A radar chart-based hole-making quality evaluation method and system, and a storage medium
By using a radar chart evaluation model, combined with box plot method and dimensional unification processing, the one-sidedness and subjectivity of hole-making quality evaluation are solved, realizing multi-dimensional correlation evaluation and visualized hole-making quality assessment, thus improving the accuracy and applicability of the evaluation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHENGDU AIRCRAFT INDUSTRY GROUP
- Filing Date
- 2026-02-25
- Publication Date
- 2026-06-09
AI Technical Summary
Existing hole-making quality evaluation methods suffer from problems such as one-sidedness, subjectivity, and poor intuitiveness. They cannot fully reflect the core influencing factors of hole-making quality and lack visual comparison methods.
A radar chart-based hole-making quality assessment method is adopted. By identifying multiple characteristic data affecting hole-making quality, the data are corrected using the box plot method. After unifying the dimensions, a radar chart is drawn. The area size measures the quality. The radar chart is drawn using negative correlation of data. The coordinate axes are sorted according to the process influence weight, so as to achieve an objective and intuitive quality evaluation.
It enables a comprehensive, objective, and intuitive evaluation of hole-making quality, eliminates extreme outliers and dimensional differences, improves the accuracy and consistency of the evaluation, and supports flexible adaptation to different on-site production needs.
Smart Images
Figure CN122175433A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the technical field of hole quality evaluation, specifically relating to a hole quality evaluation method, system, and storage medium based on radar charts. Background Technology
[0002] In the field of machining, the quality of hole making directly determines the connection reliability and service life of mechanical parts. Defects such as burrs, splitting, tearing, and roundness errors can easily cause assembly interference, stress concentration, and other problems, affecting product performance. Current hole making quality evaluation methods in the industry have the following limitations: (1) Traditional single-parameter testing (such as measuring only roundness error) cannot fully reflect the overall quality; (2) Multi-parameter evaluation requires manual setting of weights, which mainly relies on processing experience, resulting in strong subjectivity in the results; (3) The current evaluation method lacks visual comparison means and cannot intuitively compare the quality of hole making of the two sets of holes.
[0003] Currently, the core influencing factors of hole-making quality are interrelated yet independent. Specifically, the passing of a single parameter cannot represent the overall quality standard, and existing technologies have not yet formed a comprehensive evaluation system that can integrate these parameters. Therefore, how to construct a hole-making quality evaluation method that can cover the core influencing factors and is objective, intuitive, and quantifiable has become a core problem that urgently needs to be solved in this technical field. Summary of the Invention
[0004] The purpose of this invention is to provide a method, system, and storage medium for evaluating hole quality based on radar charts, aiming to solve the above-mentioned problems and overcome the shortcomings of existing hole quality evaluation methods, such as being one-sided, subjective, and lacking intuitiveness.
[0005] This invention is mainly achieved through the following technical solutions: A method for evaluating hole quality based on radar charts includes the following steps: Step S1: Determine N characteristic data that affect hole quality, and draw an N-dimensional radar chart as the radar chart evaluation model. Step S2: Obtain the characteristic data of hole making, and correct it using the box plot method for several measurement data of each characteristic data; Step S3: Based on the corrected feature data and after unifying the dimensions, draw a closed radar chart using the radar chart evaluation model; Step S4: Calculate the area of the radar chart and measure the hole quality based on the area size; if all N feature data are negatively correlated with hole quality, the smaller the area of the radar chart, the higher the hole quality; if all N feature data are positively correlated with hole quality, the larger the area of the radar chart, the higher the hole quality.
[0006] To better realize the present invention, step S2 further includes the following steps: Step S21: For each feature data, arrange the several measurement data in ascending order to obtain the measurement data sequence; Step S22: Based on the measurement data sequence, calculate the front quartile and back quartile of the feature data, and calculate the interquartile range based on the difference between the front quartile and the back quartile. IQR ; Step S23: Calculate the lower limit of the anomaly boundary IQ min and upper limit IQ max ; Abnormal boundary lower limit IQ min =First-order quartile - 1.5 × IQR ; Abnormal boundary upper limit IQ max =Back-end quartile + 1.5 × IQR ; Step S24: Place [ IQ min , IQ max Measurement data outside the range are removed to obtain the corrected feature data.
[0007] To better implement this invention, further, in step S3, the modified feature data is subjected to dimensional unification correction, and the dimensionally corrected feature data is as follows: (5) in, The i-th / group of feature data after dimension correction; The i-th / group of measurement data before dimension correction of the characteristic data; This is the correction constant that is set.
[0008] The aforementioned dimensionless correction process ensures that all feature data are numerically consistent, avoiding differences in the order of magnitude of the feature data from affecting radar chart plotting.
[0009] To better realize the present invention, further, in step S1, six characteristic data affecting hole quality are determined, and a six-dimensional radar chart is drawn accordingly as a radar chart evaluation model; the six characteristic data affecting hole quality are: average burr length, average number of burrs, average length of splitting, average number of splitting, hole tear width, and hole roundness error.
[0010] To better realize the present invention, further, in step S1, when drawing the radar chart evaluation model, the influence weight of the process is first analyzed, and the influence weight of the feature data on the hole quality is ordered from high to low. The coordinate axes corresponding to the feature data are sorted clockwise starting from the 6 o'clock position, and feature data of the same type are arranged adjacently. After determining the coordinate axes, the coordinate axis parameters are fixed and applied to the hole quality evaluation of the same batch or the same group.
[0011] Specifically, in step S1, the selection of feature data follows the principles of correlation, independence, and measurability, while prioritizing feature data that significantly impacts assembly quality or is of key concern in hole-making quality inspection. In the radar chart evaluation model, the coordinate axes, after process influence weight analysis, are sorted clockwise from the 6 o'clock position according to the influence weight of parameters on hole-making quality, with parameters of the same type of defect arranged adjacently. After determining the coordinate axes, the coordinate axis parameters need to be fixed before being applied to the hole-making quality evaluation of the same batch or group, avoiding the influence of the dimensional arrangement order on the area size. In summary, the advantages of this sorting scheme for model construction and quality evaluation are reflected in four aspects: ① Improve the accuracy of evaluation by prioritizing core parameters with high weights. This can highlight the impact of key defects on hole quality, avoid secondary parameters interfering with core judgments, and make the conclusions on quality more accurate. ② Improve comparison efficiency. The adjacent distribution of defect parameters of the same type can quickly identify the overall level of a certain type of defect. When comparing radar charts of multiple samples, the core source of quality difference can be located in a short time. ③ Enhance model consistency: Fixed sorting rules ensure that radar charts drawn by different batches and different personnel have a unified reference standard, eliminating evaluation bias caused by arbitrary parameter arrangement; ④ In the evaluation of hole-making quality in the same batch or group, the influence of the arrangement order of dimensions on the area size is avoided.
[0012] To better realize the present invention, in step S3, a radar chart is drawn using negative data correlation; in step S4, the smaller the area of the drawn radar chart, the higher the hole quality.
[0013] In step S3, a radar chart is drawn using negative correlation of data. That is, if a feature data point indicates that a larger value corresponds to better hole quality, then the larger the value, the closer it is to the center of the radar chart; conversely, if a feature data point indicates that a smaller value corresponds to better hole quality, then the smaller the value, the closer it is to the center of the radar chart. Therefore, the smaller the radar chart area, the closer the detected values of each feature data point are to the ideal values, and the better the hole quality of that group of holes.
[0014] This invention is mainly achieved through the following technical solutions: A radar chart-based hole-making quality assessment system, implemented based on the aforementioned radar chart-based hole-making quality assessment method, includes: The data acquisition and preprocessing module is used to acquire the feature data of hole making and preprocess the feature data using the box plot method; The radar plotting module is used to plot radar charts based on preprocessed feature data and a radar chart evaluation model. The hole-making evaluation module is used to calculate the area of the radar image and evaluate the hole-making quality based on the size of the area.
[0015] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned radar chart-based hole quality assessment method.
[0016] The beneficial effects of this invention are as follows: (1) This invention avoids the one-sidedness of single-parameter evaluation by constructing a radar chart evaluation model containing multiple core parameters, achieving multi-dimensional correlation evaluation and avoiding differences caused by subjective evaluation. Secondly, this invention transforms hole-making quality data into a visual radar chart and quantifies the area, using the area size to determine the quality of hole-making, thus achieving a comprehensive, objective, and intuitive evaluation of hole-making quality. Therefore, this invention can intuitively compare hole-making quality through radar chart area, quickly identifying the quality of hole-making. This invention has good intuitiveness; the radar chart can clearly present parameter differences and overall quality levels, adapting to rapid comparison of multiple sets of samples.
[0017] (2) This invention uses box plots to correct data, eliminating extreme outliers and ensuring that the data accurately reflects the hole-making quality. This invention also eliminates dimensional differences, improving objectivity and ensuring high data reliability. This invention uses the corrected actual data and radar chart area as the evaluation basis, without manual weighting, ensuring consistent evaluation results across different batches. Furthermore, this invention supports dimension / parameter adjustment, flexibly adapting to different on-site production needs. The coordinate axis scale and area threshold can be adjusted according to different materials, hole diameters, or hole-making processes without changing the model framework, demonstrating good practicality.
[0018] (3) The radar chart evaluation model innovatively introduces a dual preprocessing procedure of "box plot outlier removal + dimensionless correction" to remove extreme data caused by accidental factors in hole making, ensuring that the parameters truly reflect the stable state of the process. At the same time, it eliminates the numerical deviation of parameters with different dimensions, such as burr length and roundness error, laying a precise data foundation for subsequent evaluation. Secondly, the radar chart evaluation model innovatively formulates a negative correlation data representation form and solidifies coordinate axis parameters, establishing the unique criterion of "the smaller the area, the better the quality," avoiding the evaluation distortion caused by the mixed positive and negative parameters and the difference in the order of dimension arrangement in conventional radar charts. The radar chart evaluation model forms a complete quantitative closed-loop system of "data-model-evaluation." This closed-loop system realizes the objective evaluation of hole making quality and can guide the optimization of process parameters in reverse, breaking through the application limitation of conventional radar charts that are only used for display. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of a six-dimensional radar chart evaluation model; Figure 2 This is a schematic diagram of the radar map area in Example 2; Figure 3 This is a schematic radar chart of sample 1 in Example 2; Figure 4 This is a schematic radar chart of sample 2 in Example 2; Figure 5 This is a schematic radar chart of sample 3 in Example 2; Figure 6 This is a schematic radar chart of sample 4 in Example 2; Figure 7 This is a schematic diagram of the radar images of samples 1 to 4 in Example 2 overlay; Figure 8 Images of samples 1 to 4 taken during hole preparation in Example 2; Figure 9 This is a flowchart of the hole quality assessment method based on radar charts according to the present invention. Detailed Implementation
[0020] Example 1: A method for evaluating hole quality based on radar charts, such as Figure 9 As shown, it includes the following steps: (1) Determine the feature data and construct the radar chart evaluation model; such as Figure 1 As shown, the definitions and detection dimensions of the six characteristic data points for hole quality are determined; and a radar chart evaluation model, namely a 6-dimensional radar chart, is constructed using a common radar chart model. The six characteristic data points for hole quality are used as the six coordinate axes of the radar chart. The six coordinate axes in the radar chart are distributed at equal angles (the angle between adjacent coordinate axes is 60°), forming a hexagonal radar chart framework.
[0021] Preferably, the six characteristic data points of the hole-making quality are: 1. Average burr length (unit: mm), reflecting the overall length level of burrs at the orifice; 2. Average number of burrs (unit: burrs / hole), reflecting the burr distribution density; 3. Average splitting length (unit: mm), reflecting the extent of splitting defects in the material; 4. Average number of splits (unit: strips / hole), reflecting the frequency of splitting defects; 5. Hole tear width (unit: mm), reflecting the severity of hole wall tearing defects; 6. Hole roundness error reflects the accuracy of hole shape.
[0022] (2) Data acquisition and preprocessing, and removal of abnormal measurement data; the characteristic data of each group of holes are measured, and extreme outlier data are removed by box plot method; (3) Draw a radar chart based on the radar chart evaluation model, solve for the area, and evaluate the hole quality; then draw a closed radar chart based on the corrected feature data, such as Figure 2 As shown, the area of the radar chart is used to measure the hole quality. The smaller the area of the radar chart, the closer the detected values of each feature data are to the ideal values, and the better the hole quality of that group of holes. Among them, the six feature data determined by the radar chart evaluation model are all negatively correlated with hole quality (the smaller the better). Therefore, holes with smaller radar chart areas have better hole quality.
[0023] Specifically, for a set of borehole samples to be evaluated, experimental data of the above six characteristic parameters were obtained using equipment such as an optical microscope. Then, each characteristic parameter was corrected using a box plot to remove extreme outliers and ensure the generalizability of the measurement results.
[0024] Preferably, the box plot method is as follows: (1) Arrange the measurement results of a characteristic parameter in ascending order. For example, for the measurement data x1~x8 of the characteristic parameter x, the following sequence is obtained: (1) (2) Calculate the first quartile of the characteristic parameter ( ) and the third quartile ( ); (2) (3) Substituting the data from step (1) into the formula, we can calculate the following: Interquartile range: ; median .
[0025] in: The first half of the data ( ); The second half of the data ( ) is the median.
[0026] (3) Calculate the lower limit of the anomaly boundary IQ min and upper limit IQ max ; (4) When the data is less than 29.5 or greater than 45.5, it is considered an extreme outlier and needs to be removed.
[0027] Preferably, in this invention, to avoid evaluation deviations due to differences in parameter dimensions, it is necessary to perform dimension correction calculations on each parameter. The parameter dimension correction calculation formula is shown in formula (5), and in the next evaluation, each parameter remains unchanged.
[0028] (5) in, This indicates the corrected parameters. This indicates the parameters before correction. This is the correction constant that is set.
[0029] For example, regarding the average burr length, the average burr length of the first group of holes is... The average length of the burrs in the second group of holes is , corrected constant After correction, the average burr length of the first group of holes is... The average length of the burrs in the second group of holes is .
[0030] In the radar chart evaluation model, a set of six feature data points after hole correction are marked at corresponding positions on the corresponding coordinate axes. A smooth curve connects the marked points to form a closed polygon, completing the radar chart drawing and calculating its area. Specifically, the radar chart evaluation model can modify the radar chart dimensions according to actual processing and production needs; for example, it can be reduced to a four-dimensional radar chart or increased to an eight-dimensional radar chart. Simultaneously, radar chart parameters can be replaced; for example, aperture error can be used to replace burr length. This invention can be flexibly applied to various scenarios through dimensional changes and parameter replacements.
[0031] Example 2: A radar chart-based method for evaluating hole quality is proposed. In this embodiment, carbon fiber composite materials commonly used in the aerospace field are used as the evaluation object for hole making. The hole making parameters are set as follows: hole diameter 5mm, hole depth 3mm, and rotation speed 18500r / min.
[0032] In this embodiment, a total of 4 sets of hole samples were prepared, each set containing 20 holes processed under the same process conditions. By setting different hole conditions, 4 sets of holes with different qualities were constructed.
[0033] In this embodiment, the six selected feature data are: 1. Average burr length (unit: mm); 2. Average number of burrs (unit: strips / hole); 3. Average split length (unit: mm); 4. Average number of splits (unit: strips / hole); 5. Hole tear width (unit: mm); 6. Hole roundness error.
[0034] The hole-making data was observed using an electron microscope, and experimental data were measured and recorded. In this embodiment, based on process influence weight analysis, the coordinate axes were determined to be ordered clockwise as follows: average number of splits → average length of splits → average number of burrs → average length of burrs → hole roundness error → hole tear width. This order prioritizes the core parameters affecting assembly reliability (split-related features), while arranging burr-type and split-type defect parameters adjacent to each other.
[0035] As shown in Table 1, in this embodiment, the experimental data were corrected using the box plot method, and the following four sets of experimental results were finally obtained.
[0036] In this embodiment, the correction constant B=10. Referring to formula (5), the parameters of the four groups of samples are corrected, as shown in Table 2, and the corrected experimental data are obtained.
[0037] like Figures 3-7 As shown, radar charts were created based on the data from the four sets of samples in Table 2 and then overlaid together.
[0038] Table 1
[0039]
[0040] Table 2
[0041]
[0042] like Figure 2 As shown, in this embodiment, for the area of a single radar image, the "polar radius" (i.e., the value from the center to the vertex) corresponding to each vertex can be obtained. Then, the hexagon in the radar image is divided into 6 triangles from the center (center → vertex 1 → vertex 2, center → vertex 2 → vertex 3... center → vertex 6 → vertex 1) and the area is calculated.
[0043] The formula for calculating a single triangle is as follows: (6) In formula (6), It is the polar radius of the i-th vertex. To ensure closure, we have... Because radar images are regular hexagons, therefore .
[0044] like Figure 2 As shown, in this embodiment, sample 1 is taken as an example. , , , , , The area of each triangle is calculated using formula (6), and the radar image area is obtained by summing the results.
[0045] ; ; Repeating the above method, the areas of the four samples in this embodiment were calculated as follows: , , , .
[0046] like Figure 7 As shown, in this embodiment, sample 1 has the smallest radar image area and the best hole quality. Sample 2's quality is second to sample 1 but better than sample 3, while sample 4 has the worst hole quality. Comparing the radar images of the four samples simultaneously reveals that sample 4 has the largest number of splits and the longest split length, which is the core reason for its poor quality, effectively improving the efficiency of quality problem diagnosis. Figure 8 As shown, the above evaluation results are consistent with actual processing. Furthermore, the radar chart evaluation model of the present invention can modify the radar chart dimensions and replace the radar chart parameters according to the needs of actual processing and production, and can be flexibly applied to various hole-making scenarios.
[0047] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications or equivalent changes made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.
Claims
1. A method for evaluating hole quality based on radar charts, characterized in that, Includes the following steps: Step S1: Determine N characteristic data that affect hole quality, and draw an N-dimensional radar chart as the radar chart evaluation model. Step S2: Obtain the characteristic data of hole making, and correct it using the box plot method for several measurement data of each characteristic data; Step S3: Based on the corrected feature data and after unifying the dimensions, draw a closed radar chart using the radar chart evaluation model; Step S4: Calculate the area of the radar chart and measure the hole quality based on the area size; if all N feature data are negatively correlated with hole quality, the smaller the area of the radar chart, the higher the hole quality; if all N feature data are positively correlated with hole quality, the larger the area of the radar chart, the higher the hole quality.
2. The method for evaluating hole quality based on radar charts according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: For each feature data, arrange the several measurement data in ascending order to obtain the measurement data sequence; Step S22: Based on the measurement data sequence, calculate the front quartile and back quartile of the feature data, and calculate the interquartile range based on the difference between the front quartile and the back quartile. IQR ; Step S23: Calculate the lower limit of the anomaly boundary IQ min and upper limit IQ max ; Abnormal boundary lower limit IQ min =First-order quartile - 1.5 × IQR ; Abnormal boundary upper limit IQ max =Back-end quartile + 1.5 × IQR ; Step S24: Place [ IQ min , IQ max Measurement data outside the range are removed to obtain the corrected feature data.
3. A method for evaluating hole quality based on radar charts according to claim 1 or 2, characterized in that, In step S3, the modified feature data undergoes dimensional unification correction, and the dimensionally corrected feature data is as follows: （5） in, This is the i-th feature data after dimension correction; The i-th / group of measurement data before dimension correction of the characteristic data; This is the correction constant that is set.
4. The method for evaluating hole quality based on radar charts according to claim 1, characterized in that, In step S1, six characteristic data affecting hole quality are determined, and a six-dimensional radar chart is plotted accordingly as a radar chart evaluation model. The six characteristic data affecting hole quality are: average burr length, average number of burrs, average length of splitting, average number of splitting, hole tear width, and hole roundness error.
5. A method for evaluating hole quality based on radar charts according to claim 1 or 4, characterized in that, In step S1, when drawing the radar chart evaluation model, the influence weight of the process is first analyzed. The influence weight of the feature data on the hole quality is ordered from high to low. The coordinate axis corresponding to the feature data starts from the 6 o'clock position and is sorted clockwise, and the feature data of the same type are arranged adjacently. After determining the coordinate axis, the coordinate axis parameters are fixed and applied to the hole quality evaluation of the same batch or the same group.
6. The method for evaluating hole quality based on radar charts according to claim 1, characterized in that, In step S3, a radar chart is drawn using negative data correlation; in step S4, the smaller the area of the drawn radar chart, the higher the hole quality.
7. The method for evaluating hole quality based on radar charts according to claim 1, characterized in that, In step S4, the N-dimensional radar image is divided into N triangles from the center, and the sum of the areas of the N triangles is calculated as the area of the radar image; wherein, the area of the i-th triangle is: （6） ； in: It is the polar radius of the i-th vertex, and ; It is the polar radius of the (i+1)th vertex.
8. A radar chart-based hole-making quality assessment system, implemented based on the radar chart-based hole-making quality assessment method according to any one of claims 1 to 7, characterized in that, include: The data acquisition and preprocessing module is used to acquire the feature data of hole making and preprocess the feature data using the box plot method; The radar plotting module is used to plot radar charts based on preprocessed feature data and a radar chart evaluation model. The hole-making evaluation module is used to calculate the area of the radar image and evaluate the hole-making quality based on the size of the area.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements a radar chart-based hole quality assessment method as described in any one of claims 1 to 7.