / rev falls between the historical groups (0.15 mm / rev and 0.20 mm / rev), the reference parameter variable is 0.05 mm / rev (interval length), and the current parameter variable is 0.03 mm / rev (position from the lower limit group). This operation transforms the abstract parameter relationship into a proportional spatial scale (e.g., 0.03 / 0.05 = 60% position), allowing nonlinear effects to be approximated through linear relationships, directly addressing the problem of parameter boundary prediction distortion.

[0055] In S33, based on the intermediate variables from the previous three steps, the accuracy estimate under the independent influence of the current parameter is derived. This step uses the maximum error ratio of the control group as a benchmark. Based on the position of the current parameter in the value range (the offset ratio in S32), the total error variable is proportionally reduced to obtain the subdivided error ratio. In other words, when the current parameter approaches the control group with a larger error ratio, the reduction is small, and the result is close to the upper risk limit; when the current parameter moves away from the high-risk control group, the reduction is large, and the result shifts toward lower risk.

[0056] In one embodiment, in S3, the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group is obtained based on the error ratios corresponding to the two comparison parameter groups of the i-th subdivision parameter, and is expressed as:

[0057] ;in,

[0058] is the subdivision error ratio corresponding to the i-th subdivision parameter, is the error ratio corresponding to a control parameter group, is the error ratio corresponding to another control parameter group, is the larger value of the i-th subdivision parameter of the two control parameter groups, is the smaller value of the i-th subdivision parameter in the two control parameter groups, The value of the i-th subdivision parameter of the current processing parameter group.

[0059] In this embodiment, it should be noted that To normalize the parameter position, quantify the relative position of the current parameter value in the historical sequence interval, and convert the physical quantity (such as 90m / min→100m / min) into a continuous ratio of 0-1. ,when When (such as the cutting speed ascending sequence error ratio 0.3→0.6), the expression is Generate a positive gradient, such as increasing it by 0.3 times , superimpose increments according to position proportion; when When (such as the material hardness ascending sequence error ratio 0.9→0.5), the expression is Generate a negative gradient, such as reducing it by 0.4 times , proportionally reducing the base value. This achieves unified handling of rising / falling trends and avoids pre-set trend direction restrictions.

[0060] Further, Dynamic gradient generation is implemented, constructing a linear response gradient based on the difference in error ratios within the historical control group. In nonlinearly sensitive ranges (e.g., feed rate 0.15 → 0.20 mm / rev, error 0.4 → 0.8), the gradient is +0.4, with a superimposed increment of +0.24 at a position ratio of 0.6. In negative correlation ranges (e.g., coolant flow increases → error decreases), the gradient automatically becomes negative to implement attenuation calculations. Dynamically adapting to any monotonic trend (positive or negative correlation), eliminating the need for manual judgment.

[0061] Further, Implemented a baseline + incremental mechanism with a historical control group As the benchmark, the gradient effect is superimposed according to the displacement.

[0062] In summary, the existing methods ignore the order of parameters and regard the subdivision parameters that do not appear in the historical processing parameter group as discrete points. However, this expression realizes forced recognition. For example, 95m / min is located at the midpoint of the historical data from 90m / min to 100m / min, and the trigger error ratio is estimated to be 0.45 from the midpoint of 0.3 to 0.6. By removing the coupling of other parameters (e.g., ignoring feed rate fluctuations in speed calculations), the weighted aggregation of S4 truly reflects the weights of various parameters. In summary, the expression in this implementation is compatible with both positively and negatively correlated parameters, outputting a "decoupled" subdivision error ratio, and ensuring that the multi-parameter comprehensive prediction (S4) is physically reasonable.

[0063] like Figure 4 As shown, in one embodiment, it also includes:

[0064] S5. Obtain the gear design size of the target gear, and obtain prediction accuracy data based on the prediction error ratio.

[0065] In this embodiment, it should be noted that in S5, the predicted theoretical risk value is converted into physical dimensional instructions that can directly guide production, completing the final closed loop of the prediction model. This step first calls the gear design dimensions of the target gear (i.e., the theoretical standard values ​​determined in S11, such as a tooth thickness of 5.000mm). Then, combined with the predicted error ratio output from S4 (e.g., 0.59), the expected fluctuation range of the processed dimensions is inferred. Specifically, the predicted error ratio is multiplied by the maximum allowable tolerance to calculate the expected absolute dimensional deviation under the current parameter combination (e.g., for a maximum allowable tolerance of 0.02mm, 0.59 × 0.02mm ≈ 0.012mm). Finally, this deviation is superimposed on the baseline design dimensions (e.g., the theoretical tooth thickness value of 5.000mm ± 0.012mm), generating predicted accuracy data that can be directly used for quality inspection and comparison (e.g., the actual measured dimensions are expected to fall within the range of 4.988-5.012mm).

[0066] In one embodiment, in S4, the prediction error ratio corresponding to the current processing parameter group is obtained according to the subdivision error ratio corresponding to each subdivision parameter in the current processing parameter group, which is expressed as:

[0067] , ;in,

[0068] is the prediction error ratio corresponding to the current processing parameter group, The number of types of subdivided parameters in the current processing parameter group. is the contribution weight of the i-th subdivision parameter, is the subdivision error ratio corresponding to the i-th subdivision parameter.

[0069] In this embodiment, it should be noted that The purpose of modeling the differences in parameter influence is to quantify the differential contribution of each machining parameter to accuracy, breaking through the limitations of traditional methods that treat it equally. A slight change in cutting speed (e.g., 90m / min to 95m / min) may increase the error ratio by 0.2, while the same change in feed rate only increases it by 0.05 - the influence of the two is essentially different. (Such as speed weight ( =0.6), feed rate ( =0.3), forcing the model to focus on the core parameter (speed) and suppress the interference of the secondary parameter (feed rate). The weight distribution is based on physical mechanisms (for example, cutting thermal deformation is twice as sensitive to speed as feed rate).

[0070] Further, For the decoupled integration of the subdivision error, the independent parameter impact estimation of the fusion S3 output is integrated, that is, the subdivision error ratio , building a global accuracy prediction. The weighted summation ensures that the speed-dominated error fluctuations are not diluted by the secondary parameters. The interference of other parameters has been removed (completed in S3), so that the weighted results can truly reflect the multi-parameter coupling effect.

[0071] Further, For the weight normalization constraint, the prediction error ratio Strictly limited to the interval [0, 1], consistent with the error ratio definition domain of S1. If not normalized, the output may exceed the upper limit of the tolerance ratio (100%), losing the risk warning significance. After normalization: Ensure It always indicates “how much proportion of tolerance is occupied” and can directly trigger a graded alarm.

[0072] A gear hobbing machine tool machining accuracy prediction data processing system is also provided, the system comprising:

[0073] a data acquisition module for acquiring a target gear, acquiring multiple historical processing parameter groups corresponding to the target gear, acquiring historical accuracy data after gear cutting processing is completed according to each historical processing parameter group, and acquiring an error ratio corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group, wherein each historical processing parameter group includes multiple subdivided parameters;

[0074] A sequence generation module is used to arrange multiple historical processing parameter groups in ascending order of the values ​​of each subdivided parameter in the historical processing parameter group and form a parameter group sequence corresponding to each subdivided parameter;

[0075] a data processing module for obtaining a current machining parameter group corresponding to a target gear and including a plurality of subdivided parameters, obtaining the position of the value of the i-th subdivided parameter of the current machining parameter group in the parameter group sequence corresponding to the i-th subdivided parameter, obtaining historical machining parameter groups corresponding to two adjacent values ​​at the position and using them as two comparison parameter groups for the i-th subdivided parameter, and obtaining a subdivision error ratio corresponding to the i-th subdivided parameter in the current machining parameter group based on the error ratios corresponding to the two comparison parameter groups for the i-th subdivided parameter;

[0076] The data prediction module is used to obtain the prediction error ratio corresponding to the current processing parameter group according to the subdivision error ratio corresponding to each subdivision parameter in the current processing parameter group.

[0077] In one embodiment, the data acquisition module is also used to: obtain the gear design dimensions of the target gear, and obtain the historical processing dimensions corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group; obtain the error ratio based on the gear design dimensions and the historical processing dimensions.

[0078] In one embodiment, the data processing module is also used to: obtain the difference in error ratios corresponding to the two control parameter groups of the i-th subdivision parameter and use it as the total error variable, and obtain the error ratio corresponding to the smaller value of the two control parameter groups of the i-th subdivision parameter; obtain the difference in the value of the i-th subdivision parameter in the two control parameter groups and use it as the control parameter variable, obtain the difference between the value of the i-th subdivision parameter in the current processing parameter group and the value of the i-th subdivision parameter in the smaller value control parameter group and use it as the current parameter variable; obtain the subdivision error ratio corresponding to the i-th subdivision parameter based on the current parameter variable, control parameter variable, total error variable and error ratio corresponding to the smaller value of the two control parameter groups corresponding to the i-th subdivision parameter.

[0079] In this embodiment, it should be noted that, regarding the above-mentioned gear hobbing machine processing accuracy prediction data processing system, the specific method of performing operations therein has been described in detail in the implementation of the gear hobbing machine processing accuracy prediction data processing method, and will not be elaborated here.

[0080] The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various simple modifications can be made to the technical solution of the present invention, and these simple modifications all fall within the scope of protection of the present invention.

[0081] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction. To avoid unnecessary repetition, the present invention will not further describe various possible combinations.

[0082] In addition, the various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the present invention, they should also be regarded as the contents disclosed by the present invention.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and description of the present invention.

Claims

1. A method for processing data for predicting machining accuracy of a gear hobbing machine tool, characterized in that: include: Obtain a target gear, obtain multiple historical processing parameter groups corresponding to the target gear, obtain historical accuracy data of the target gear after processing according to each historical processing parameter group, and obtain an error ratio corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group, wherein each historical processing parameter group includes multiple subdivided parameters; Arrange the plurality of historical processing parameter groups in ascending order of values ​​of each subdivided parameter in the historical processing parameter group and form a parameter group sequence corresponding to each subdivided parameter; Obtain a current machining parameter group corresponding to the target gear and including multiple subdivision parameters, obtain the position of the value of the i-th subdivision parameter of the current machining parameter group in the parameter group sequence corresponding to the i-th subdivision parameter, obtain the historical machining parameter groups corresponding to the two adjacent values ​​at this position and use them as two comparison parameter groups for the i-th subdivision parameter, and obtain the subdivision error ratio corresponding to the i-th subdivision parameter in the current machining parameter group based on the error ratios corresponding to the two comparison parameter groups for the i-th subdivision parameter; Wherein, obtaining the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group according to the error ratio corresponding to the two control parameter groups of the i-th subdivision parameter includes: obtaining the difference between the error ratios corresponding to the two control parameter groups of the i-th subdivision parameter and using it as the total error variable, and obtaining the error ratio corresponding to the smaller value of the two control parameter groups of the i-th subdivision parameter; obtaining the difference between the values ​​of the i-th subdivision parameter in the two control parameter groups and using it as the control parameter variable, obtaining the difference between the value of the i-th subdivision parameter in the current processing parameter group and the value of the i-th subdivision parameter in the smaller value control parameter group and using it as the current parameter variable; obtaining the subdivision error ratio corresponding to the i-th subdivision parameter according to the current parameter variable, the control parameter variable, the total error variable and the error ratio corresponding to the smaller value of the two control parameter groups corresponding to the i-th subdivision parameter; Obtain the prediction error ratio corresponding to the current processing parameter group according to the subdivision error ratio corresponding to each subdivision parameter in the current processing parameter group; Among them, the prediction error ratio is expressed as: , ;in, is the prediction error ratio corresponding to the current processing parameter group, The number of types of subdivided parameters in the current processing parameter group. is the contribution weight of the i-th subdivision parameter, is the subdivision error ratio corresponding to the i-th subdivision parameter.

2. The method for processing data for predicting machining accuracy of a gear hobbing machine tool according to claim 1, wherein: The step of obtaining the error ratio corresponding to each historical processing parameter group according to the historical accuracy data corresponding to each historical processing parameter group includes: Obtain the gear design dimensions of the target gear, and obtain the historical processing dimensions corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group; Obtain the error ratio based on the gear design dimensions and historical processing dimensions.

3. The method for processing data for predicting machining accuracy of a gear hobbing machine tool according to claim 2, wherein: The error ratio obtained based on the gear design size and historical processing size is expressed as: ;in, is the error ratio corresponding to the jth group of historical processing parameter groups, Design dimensions for the gears, is the historical processing size corresponding to the jth group of historical processing parameter groups, is the maximum error.

4. The method for processing data for predicting machining accuracy of a gear hobbing machine tool according to claim 1, wherein: The subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group is obtained based on the error ratio corresponding to the two comparison parameter groups of the i-th subdivision parameter, and is expressed as: ;in, is the subdivision error ratio corresponding to the i-th subdivision parameter, is the error ratio of a control parameter group, is the error ratio of another control parameter group, is the larger value of the i-th subdivision parameter of the two control parameter groups, is the smaller value of the i-th subdivision parameter in the two control parameter groups, The value of the i-th subdivision parameter of the current processing parameter group.

5. The method for processing data for predicting machining accuracy of a gear hobbing machine tool according to claim 1, wherein: Also includes: The gear design dimensions of the target gear are obtained, and the prediction accuracy data is obtained based on the prediction error ratio.

6. A gear hobbing machine tool machining accuracy prediction data processing system, characterized in that: The system comprises: a data acquisition module for acquiring a target gear, acquiring multiple historical processing parameter groups corresponding to the target gear, acquiring historical accuracy data after gear cutting processing is completed according to each historical processing parameter group, and acquiring an error ratio corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group, wherein each historical processing parameter group includes multiple subdivided parameters; A sequence generation module is used to arrange multiple historical processing parameter groups in ascending order of the values ​​of each subdivided parameter in the historical processing parameter group and form a parameter group sequence corresponding to each subdivided parameter; a data processing module for obtaining a current machining parameter group corresponding to a target gear and including a plurality of subdivided parameters, obtaining the position of the value of the i-th subdivided parameter of the current machining parameter group in the parameter group sequence corresponding to the i-th subdivided parameter, obtaining historical machining parameter groups corresponding to two adjacent values ​​at the position and using them as two comparison parameter groups for the i-th subdivided parameter, and obtaining a subdivision error ratio corresponding to the i-th subdivided parameter in the current machining parameter group based on the error ratios corresponding to the two comparison parameter groups for the i-th subdivided parameter; Wherein, obtaining the subdivision error ratio corresponding to the i-th subdivision parameter in the current processing parameter group according to the error ratio corresponding to the two control parameter groups of the i-th subdivision parameter includes: obtaining the difference between the error ratios corresponding to the two control parameter groups of the i-th subdivision parameter and using it as the total error variable, and obtaining the error ratio corresponding to the smaller value of the two control parameter groups of the i-th subdivision parameter; obtaining the difference between the values ​​of the i-th subdivision parameter in the two control parameter groups and using it as the control parameter variable, obtaining the difference between the value of the i-th subdivision parameter in the current processing parameter group and the value of the i-th subdivision parameter in the smaller value control parameter group and using it as the current parameter variable; obtaining the subdivision error ratio corresponding to the i-th subdivision parameter according to the current parameter variable, the control parameter variable, the total error variable and the error ratio corresponding to the smaller value of the two control parameter groups corresponding to the i-th subdivision parameter; A data prediction module is used to obtain a prediction error ratio corresponding to the current processing parameter group according to the subdivision error ratio corresponding to each subdivision parameter in the current processing parameter group; Among them, the prediction error ratio is expressed as: , ;in, is the prediction error ratio corresponding to the current processing parameter group, The number of types of subdivided parameters in the current processing parameter group. is the contribution weight of the i-th subdivision parameter, is the subdivision error ratio corresponding to the i-th subdivision parameter.

7. The gear hobbing machine tool machining accuracy prediction data processing system according to claim 6, characterized in that: The data acquisition module is also used for: Obtain the gear design dimensions of the target gear, and obtain the historical processing dimensions corresponding to each historical processing parameter group based on the historical accuracy data corresponding to each historical processing parameter group; Obtain the error ratio based on the gear design dimensions and historical processing dimensions.

Images