A method and system for controlling near-net-shape grinding allowance

Abstract

This invention relates to a near-net-shape grinding allowance control method and system, belonging to the field of grinding technology. It includes: acquiring laser detection information, image detection information, and workpiece requirements specifications of a workpiece on a grinding device; retrieving single-sided features and single-sided required distance values ​​of the workpiece based on the workpiece requirements specifications; determining the image-estimated single-sided quality and single-sided quality by combining the single-sided features and image detection information; determining the single-sided detection distance value based on the laser detection information; determining the quality adjustment value based on the single-sided quality; determining the operating control parameters by combining the single-sided detection distance value, quality adjustment value, and single-sided required distance value, and outputting the operating control parameters to the grinding device. This invention improves the overall grinding efficiency when grinding large batches of workpieces.

Description

Technical Field

[0001] This invention relates to the field of grinding technology, and in particular to a method and system for controlling near-net-shape grinding allowance. Background Technology

[0002] Grinding is a machining process that uses abrasive grains on the surface of a grinding wheel to perform precision cutting on the surface of a workpiece through a combination of micro-cutting, scribing, and polishing.

[0003] After a workpiece undergoes casting, forging, or other machining processes, grinding is generally required to compensate for shortcomings in workpiece precision and surface quality, bringing the workpiece's dimensions, shape, and surface properties closer to the finished product requirements, thus forming a near-net-shape workpiece. During grinding, operators typically use calipers and micrometers to pre-inspect the blank's allowance and surface defects. After clamping the workpiece, operators set basic parameters such as the grinding wheel speed and table speed before performing the grinding process.

[0004] Currently, when grinding workpieces, operators need to inspect the workpieces first and then input operating parameters to perform the grinding process, which increases the overall processing time and results in low overall grinding efficiency when processing large batches of workpieces. Summary of the Invention

[0005] To improve the overall grinding efficiency when grinding a large number of workpieces, this invention provides a near-net-shape grinding allowance control method and system.

[0006] In a first aspect, the present invention provides a method for controlling near-net-shape grinding allowance, which adopts the following technical solution:

[0007] A method for controlling near-net-shape grinding allowance includes:

[0008] S1: Collect laser detection information, image detection information and workpiece specifications required for the workpiece on the grinding device;

[0009] S2: Retrieve the single-sided features and single-sided distance values ​​of the workpiece based on the workpiece requirements and specifications;

[0010] S3: Combine the single-sided features of the workpiece with image detection information to determine the image prediction of the single-sided surface and the single-sided quality.

[0011] S4: Determine the single-sided detection distance value based on the laser detection information;

[0012] S5: Determine the quality adjustment value based on the single-sided quality;

[0013] S6: Combine the single-sided detection distance value, quality adjustment value and single-sided required distance value to determine the operation control parameters, and output the operation control parameters to the grinding device;

[0014] Methods for predicting a single-sided image and determining its quality include:

[0015] S31: Image contour features and image defect features are obtained based on image detection information;

[0016] S32: Determine the feature matching rate based on the consistency between the image contour features and the single-sided features of the workpiece;

[0017] S33: Determine the feature matching facet based on the feature matching rate, and use the feature matching facet as the image prediction facet;

[0018] S34: Determine the baseline features of single-sided defects based on the image prediction of a single side;

[0019] S35: Combine image defect features with single-sided defect baseline features to determine the defect prediction status, and use the defect prediction status as the single-sided quality status.

[0020] By employing the above technical solution, laser detection information, image detection information, and workpiece specifications are collected and analyzed to determine the single-sided detection distance, quality adjustment value, and single-sided required distance. These are then combined with the determined operating control parameters, which are output to the grinding device. This automatically ensures that the operating control parameters accurately match the actual state and required specifications of the workpiece, guaranteeing that the ground workpiece meets the required specifications and improving the overall grinding efficiency for mass production. Image contour features and image defect features are identified, and the feature matching rate is determined by the consistency between the image contour features and the single-sided features of the workpiece. The feature-matched single-sided surface is then identified, and the defect prediction is determined by combining the single-sided defect baseline features. The feature-matched single-sided surface is used as the image prediction single-sided surface, and the defect prediction is used as the single-sided quality status. This dual verification of feature matching and defect comparison provides a reliable basis for determining the quality adjustment value.

[0021] Optional methods for determining one side of feature matching include:

[0022] S331: Retrieve single-sided reference matching rate based on workpiece requirements and specifications;

[0023] S332: Determine whether the feature matching rate is greater than the single-sided baseline matching rate;

[0024] S333: If yes, then select the feature matching rate that is greater than the single-sided baseline matching rate and use it as the matching rate.

[0025] S334: Retrieve the number of values ​​that satisfy the matching rate;

[0026] S335: Select the matching rate based on the number of satisfied values ​​and use it as the selected matching rate, and use the single side of the workpiece single-sided feature corresponding to the selected matching rate as the feature matching single side.

[0027] S336: If not, collect historical requirement specifications;

[0028] S337: Combine image contour features, image defect features and historical requirement specifications to determine the historical specification facet, and use the historical specification facet as the feature matching facet.

[0029] By adopting the above technical solution, the reliability of feature matching single face is ensured by retrieving the single face reference matching rate and performing quantitative screening. Furthermore, by introducing historical requirement specifications, the shortcomings of the current insufficient matching rate are made up for, thereby improving the applicability and fault tolerance of the determined feature matching single face and ensuring the continuity of the grinding allowance control process.

[0030] Optionally, the methods for selecting the matching rate include:

[0031] S3351: Determine whether the value that satisfies the condition is only one;

[0032] S3352: If yes, then the matching rate is directly used as the selection matching rate;

[0033] S3353: If not, select the two highest matching rates as candidate matching rates;

[0034] S3354: Calculate the difference between the matching rates of two candidate matches and use it as the matching difference;

[0035] S3355: Determines the image brightness value based on image detection information;

[0036] S3356: Determine the brightness adaptation matching rate by combining the image brightness value and the matching difference, and use the brightness adaptation matching rate as the selection matching rate.

[0037] By adopting the above technical solution, when there are multiple values, the collaborative analysis of matching difference and image brightness value avoids the deviation caused by selecting solely based on the matching rate, making the determination of the matching rate more in line with the actual detection scenario, and further improving the accuracy of feature matching.

[0038] Optional methods for determining the brightness adaptation matching rate include:

[0039] S33561: Determine the matching benchmark difference based on the single-sided benchmark matching rate;

[0040] S33562: Determine whether the matching difference is greater than the matching baseline difference;

[0041] S33563: If yes, select the larger candidate matching rate as the brightness adaptation matching rate;

[0042] S33564: If not, then determine the candidate single face based on the candidate matching rate;

[0043] S33565: Determine the brightness influence range of a single surface based on the candidate single surface;

[0044] S33566: Determine the brightness reference value by combining the image brightness value with the single-sided brightness influence range;

[0045] S33567: Select the candidate matching rate corresponding to the smaller brightness reference value and use it as the brightness adaptation matching rate.

[0046] By adopting the above technical solution, the selection of brightness adaptation matching rate is differentiated and refined by judging whether the matching difference is greater than the matching benchmark difference. When the matching difference is significant, a high matching rate is selected first to ensure basic reliability. When the matching difference is small, the brightness adaptation matching rate is selected by combining the brightness characteristics of the single side and the actual image brightness. This effectively reduces the impact of brightness interference on the matching results and improves the rationality of the selected matching rate.

[0047] Optional methods for determining the historical specification single side include:

[0048] S3371: Retrieve the historical specification outline features and historical specification defect features of each historical single-sided document based on historical requirement specifications.

[0049] S3372: Analyze the similarity between historical specification contour features and image contour features to obtain contour similarity reference values;

[0050] S3373: Analyze the similarity between historical specification defect features and image defect features to obtain defect similarity reference values;

[0051] S3374: Determine the comprehensive similarity reference value by combining the contour similarity reference value and the defect similarity reference value;

[0052] S3375: Sort the similar comprehensive reference values ​​from largest to smallest, and take the historical single face of the historical requirement specification corresponding to the first-ranked similar comprehensive reference value as the historical selected single face, and take the historical selected single face as the historical specification single face.

[0053] By adopting the above technical solution, through the similarity analysis of contour and defect in two dimensions, reliable historical data support is provided for the determination of single-sided image prediction when the current matching rate is not up to standard, thus ensuring the consistency of grinding allowance control.

[0054] Optional methods for determining similarity comprehensive reference values ​​include:

[0055] S33741: Calculate the difference between the contour similarity reference value and the defect similarity reference value and use it as the similarity reference deviation value;

[0056] S33742: Collect the current time point;

[0057] S33743: Retrieve historical grinding time intervals based on historical requirement specifications;

[0058] S33744: Determine the time reference value by combining the current time point with the historical grinding time range;

[0059] S33745: Determine the time reference coefficient based on the time reference value;

[0060] S33746: Based on the time reference coefficient, the contour similarity reference value and the defect similarity reference value are weighted and calculated to obtain the time comprehensive reference value, and the time comprehensive reference value is used as the similarity comprehensive reference value.

[0061] By adopting the above technical solution, the time dimension is incorporated into the similarity analysis system. It takes into account the different historical grinding time intervals corresponding to grinding different historical requirements and specifications, and performs weighted adjustment to make the comprehensive similarity reference value more in line with the current grinding scenario. This improves the accuracy of single-sided selection of historical specifications and ensures the reliability of historical data reuse.

[0062] Optional methods for determining the time reference value include:

[0063] S337441: Determine whether the current time point falls within the historical grinding time interval;

[0064] S337442: If yes, then the historical grinding time interval into which the current time point falls shall be taken as the time interval into which it falls;

[0065] S337443: Determine the landing reference value based on the landing time interval, and use the landing reference value as the time reference value;

[0066] S337444: If not, calculate the difference between the current time point and the historical grinding time interval and use it as the time deviation value;

[0067] S337445: Sort the time deviation values ​​from smallest to largest, and take the historical grinding time interval corresponding to the first time deviation value as the deviation time interval.

[0068] S337446: Determine the deviation reference value based on the deviation time interval, and use the deviation reference value as the time reference value.

[0069] By adopting the above technical solution, the relationship between the current time point and the historical grinding time interval is determined, and the falling reference value or the deviation reference value is used as the time reference value respectively. Through quantitative analysis of the falling situation and the magnitude of the deviation, it is ensured that the time reference value can accurately reflect the time correlation between the current scenario and the historical grinding scenario, providing a scientific basis for the determination of the time reference coefficient and further improving the rationality of the similar comprehensive reference value.

[0070] Optionally, methods for determining whether a value falls within the reference value include:

[0071] S3374431: Determine the reference value for a single interval and the number of values ​​for the interval based on the time interval in which the value falls;

[0072] S3374432: Determine whether the number of values ​​falling into the interval is only one;

[0073] S3374433: If yes, then the single-interval reference value will be used as the reference value to fall into;

[0074] S3374434: If not, calculate the midpoint of the time interval into which the time falls and use it as the midpoint of the time interval into which the time falls.

[0075] S3374435: Calculate the deviation between the midpoint of the landing time and the current time point and use it as the midpoint deviation value;

[0076] S3374436: Determine the intermediate selection coefficient based on the intermediate deviation value;

[0077] S3374437: Combine the intermediate selection coefficient with the single interval reference value to determine the intermediate selection reference value, and use the intermediate selection reference value as the falling reference value.

[0078] By adopting the above technical solution, the reference value for falling into the interval is determined by using a single interval reference value or an intermediate deviation value and an intermediate selection coefficient, taking into full account the complex scenario that the current time point may fall into multiple historical grinding time intervals. Through the collaborative calculation of the interval midpoint deviation and the intermediate selection coefficient, the reference value that best fits the current time point is accurately selected, avoiding selection chaos when falling into multiple intervals, further improving the reliability of the time reference value, and providing more accurate support for the calculation of similar comprehensive reference values.

[0079] Secondly, the present invention provides a near-net-shape grinding allowance control system, which adopts the following technical solution:

[0080] A near-net-shape grinding allowance control system includes:

[0081] The data acquisition module is used to collect laser detection information, image detection information, workpiece requirements and specifications, historical requirements and specifications, and the current time point.

[0082] The memory stores a program for implementing a near-net-shape grinding allowance control method as described in any one of the first aspects;

[0083] The processor loads and executes programs stored in memory.

[0084] In summary, the present invention has at least one of the following beneficial technical effects:

[0085] 1. By collecting laser detection information, image detection information and workpiece requirements, and analyzing and determining the single-sided detection distance value, quality adjustment value and single-sided required distance value, and then combining the determination of operating control parameters and outputting them to the grinding device, the operating control parameters are automatically and accurately matched with the actual state and required specifications of the workpiece, ensuring that the workpiece meets the required specifications after grinding, and improving the overall grinding efficiency when grinding a large number of workpieces.

[0086] 2. By identifying image contour features and image defect features, and determining the feature matching rate by comparing the image contour features with the single-sided features of the workpiece, the feature matching single-sided features are determined and the defect prediction is determined by combining the single-sided defect reference features. The feature matching single-sided features are used as the image prediction single-sided features and the defect prediction is used as the single-sided quality status. Thus, through the dual verification of feature matching and defect comparison, a reliable basis is provided for determining the quality adjustment value.

[0087] 3. By retrieving the single-sided benchmark matching rate and performing quantitative screening, the reliability of the feature matching single-sided surface is ensured. Furthermore, by introducing historical requirement specifications, the shortcomings of the current insufficient matching rate are compensated for, thereby improving the applicability and fault tolerance of the determined feature matching single-sided surface and ensuring the continuity of the grinding allowance control process. Attached Figure Description

[0088] Figure 1 This is a flowchart of a method for controlling near-net-shape grinding allowance;

[0089] Figure 2 This is a flowchart illustrating the method for determining the single-sided image prediction and its quality.

[0090] Figure 3 This is a flowchart of the method for determining one side of feature matching. Detailed Implementation

[0091] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0092] A near-net-shape grinding allowance control method collects laser detection information, image detection information, workpiece requirements, historical requirements, and the current time point. First, based on the workpiece requirements, it clarifies the single-sided features and required distance values. Then, it identifies contours and defect features using image detection information, determines the estimated single-sided area through feature matching rate, and judges the single-sided quality by comparing defects. Next, it obtains the single-sided detection distance value based on laser detection information, determines the quality adjustment value, and integrates these to generate operating control parameters. When no matching features are available, it reuses historical requirements and combines time-weighted analysis to ensure reliability. This automatically and accurately matches the operating control parameters with the actual state and requirements of the workpiece, ensuring that the ground workpiece meets the requirements and improving the overall grinding efficiency when grinding large batches of workpieces.

[0093] Reference Figure 1 This invention discloses a method for controlling near-net-shape grinding allowance, comprising:

[0094] S1: Collect laser detection information, image detection information, and workpiece specifications required for the workpiece on the grinding device.

[0095] The grinding device refers to the apparatus used to grind workpieces. Laser inspection information refers to the set of quantitative data relating to the workpiece's geometric dimensions obtained after scanning and inspecting the workpiece on the grinding device using laser inspection equipment. Laser inspection information is acquired through laser inspection equipment pre-installed on the grinding device.

[0096] Image detection information refers to the set of visual data corresponding to the surface of a workpiece that needs to be ground, acquired by industrial imaging equipment. This image detection information is obtained through detection by industrial imaging equipment pre-installed on the grinding machine; the industrial imaging equipment can be a high-resolution industrial area scan camera.

[0097] Workpiece requirements specifications refer to the design standards that a workpiece must meet after processing. These specifications include core technical indicators such as single-sided workpiece features, single-sided required distance values, surface quality thresholds, dimensional tolerance ranges, and single-sided datum matching rates. Workpiece requirements specifications can be extracted from workpiece design drawings and process technology documents, or obtained through pre-input by the operator.

[0098] S2: Retrieve the single-sided features and single-sided distance values ​​of the workpiece based on the workpiece requirements and specifications.

[0099] Among them, the single-sided feature of a workpiece refers to the set of inherent geometric and technological attributes of a single surface of a workpiece to be ground. The single-sided feature of a workpiece includes the shape features of the surface (such as plane, arc surface, and step surface), contour datum markings, etc.

[0100] The single-sided required distance value refers to the target size distance parameter that a single surface of a workpiece needs to reach after machining.

[0101] The single-sided features and single-sided distance values ​​of the workpiece can be retrieved based on the workpiece requirements specifications for convenient subsequent use.

[0102] S3: Combine the single-sided features of the workpiece with image detection information to determine the image prediction of the single-sided surface and the single-sided surface quality.

[0103] Image-based single-sided prediction refers to determining the single side to be ground based on an image. Single-sided quality refers to the quality of the workpiece surface corresponding to the image.

[0104] By combining and analyzing the single-sided features of the workpiece with image detection information, the predicted single-sided quality and condition of the image can be determined, facilitating subsequent use.

[0105] To further ensure the reasonableness of the image prediction for single-sided and single-sided quality, it is necessary to perform further separate analysis and calculation on the image prediction for single-sided and single-sided quality, which will be explained in detail through the steps shown below.

[0106] Reference Figure 2 The method for predicting a single surface of an image and determining its quality includes the following steps:

[0107] S31: Image contour features and image defect features are obtained based on image detection information.

[0108] Image contour features refer to the set of visual features that reflect the shape, geometric dimensions, and other external attributes of a workpiece's single-sided edge. Image defect features refer to the set of visual features that characterize the location, shape, size, quantity, and other attributes of defects such as scratches, burns, pinholes, and burrs on the workpiece surface.

[0109] After preprocessing the image detection information by grayscale conversion and median filtering for noise reduction, the edge pixels in the image are extracted using the Canny edge detection algorithm or the Sobel operator. The shape, size and other image contour features of the workpiece are obtained by contour fitting and coordinate calculation. Then, morphological opening and closing operations are used to separate the defect region from the workpiece substrate. The image defect features are obtained by calculating parameters such as the area, perimeter and aspect ratio of the defect region.

[0110] S32: Determine the feature matching rate based on the consistency between the image contour features and the single-sided features of the workpiece.

[0111] Feature matching rate is a core indicator that quantifies the consistency between image contour features and workpiece single-sided features. A higher feature matching rate indicates a higher degree of overlap and similarity between the image contour features and the workpiece single-sided features.

[0112] This process involves comparing the image contour features with the individual features of each workpiece, covering three core dimensions: geometric shape (e.g., circle, rectangle, stepped surface), key dimensions (e.g., diameter, side length, spacing of reference holes), and contour topology (e.g., number and distribution of holes, edge transition method). A matching algorithm is then selected based on the comparison dimensions: geometric shape matching uses the shape context algorithm, size matching calculates the deviation rate of key dimensions, and contour topology matching uses the Hausdorff distance algorithm. The matching results for each dimension are then weighted (e.g., shape dimension weight 0.4, size dimension weight 0.4, topology dimension weight 0.2), and a weighted average is calculated. Finally, the calculation results are normalized and converted into a feature matching rate for convenient subsequent use.

[0113] S33: Determine the feature matching facet based on the feature matching rate, and use the feature matching facet as the image prediction facet.

[0114] Among them, feature matching single-sided refers to the single-sided surface of the workpiece to be ground selected based on the single-sided features of the workpiece.

[0115] By analyzing the feature matching rate, the feature matching facet is determined, and the feature matching facet is used as the image prediction facet, thereby improving the accuracy of the obtained image prediction facet.

[0116] To further ensure the rationality of the feature matching facet, it is necessary to perform further separate analysis and calculation on the feature matching facet, which will be explained in detail through the steps shown below.

[0117] Reference Figure 3 The method for determining one side of feature matching includes the following steps:

[0118] S331: Retrieve single-sided datum matching rate based on workpiece requirements and specifications.

[0119] The single-sided reference matching rate refers to the minimum matching rate threshold for determining whether the image contour features match the single-sided features of the workpiece. Different single-sided features of the workpiece correspond to different single-sided reference matching rates.

[0120] The single-sided datum matching rate can be retrieved based on the workpiece requirements and specifications for convenient subsequent use.

[0121] S332: Determine if the feature matching rate is greater than the single-sided baseline matching rate. If yes, proceed to S333; if no, proceed to S336.

[0122] Specifically, by judging whether the feature matching rate is greater than the single-sided reference matching rate, it is determined whether the workpiece identified by the image is the workpiece that needs to be ground.

[0123] S333: Select the feature matching rate that is greater than the single-sided baseline matching rate and use it as the satisfaction matching rate.

[0124] Among them, the matching rate refers to the feature matching rate that is greater than the single-sided benchmark matching rate.

[0125] When the feature matching rate is greater than the single-sided reference matching rate, it indicates that the workpiece identified by the image is the workpiece that needs to be ground. Therefore, the matching rate is selected for convenient subsequent use.

[0126] S334: Retrieve the number of values ​​that meet the matching rate.

[0127] Here, the number of values ​​that meet the matching rate refers to the number of values ​​that meet the matching rate.

[0128] By counting the number of matches, and retrieving the count results as the number of matches, it is convenient to use them later.

[0129] S335: Select the matching rate based on the number of satisfied values ​​and use it as the selected matching rate, and use the single-sided feature of the workpiece corresponding to the selected matching rate as the feature matching single-sided surface.

[0130] Among them, the selection matching rate refers to the matching rate corresponding to the selected matching rate.

[0131] By selecting a matching rate and using it as the selection matching rate, and then using the single-sided feature of the workpiece corresponding to the selection matching rate as the feature matching single-sided surface, the accuracy of the obtained feature matching single-sided surface is improved.

[0132] To further ensure the rationality of the selected matching rate, it is necessary to perform a further separate analysis and calculation of the selected matching rate, which will be explained in detail through the steps shown below.

[0133] The method for selecting the matching rate includes the following steps:

[0134] S3351: Determine whether the given value is exactly one. If yes, proceed to S3352; if no, proceed to S3353.

[0135] Specifically, by determining whether there is only one matching value, it is possible to determine whether further selection of the matching rate is needed.

[0136] S3352: Directly use the matching rate as the selection matching rate.

[0137] When there is only one value that satisfies the condition, it means that there is no need to further select the matching rate. Therefore, the matching rate is directly used as the selection matching rate.

[0138] S3353: Select the two highest matching rates and use them as candidate matching rates.

[0139] Among them, the candidate matching rate refers to the matching rate from which further selection is needed.

[0140] When there is more than one satisfying value, it means that we need to further select the satisfying matching rate. Therefore, we sort the satisfying matching rates from largest to smallest and select the first and second satisfying matching rates as candidate matching rates for later use.

[0141] S3354: Calculate the difference between the matching rates of two candidate matches and use it as the matching difference.

[0142] The matching difference refers to the difference between the matching rates of two candidate matches.

[0143] Calculating the matching difference facilitates subsequent use.

[0144] S3355: Determines the image brightness value based on image detection information.

[0145] Among them, the image brightness value is a quantitative indicator that reflects the overall brightness of the acquired image.

[0146] By converting the image detection information into a grayscale image, and then using the grayscale mean calculation method to determine the image brightness value, it is convenient for subsequent use.

[0147] S3356: Determine the brightness adaptation matching rate by combining the image brightness value and the matching difference, and use the brightness adaptation matching rate as the selection matching rate.

[0148] Among them, the brightness adaptation matching rate refers to the matching rate that best matches the actual image detection scenario, selected from the candidate matching rates based on brightness.

[0149] By combining and analyzing the image brightness value with the matching difference, the brightness adaptation matching rate is determined, and the brightness adaptation matching rate is used as the selection matching rate to improve the accuracy of the obtained selection matching rate.

[0150] To further ensure the reasonableness of the brightness adaptation matching rate, it is necessary to perform a further separate analysis and calculation of the brightness adaptation matching rate, which will be explained in detail through the following steps.

[0151] The method for determining the brightness adaptation matching rate includes the following steps:

[0152] S33561: Determine the matching benchmark difference based on the single-sided benchmark matching rate.

[0153] The matching baseline difference refers to the critical threshold used to determine the magnitude of the difference between two candidate matching rates. The higher the single-sided baseline matching rate, the greater the tolerance for differences in candidate matching rates.

[0154] The product of the single-sided reference matching rate and the preset difference coefficient is calculated, and the calculation result is used as the matching reference difference for convenient subsequent use.

[0155] The difference coefficient is a coefficient used to convert the single-sided reference matching rate into the matching reference difference. The difference coefficient is preset by the operator according to actual needs.

[0156] S33562: Determine whether the matching difference is greater than the matching baseline difference. If yes, proceed to S33563; otherwise, proceed to S33564.

[0157] Specifically, by judging whether the matching difference is greater than the matching benchmark difference, it is determined whether the candidate matching rate can be directly selected.

[0158] S33563: Select the larger candidate matching rate and use it as the brightness adaptation matching rate.

[0159] When the matching difference is greater than the matching benchmark difference, it means that the candidate matching rate can be directly selected. Therefore, the candidate matching rates are compared and the larger candidate matching rate is selected as the brightness adaptation matching rate to improve the accuracy of the obtained brightness adaptation matching rate.

[0160] S33564: Determine candidate single facets based on candidate matching rate.

[0161] Among them, candidate single face refers to the single face of the workpiece corresponding to the candidate matching rate in the workpiece requirement specification.

[0162] When the matching difference is not greater than the matching benchmark difference, it means that the candidate matching rate cannot be selected directly. Therefore, the single-sided feature of the corresponding workpiece is retrieved through the candidate matching rate, and the single-sided feature corresponding to the single-sided feature of the workpiece is taken as the candidate single-sided feature for convenient use later.

[0163] S33565: Determine the brightness influence range of a single surface based on the candidate single surface.

[0164] Among them, the single-sided brightness influence range refers to the brightness range that affects the identification of a candidate single side.

[0165] Because different candidate single-sided surfaces correspond to different material types (such as aluminum alloy, carbon steel), surface treatment processes (such as polishing, sanding, spraying), and geometric shapes (such as flat surfaces, curved surfaces, stepped surfaces), they correspond to different single-sided brightness influence ranges.

[0166] By inputting candidate single-sided surfaces into a preset single-sided brightness database, the influence range of single-sided brightness is obtained for easy subsequent use.

[0167] The single-sided brightness database has a pre-stored lookup table of different candidate single sides and their corresponding single-sided brightness influence ranges. The single-sided brightness database is obtained after the operator pre-inputs the data.

[0168] S33566: Determine the brightness reference value by combining the image brightness value with the single-sided brightness influence range.

[0169] Among them, the brightness reference value refers to the reference value that quantifies the degree of influence of brightness. The smaller the brightness reference value, the smaller the influence of brightness on the identification of candidate single face.

[0170] By analyzing whether the image brightness value falls within the single-sided brightness influence range, when it does not fall within the range, the brightness reference value is less than 1, and the larger the difference between the image brightness value and the nearest value of the single-sided brightness influence range, the smaller the corresponding brightness reference value. When it falls within the range, the brightness reference value is greater than 1, and the larger the difference between the image brightness value and the nearest value of the single-sided brightness influence range, the larger the corresponding brightness reference value.

[0171] S33567: Select the candidate matching rate corresponding to the smaller brightness reference value and use it as the brightness adaptation matching rate.

[0172] The accuracy of the obtained brightness adaptation matching rate is improved by sorting the brightness reference values ​​from smallest to largest and selecting the candidate matching rate corresponding to the first ranked brightness reference value.

[0173] S336: Collect historical requirements specifications.

[0174] Among them, historical requirement specifications refer to the design standards that workpieces ground in the past must meet after processing.

[0175] When the feature matching rate is not greater than the single-sided reference matching rate, it means that the workpiece recognized by the image at this time is not the workpiece that needs to be ground at present. Therefore, the workpiece requirement specifications in the historical time are retrieved and used as historical requirement specifications for convenient use in the future.

[0176] S337: Combine image contour features, image defect features and historical requirement specifications to determine the historical specification facet, and use the historical specification facet as the feature matching facet.

[0177] Among them, the historical specification single side refers to the single side corresponding to the workpiece after matching and screening based on the image and the historical time of grinding.

[0178] By combining and analyzing image contour features, image defect features, and historical requirement specifications, the historical specification facet is determined and used as the feature matching facet, thereby improving the accuracy of the obtained feature matching facet.

[0179] To further ensure the rationality of the historical specification single side, it is necessary to conduct further separate analysis and calculation on the historical specification single side, which will be explained in detail through the following steps.

[0180] The method for determining the historical single-sided specification includes the following steps:

[0181] S3371: Retrieve the historical specification outline features and historical specification defect features of each historical single side based on historical demand specifications.

[0182] Among them, "historical single-sided surface" refers to a single side of the workpiece corresponding to the historical requirement specifications. "Historical specification contour features" refers to the set of baseline visual features in the historical single-sided surface that reflect the shape, geometric dimensions, and other external attributes of the workpiece's single-sided edge. "Historical specification defect features" refers to the set of visual baseline features in the historical single-sided surface that characterize the location, shape, size, and quantity of defects such as scratches, burns, pinholes, and burrs on the workpiece surface. Historical requirement specifications include the historical specification contour features and historical specification defect features of each historical single-sided surface.

[0183] By retrieving the historical specification outline features and historical specification defect features of each historical single-sided component through historical requirement specifications, it is convenient for subsequent use.

[0184] S3372: Analyze the similarity between historical specification contour features and image contour features to obtain contour similarity reference values.

[0185] Among them, the contour similarity reference value refers to a quantitative index that quantifies the degree of similarity in geometric shape and topological structure between the contour features of historical specifications and the contour features of the current image.

[0186] By analyzing the similarity between historical specification contour features and image contour features, contour similarity reference values ​​are obtained for subsequent use. The specific analysis method for contour similarity reference values ​​is described in S32 and will not be elaborated further here.

[0187] S3373: Analyze the similarity between historical specification defect features and image defect features to obtain defect similarity reference values.

[0188] Among them, the defect similarity reference value refers to the core indicator that quantifies the similarity between historical specification defect features and current image defect features in terms of defect attributes.

[0189] By comparing historical specification defect features with image defect features according to three core dimensions—defect type (such as consistency of categories like scratches, burns, and pinholes), dimensional parameters (such as the degree of deviation in the width, area, and depth of defects), and distribution patterns (such as whether the defect is located in the edge area or the core mating area, and the difference in the number of defects in a single area)—the similarity of each dimension is obtained. Then, weighted coefficients are assigned to each dimension based on the degree of impact of the defect on the workpiece quality, with the defect type matching degree having the highest weight (e.g., 0.4), followed by dimensional parameters (e.g., 0.3), and the distribution pattern having a weight of 0.3. Finally, the similarity of each dimension is multiplied by its corresponding weight and summed to obtain the final defect similarity reference value.

[0190] S3374: Determine the comprehensive similarity reference value by combining the contour similarity reference value and the defect similarity reference value.

[0191] Among them, the similarity comprehensive reference value refers to the reference value corresponding to the comprehensive analysis of the contour and defects.

[0192] By weighting the contour similarity reference value and the defect similarity reference value, and using the result as the comprehensive similarity reference value, it is convenient for subsequent use. The weighting coefficients can be preset by the operator.

[0193] To further ensure the rationality of the similarity comprehensive reference value, it is necessary to perform a further separate analysis and calculation on the similarity comprehensive reference value, which will be explained in detail through the steps shown below.

[0194] The method for determining the similarity comprehensive reference value includes the following steps:

[0195] S33741: Calculate the difference between the contour similarity reference value and the defect similarity reference value and use it as the similarity reference deviation value.

[0196] The similarity reference deviation value refers to the difference between the contour similarity reference value and the defect similarity reference value.

[0197] Calculating the similarity reference deviation value facilitates subsequent use.

[0198] S33742: Collect the current time point.

[0199] The current time point refers to the specific point in time corresponding to the current moment. This current time point is obtained by querying a time database. The time database keeps track of time in real time and stores the data.

[0200] S33743: Retrieve historical grinding time range based on historical requirement specifications.

[0201] The historical grinding time interval refers to the time range during which a grinding process was performed on a corresponding single side in the past. Historical requirement specifications include the historical grinding time interval.

[0202] Historical grinding time ranges can be retrieved by using historical requirement specifications, making it convenient for subsequent use.

[0203] S33744: Determine the time reference value by combining the current time point with the historical grinding time interval.

[0204] Among them, the time reference value is the core indicator that quantifies the degree of matching between the current time point and the historical grinding time interval.

[0205] By combining the current time point with the historical grinding time interval, a time reference value is determined for convenient subsequent use.

[0206] To further ensure the rationality of the time reference value, it is necessary to perform further separate analysis and calculation on the time reference value, which will be explained in detail through the steps shown below.

[0207] The method for determining the time reference value includes the following steps:

[0208] S337441: Determine if the current time point falls within the historical grinding time interval. If yes, execute S337442; if no, execute S337444.

[0209] Specifically, by determining whether the current time point falls within the historical grinding time range, it can be determined whether the historical grinding time range can be directly used for reference analysis.

[0210] S337442: Use the historical grinding time interval into which the current time point falls as the falling time interval.

[0211] The time interval to which the current time point falls refers to the historical grinding time interval to which the current time point falls.

[0212] If there exists a historical grinding time interval that the current time point falls into, it means that the historical grinding time interval can be directly used for reference analysis. Therefore, the time interval that falls into is defined to facilitate its subsequent use.

[0213] S337443: Determine the landing reference value based on the landing time interval, and use the landing reference value as the time reference value.

[0214] Among them, the reference value refers to the core indicator of the degree of working condition matching quantified based on the time interval of the fall.

[0215] By analyzing the time intervals in which the data falls, a reference value for the data falls is determined, and this reference value is used as a time reference value to improve the accuracy of the obtained time reference value.

[0216] To further ensure the reasonableness of falling within the reference value, it is necessary to perform further separate analysis and calculation on the reference value, which will be explained in detail through the steps shown below.

[0217] The method for determining whether a value falls within the reference range includes the following steps:

[0218] S3374431: Determine the reference value for a single interval and the number of values ​​for the interval based on the time interval in which the value falls.

[0219] Here, a single-interval reference value refers to the reference value corresponding to a single time interval. The number of values ​​within a time interval refers to the number of values ​​within that time interval. Each time interval corresponds to its own unique single-interval reference value.

[0220] By counting the time intervals that fall into the range, and using the count results as the number of intervals that fall into the range, and by inputting the time intervals into a preset time interval database to obtain a reference value for each interval, it is convenient for subsequent use.

[0221] The time interval database pre-stores a table of different time intervals and their corresponding single-interval reference values. The time interval database is pre-set by the operator according to their needs.

[0222] For example, the time interval database can be set such that when the time interval falls within 8:00-12:00, the single interval reference value is 1; when the time interval falls within 13:00-17:00, the single interval reference value is 2; and when the time interval falls within 18:00-20:00, the single interval reference value is 3.

[0223] S3374432: Determine whether there is only one value falling within the interval. If yes, execute S3374433; if no, execute S3374434.

[0224] In this process, by determining whether there is only one value falling within the interval, it can be determined whether further analysis of the reference value for a single interval is needed.

[0225] S3374433: Use the single-interval reference value as the reference value for falling into the range.

[0226] When only one value falls within the interval, it means that no further analysis of the single interval reference value is needed, so the single interval reference value is used as the falling reference value.

[0227] S3374434: Calculate the midpoint of the time interval into which the fall occurs and use it as the midpoint of the time interval into the fall.

[0228] The midpoint of the time interval refers to the midpoint of the time interval.

[0229] When there is more than one value falling into the interval, it indicates that further analysis of the reference value of the single interval is needed. Therefore, the midpoint of the falling time is calculated for convenient use later.

[0230] S3374435: Calculate the deviation between the midpoint of the landing time and the current time point and use it as the midpoint deviation value.

[0231] The intermediate deviation value refers to the deviation between the midpoint of the time point and the current time point.

[0232] Calculating the intermediate deviation value facilitates subsequent use.

[0233] S3374436: Determine the intermediate selection coefficient based on the intermediate deviation value.

[0234] The intermediate selection coefficient refers to the coefficient used in the weighted calculation of reference values ​​within a single interval. Different intermediate deviation values ​​correspond to different intermediate selection coefficients.

[0235] The intermediate deviation value is input into a preset intermediate selection database to obtain the intermediate selection coefficient, which is convenient for subsequent use.

[0236] The intermediate selection database pre-stores a table of different intermediate deviation values ​​and their corresponding intermediate selection coefficients. The intermediate selection database is pre-set by the operator according to actual needs.

[0237] S3374437: Combine the intermediate selection coefficient with the single interval reference value to determine the intermediate selection reference value, and use the intermediate selection reference value as the falling reference value.

[0238] Among them, the intermediate selection reference value refers to the quantitative indicator obtained by weighting and correcting the single-interval reference value through the intermediate selection coefficient.

[0239] By multiplying each single-interval reference value with its corresponding intermediate selection coefficient, summing the results, and using the summation as the intermediate selection reference value, the accuracy of the obtained entry reference value is improved.

[0240] S337444: Calculate the difference between the current time point and the historical grinding time interval and use it as the time deviation value.

[0241] The time deviation value refers to the difference between the current time point and the historical grinding time interval.

[0242] If there is a historical grinding time interval that the current time point does not fall into, it means that the historical grinding time interval cannot be directly used for reference analysis. Therefore, the time deviation value is calculated to facilitate subsequent use.

[0243] S337445: Sort the time deviation values ​​from smallest to largest, and take the historical grinding time interval corresponding to the first time deviation value as the deviation time interval.

[0244] Among them, the deviation time interval refers to the historical grinding time interval corresponding to the smallest deviation.

[0245] By sorting the time deviation values ​​from smallest to largest, and taking the historical grinding time interval corresponding to the first-ranked time deviation value as the deviation time interval, it is convenient for subsequent use.

[0246] S337446: Determine the deviation reference value based on the deviation time interval, and use the deviation reference value as the time reference value.

[0247] The deviation reference value refers to the reference value corresponding to the deviation time interval.

[0248] By inputting the deviation time interval into a preset time interval database, a deviation reference value is obtained for easy subsequent use.

[0249] S33745: Determine the time reference coefficient based on the time reference value.

[0250] Among them, the time reference coefficient refers to the quantization parameter that maps the time reference value to the weighted calculation coefficient.

[0251] By inputting the time reference value into a preset time reference database, a time reference coefficient is obtained for easy subsequent use.

[0252] The time reference database pre-stores a table of different time reference values ​​and their corresponding time reference coefficients. The time reference database is pre-set by the operator according to actual needs.

[0253] S33746: Based on the time reference coefficient, the contour similarity reference value and the defect similarity reference value are weighted and calculated to obtain the time comprehensive reference value, and the time comprehensive reference value is used as the similarity comprehensive reference value.

[0254] Among them, the time-based comprehensive reference value refers to the comprehensive reference value after being weighted and corrected based on time.

[0255] By using the time reference coefficient as a weighting coefficient and weighting the contour similarity reference value and the defect similarity reference value, a time comprehensive reference value is obtained. This time comprehensive reference value is then used as the similarity comprehensive reference value, thereby improving the accuracy of the obtained similarity comprehensive reference value.

[0256] For example, when the time reference factor is 0.6, the overall time reference value is 0.6 * profile similarity reference value + 0.4 * defect similarity reference value.

[0257] S3375: Sort the similar comprehensive reference values ​​from largest to smallest, and take the historical single face of the historical requirement specification corresponding to the first-ranked similar comprehensive reference value as the historical selected single face, and take the historical selected single face as the historical specification single face.

[0258] Among them, the historical selected single side refers to the single side corresponding to the selection of the historical single side in the historical requirement specification.

[0259] By sorting the similar comprehensive reference values ​​from largest to smallest, and taking the historical single-sided page of the historical requirement specification corresponding to the first-ranked similar comprehensive reference value as the historical selected single-sided page, and taking the historical selected single-sided page as the historical specification single-sided page, the accuracy of obtaining the historical specification single-sided page is improved.

[0260] S34: Determine the baseline features of single-sided defects based on the image prediction of a single side.

[0261] Among them, the single-sided defect benchmark features refer to the set of benchmark attributes used to determine whether a single-sided defect is qualified. The single-sided defect benchmark features include three core elements: defect type limitation, size threshold, and distribution range.

[0262] By retrieving the single-sided defect reference features from the workpiece requirements specifications based on the image prediction of the single side, it is convenient for subsequent use.

[0263] S35: Combine image defect features with single-sided defect baseline features to determine the defect prediction status, and use the defect prediction status as the single-sided quality status.

[0264] Among them, the defect prediction status refers to the quantitative judgment result of whether the current single-sided defect meets the quality requirements.

[0265] By comparing the image defect features with the single-sided defect baseline features one by one, similarity values ​​corresponding to the defect type, size parameters, and distribution patterns are obtained. The similarity values ​​of each dimension are then weighted and calculated as the defect prediction, and the defect prediction is used as the single-sided quality status, thereby improving the accuracy of the obtained single-sided quality status.

[0266] For specific analysis methods of defect prediction, please refer to S3373, which will not be elaborated here.

[0267] S4: Determine the single-sided detection distance value based on the laser detection information.

[0268] Among them, the single-sided detection distance value refers to the actual vertical distance between the laser emitter and the single side of the workpiece to be detected.

[0269] The distance detection value is retrieved by laser detection information and used as the single-sided detection distance value for convenient subsequent use.

[0270] S5: Determine the quality adjustment value based on the single-sided quality.

[0271] The quality adjustment value refers to a quantitative adjustment parameter set based on the grade determination result of the single-sided quality condition. Different single-sided quality conditions correspond to different quality adjustment values. The higher the single-sided quality condition, the more defects there are, and therefore the larger the quality adjustment value.

[0272] By inputting the single-sided quality information into a preset quality adjustment database, a quality adjustment value is obtained for convenient subsequent use.

[0273] The quality adjustment database pre-stores a table showing the correspondence between different single-sided quality conditions and their corresponding quality adjustment values. The quality adjustment database is pre-set by the operator according to actual needs.

[0274] S6: Combine the single-sided detection distance value, quality adjustment value and single-sided required distance value to determine the operation control parameters, and output the operation control parameters to the grinding device.

[0275] Among them, the operating control parameters refer to the feed depth used to guide the grinding device in grinding and cutting.

[0276] The initial adjustment value is obtained by calculating the difference between the single-sided detection distance value and the single-sided required distance value. Then, the product value between the initial adjustment value and the quality adjustment value is calculated and used as the final adjustment value. The final adjustment value is then used as the operation control parameter and output to the grinding device. This automatically and accurately matches the operation control parameter with the actual state and required specifications of the workpiece, ensuring that the workpiece meets the required specifications after grinding and improving the overall grinding efficiency when grinding a large number of workpieces.

[0277] Based on the same inventive concept, embodiments of the present invention provide a near-net-shape grinding allowance control system, comprising:

[0278] The data acquisition module is used to collect laser detection information, image detection information, workpiece requirements and specifications, historical requirements and specifications, and the current time point.

[0279] The memory stores a program for implementing a near-net-shape grinding allowance control method as described above;

[0280] The processor loads and executes programs stored in memory.

[0281] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0282] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A near-net-shape grinding allowance control method, characterized by, include: S1: Collect laser detection information, image detection information and workpiece specifications required for the workpiece on the grinding device; S2: Retrieve the single-sided features and single-sided distance values ​​of the workpiece based on the workpiece requirements and specifications; S3: Combine the single-sided features of the workpiece with image detection information to determine the image prediction of the single-sided surface and the single-sided quality. S4: Determine the single-sided detection distance value based on the laser detection information; S5: Determine the quality adjustment value based on the single-sided quality; S6: Combine the single-sided detection distance value, quality adjustment value and single-sided required distance value to determine the operation control parameters, and output the operation control parameters to the grinding device; Methods for predicting a single-sided image and determining its quality include: S31: Image contour features and image defect features are obtained based on image detection information; S32: Determine the feature matching rate based on the consistency between the image contour features and the single-sided features of the workpiece; S33: Determine the feature matching facet based on the feature matching rate, and use the feature matching facet as the image prediction facet; S34: Determine the baseline features of single-sided defects based on the image prediction of a single side; S35: Combine image defect features with single-sided defect baseline features to determine the defect prediction status, and use the defect prediction status as the single-sided quality status; Methods for determining one side of feature matching include: S331: Retrieve single-sided reference matching rate based on workpiece requirements and specifications; S332: Determine whether the feature matching rate is greater than the single-sided baseline matching rate; S333: If yes, then select the feature matching rate that is greater than the single-sided baseline matching rate and use it as the matching rate. S334: Retrieve the number of values ​​that satisfy the matching rate; S335: Select the matching rate based on the number of satisfied values ​​and use it as the selected matching rate, and use the single side of the workpiece single-sided feature corresponding to the selected matching rate as the feature matching single side. S336: If not, collect historical requirement specifications; S337: Combine image contour features, image defect features and historical requirement specifications to determine the historical specification facet, and use the historical specification facet as the feature matching facet.

2. The near-net-shape grinding allowance control method according to claim 1, characterized in that, Methods for selecting the matching rate include: S3351: Determine whether the value that satisfies the condition is only one; S3352: If yes, then the matching rate is directly used as the selection matching rate; S3353: If not, select the two highest matching rates as candidate matching rates; S3354: Calculate the difference between the matching rates of two candidate matches and use it as the matching difference; S3355: Determines the image brightness value based on image detection information; S3356: Determine the brightness adaptation matching rate by combining the image brightness value and the matching difference, and use the brightness adaptation matching rate as the selection matching rate.

3. The near-net-shape grinding allowance control method according to claim 2, characterized in that, Methods for determining the brightness adaptation matching rate include: S33561: Determine the matching benchmark difference based on the single-sided benchmark matching rate; S33562: Determine whether the matching difference is greater than the matching baseline difference; S33563: If yes, select the larger candidate matching rate as the brightness adaptation matching rate; S33564: If not, then determine the candidate single face based on the candidate matching rate; S33565: Determine the brightness influence range of a single surface based on the candidate single surface; S33566: Determine the brightness reference value by combining the image brightness value with the single-sided brightness influence range; S33567: Select the candidate matching rate corresponding to the smaller brightness reference value and use it as the brightness adaptation matching rate.

4. The near-net-shape grinding allowance control method according to claim 1, characterized in that, Methods for determining the historical single-sided specification include: S3371: Retrieve the historical specification outline features and historical specification defect features of each historical single-sided document based on historical requirement specifications. S3372: Analyze the similarity between historical specification contour features and image contour features to obtain contour similarity reference values; S3373: Analyze the similarity between historical specification defect features and image defect features to obtain defect similarity reference values; S3374: Determine the comprehensive similarity reference value by combining the contour similarity reference value and the defect similarity reference value; S3375: Sort the similar comprehensive reference values ​​from largest to smallest, and take the historical single face of the historical requirement specification corresponding to the first-ranked similar comprehensive reference value as the historical selected single face, and take the historical selected single face as the historical specification single face.

5. The near-net-shape grinding allowance control method according to claim 4, characterized in that, The methods for determining similarity comprehensive reference values ​​include: S33741: Calculate the difference between the contour similarity reference value and the defect similarity reference value and use it as the similarity reference deviation value; S33742: Collect the current time point; S33743: Retrieve historical grinding time intervals based on historical requirement specifications; S33744: Determine the time reference value by combining the current time point with the historical grinding time range; S33745: Determine the time reference coefficient based on the time reference value; S33746: Based on the time reference coefficient, the contour similarity reference value and the defect similarity reference value are weighted and calculated to obtain the time comprehensive reference value, and the time comprehensive reference value is used as the similarity comprehensive reference value.

6. The near-net-shape grinding allowance control method according to claim 5, characterized in that, Methods for determining time reference values ​​include: S337441: Determine whether the current time point falls within the historical grinding time interval; S337442: If yes, then the historical grinding time interval into which the current time point falls shall be taken as the time interval into which it falls; S337443: Determine the landing reference value based on the landing time interval, and use the landing reference value as the time reference value; S337444: If not, calculate the difference between the current time point and the historical grinding time interval and use it as the time deviation value; S337445: Sort the time deviation values ​​from smallest to largest, and take the historical grinding time interval corresponding to the first time deviation value as the deviation time interval. S337446: Determine the deviation reference value based on the deviation time interval, and use the deviation reference value as the time reference value.

7. The near-net-shape grinding allowance control method according to claim 6, characterized in that, Methods for determining whether a value falls within the reference range include: S3374431: Determine the reference value for a single interval and the number of values ​​for the interval based on the time interval in which the value falls; S3374432: Determine whether the number of values ​​falling into the interval is only one; S3374433: If yes, then the single-interval reference value will be used as the reference value to fall into; S3374434: If not, calculate the midpoint of the time interval into which the time falls and use it as the midpoint of the time interval into which the time falls. S3374435: Calculate the deviation between the midpoint of the landing time and the current time point and use it as the midpoint deviation value; S3374436: Determine the intermediate selection coefficient based on the intermediate deviation value; S3374437: Combine the intermediate selection coefficient with the single interval reference value to determine the intermediate selection reference value, and use the intermediate selection reference value as the falling reference value.

8. A near-net-shape grinding allowance control system, characterized in that, include: The data acquisition module is used to collect laser detection information, image detection information, workpiece requirements and specifications, historical requirements and specifications, and the current time point. The memory stores a program for implementing a near-net-shape grinding allowance control method as described in any one of claims 1 to 7; The processor loads and executes programs stored in memory.

Images