Data processing method and system based on EXCEL logic

Abstract

The present invention relates to the field of data processing technology, and discloses a data processing method and system based on EXCEL logic, which obtains point cloud data of M angles of a target mold, and allocates M EXCEL tables, wherein one cell of each EXCEL table corresponds to one point cloud data of one angle; the M EXCEL tables are analyzed based on a preset mold quality model to obtain problem cells; and problem cells are visually marked based on a preset EXCEL command. The present invention converts complex mold quality inspection data into intuitive problem visualization based on an EXCEL table, which not only simplifies the complexity of relying on third-party professional software and manual analysis in the traditional inspection process, but also significantly improves the efficiency and intuitiveness of information transmission, making the analysis and management of mold quality problems more efficient and easy to operate.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a data processing method and system based on EXCEL logic. Background Art

[0002] The management tools of existing mold companies are usually developed based on EXCEL to record and analyze production data. However, mold quality inspection results usually rely on third-party professional software. The inspection results of such software often need to be manually summarized and recorded and uploaded to the management tool, resulting in low efficiency and lack of intuitive information transmission. The current recording method only describes in text that there is a quality problem in a certain part of a mold, lacking an intuitive reflection of the mold quality problem.

[0003] In addition, third-party professional tools usually require professional operation and are expensive to use. This not only increases operating costs but also increases process complexity for small and medium-sized mold companies. Existing technologies therefore have problems such as low information transmission efficiency, non-intuitive problem reflection, and high cost of use. Summary of the invention

[0004] The present invention provides a data processing method and system based on EXCEL logic to solve the technical problems raised in the background technology.

[0005] The present invention provides a data processing method based on EXCEL logic, comprising:

[0006] Step 1: Obtain the point cloud data of M angles of the target mold and allocate M EXCEL tables. One cell of each EXCEL table corresponds to one point cloud data of one angle, including:

[0007] Establish a target plane, and obtain a central point cloud of point cloud data at an angle based on the target plane;

[0008] Based on the central point cloud, a topological structure is constructed for the point cloud data at one angle, and clustering is performed based on the topological structure to obtain an updated topological structure, and the point cloud data is filled into the corresponding cells of the EXCEL table according to the updated topological structure;

[0009] Step 2: Analyze M EXCEL tables based on the preset mold quality model to obtain problematic cells;

[0010] Step 3: Based on the preset EXCEL command, visually mark the problem cells.

[0011] Furthermore, the point cloud data of M angles of the target mold are obtained, including:

[0012] By scanning multiple angles of the target mold with radar, point cloud data of each angle of the target mold is obtained; wherein each angle covers the entire surface of the corresponding angle of the target mold.

[0013] Furthermore, each cell of each EXCEL table corresponds to a point cloud data of one angle, including:

[0014] Step 4, obtain the maximum contact surface of the point cloud data at the i-th angle as the target plane; the target plane calculation formula is as follows:

[0015] ;in, Represents the x-axis component, y-axis component, and z-axis component of the normal vector of the target plane. , represents the intercept of the target plane, represents the contact distance threshold of the target plane, Represents the total number of point clouds at the i-th angle, express The index of Indicates the point cloud data of the i-th angle The x-axis coordinate, y-axis coordinate and z-axis coordinate of the point cloud, Express satisfaction If yes, it is 1, otherwise it is 0;

[0016] Select the point cloud that does not touch the edge of the target plane , , and project it vertically into the target plane to obtain the mapping point cloud; the projection formula is as follows:

[0017] ;

[0018] in, express The x-axis coordinate, y-axis coordinate and z-axis coordinate of the mapped point cloud, express The x-axis coordinate, y-axis coordinate and z-axis coordinate of

[0019] Connect the edges of the combined point cloud constructed by the mapped point cloud and the non-mapped point cloud on the target plane to obtain a geometric figure, obtain the center of the geometric figure by the centroid method, and use the point cloud with the smallest Euclidean distance to the center of the geometric figure as the central point cloud of the point cloud data at i angles;

[0020] Step 5, fill the center point cloud into the center table of the i-th EXCEL table;

[0021] Taking the central point cloud as the root node, establish the topological structure of the point cloud data at the i-th angle , ,in, represents a 1-degree node, express Degree node, represents a node of degree n, ;

[0022] Step 6, Degree Node Assignment Cluster centers are found, and clustering is performed according to the position coordinates of the point cloud to obtain the updated topological structure of the point cloud data at the i-th angle;

[0023] Step 7, the 1-degree to n-degree nodes of the updated topology structure are sequentially filled with cells around the central point cloud.

[0024] Furthermore, the mold quality model includes:

[0025] Within a preset time period, point cloud data of the target mold at M angles are obtained, and a three-dimensional model of the target mold is constructed based on the point cloud data. If there is a problem area inside the three-dimensional model, feature images of the three-dimensional model at M angles are obtained, and the feature images include the problem area;

[0026] Based on the point cloud data of the target mold at M angles, M EXCEL tables of the target mold are obtained through steps 4 to 7, and the feature images of the M angles of the 3D model are fitted with the M EXCEL tables. The problem cells in each EXCEL are determined according to the problem area, and M feature EXCEL tables are obtained;

[0027] The EXCEL table is used as a training sample, and the corresponding feature EXCEL table is used as a sample label to train a mold quality model. The output of the mold quality model represents the predicted problem cell in the EXCEL table;

[0028] The hyperparameters of the hidden layer of the mold quality model are back-propagated and updated through the preset loss function.

[0029] Furthermore, the characteristic images of the three-dimensional model at M angles are fitted correspondingly to M EXCEL tables, including:

[0030] By scaling the feature image at the i-th angle, a scaled image is obtained; wherein the number of rows and columns of the scaled image are equal to the number of rows and columns of the EXCEL table at the i-th angle;

[0031] Overlap the zoomed image of the i-th angle with the EXCEL table to obtain cells that overlap with the problem area in the zoomed image, and set the corresponding cells as problem cells.

[0032] Further, the mold quality model includes a first hidden layer, a second hidden layer, a third hidden layer and an output layer;

[0033] The calculation formula for the first hidden layer is as follows:

[0034] ;

[0035] in, represents the first vector of the output of the first hidden layer, Represents the vector in the cell at row g and column h in the EXCEL table. , and denote the first learning parameter, the second learning parameter and the third learning parameter respectively, represents element-wise multiplication, represents a multi-layer perceptron;

[0036] The calculation formula for the second hidden layer is as follows:

[0037] ;

[0038] in, The second vector representing the output of the second hidden layer, represents the sliding window between rows, represents the index of the sliding window between rows, represents a sliding window between columns, Represents the index of the sliding window between columns;

[0039] The calculation formula for the third hidden layer is as follows:

[0040] ;

[0041] in, A third vector representing the output of the third hidden layer;

[0042] The calculation formula of the output layer is as follows:

[0043] ;

[0044] in, represents the output of the output layer, Indicates the cosine similarity. represents the standard threshold, Represents the similarity threshold.

[0045] Furthermore, the loss function of the mold quality model is The details are as follows:

[0046] The mold quality model is used to predict the problem cell of the EXCEL table output of the M angles of the target mold, and the predicted cell is used as a feature point. The feature point is extended according to the angle of the corresponding EXCEL table to obtain Focus, count the number of feature points connected to each focus;

[0047] Loss Function The calculation formula is as follows:

[0048] ;

[0049] in, and denote the first loss coefficient and the second loss coefficient respectively, , express The index of Indicates the number of rows of cells in the EXCEL table. express The index of Indicates the number of columns of cells in the EXCEL table. express The index of The first part of the EXCEL table showing the output of the mold quality model Line Column elements, Indicates the characteristics of the EXCEL table Line Column elements, Indicates the number of focal points, express The index of Indicates The number of feature points connected by each focus.

[0050] Further, the preset EXCEL commands include:

[0051] Based on EXCEL's VBA editor, create macro commands. The macro commands are as follows:

[0052] Declare variables: declare a variable ws of Worksheet type;

[0053] Declare a variable cell of the Range type, which is used to traverse the cells of the EXCEL table in the loop;

[0054] Set the EXCEL table: Set the ws variable to the EXCEL table in "Sheet1" of the current workbook;

[0055] Traverse cells: Use ForEach loop to traverse all cells in the used range of the EXCEL table of ws;

[0056] Conditional judgment and formatting: Inside the loop, conditional judgment is performed on each cell: if the value of the cell is equal to 1, the background color of the cell is set to red (RGB value is 255, 0, 0);

[0057] End loop and macro: When the loop ends, the execution of the macro also ends;

[0058] The predicted problem cells output by the mold quality model are stored in a database, and the predicted problem cells are extracted from the database into EXCEL through SQL query, and macro commands are executed to obtain visual marks of the predicted problem cells in the EXCEL table;

[0059] The predicted problem cells of the mold quality model are stored in a database, and the predicted problem cells are extracted from the database into EXCEL through SQL query, and macro commands are executed to obtain visual marks of the predicted problem cells in the EXCEL table.

[0060] The data processing system based on EXCEL logic is applied to the method described, including:

[0061] The data processing module obtains the point cloud data of M angles of the target mold and allocates M EXCEL tables, where one cell of each EXCEL table corresponds to one point cloud data of one angle;

[0062] The data analysis module analyzes M EXCEL tables based on the preset mold quality model to obtain the problem cells;

[0063] The table marking module visually marks problem cells based on preset EXCEL commands.

[0064] The beneficial effect of the present invention is that by converting complex mold quality inspection data into intuitive problem visualization based on EXCEL tables, it not only simplifies the complexity of traditional inspection processes that rely on third-party professional software and manual analysis, but also significantly improves the efficiency and intuitiveness of information transmission, making the analysis and management of mold quality problems more efficient and easy to operate. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a flow chart of the data processing method based on EXCEL logic of the present invention. DETAILED DESCRIPTION

[0066] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0067] like Figure 1 As shown, the data processing method based on EXCEL logic includes:

[0068] Step 1: Obtain the point cloud data of M angles of the target mold and allocate M EXCEL tables. One cell of each EXCEL table corresponds to one point cloud data of one angle, including:

[0069] Establish a target plane, and obtain a central point cloud of point cloud data at an angle based on the target plane;

[0070] Based on the central point cloud, a topological structure is constructed for the point cloud data at one angle, and clustering is performed based on the topological structure to obtain an updated topological structure, and the point cloud data is filled into the corresponding cells of the EXCEL table according to the updated topological structure;

[0071] Step 2: Analyze M EXCEL tables based on the preset mold quality model to obtain problematic cells;

[0072] Step 3: Based on the preset EXCEL command, visually mark the problem cells.

[0073] Specifically, the present invention combines an EXCEL table used to display two-dimensional information with multiple angles so that each EXCEL table can reflect the quality problem of a target mold at one angle, and uses visual markings on cells to enable the EXCEL table to intuitively display the quality problem of the corresponding angle. The EXCEL tables of multiple angles of the target mold are packaged into a file package, so that a management system developed based on EXCEL can directly view the quality problem of the corresponding target mold, which is convenient for the person in charge to intuitively obtain the quality information of the mold through the management system rather than obtaining the quality information of the mold through text description, thereby improving the efficiency of information transmission.

[0074] In one embodiment of the present invention, obtaining point cloud data of M angles of a target mold includes:

[0075] By scanning multiple angles of the target mold with radar, point cloud data of each angle of the target mold is obtained; wherein each angle covers the entire surface of the corresponding angle of the target mold.

[0076] Specifically, the entire surface of the target mold corresponding to the angle is obtained to facilitate subsequent processing, which is otherwise impossible to achieve.

[0077] In one embodiment of the present invention, one cell of each EXCEL table corresponds to one point cloud data of one angle, including:

[0078] Step 4, obtain the maximum contact surface of the point cloud data at the i-th angle as the target plane; the target plane calculation formula is as follows:

[0079] ;in, Represents the x-axis component, y-axis component, and z-axis component of the normal vector of the target plane. , represents the intercept of the target plane, represents the contact distance threshold of the target plane, Represents the total number of point clouds at the i-th angle, express The index of Indicates the point cloud data of the i-th angle The x-axis coordinate, y-axis coordinate and z-axis coordinate of the point cloud, Express satisfaction If yes, it is 1, otherwise it is 0;

[0080] Select the point cloud that does not touch the edge of the target plane , , and project it vertically into the target plane to obtain the mapping point cloud; the projection formula is as follows:

[0081] ;

[0082] in, express The x-axis coordinate, y-axis coordinate and z-axis coordinate of the mapped point cloud, express The x-axis coordinate, y-axis coordinate and z-axis coordinate of

[0083] Connect the edges of the combined point cloud constructed by the mapped point cloud and the non-mapped point cloud on the target plane to obtain a geometric figure, obtain the center of the geometric figure by the centroid method, and use the point cloud with the smallest Euclidean distance to the center of the geometric figure as the central point cloud of the point cloud data at i angles;

[0084] Step 5, fill the center point cloud into the center table of the i-th EXCEL table;

[0085] Taking the central point cloud as the root node, establish the topological structure of the point cloud data at the i-th angle , ,in, represents a 1-degree node, express Degree node, represents a node of degree n, ;

[0086] Step 6, Degree Node Assignment Cluster centers are found, and clustering is performed according to the position coordinates of the point cloud to obtain the updated topological structure of the point cloud data at the i-th angle;

[0087] Step 7, the 1-degree to n-degree nodes of the updated topology structure are sequentially filled with cells around the central point cloud.

[0088] Specifically, for the point cloud data of each angle, the face with the largest number of contact point clouds is taken as the target plane, and then the remaining point clouds are vertically mapped to the target plane. In this way, the original point cloud and the mapped point cloud are obtained on the target plane, and the original point cloud and the mapped point cloud are regarded as a whole, and then the outermost point clouds of the overall point cloud on the target plane are connected to obtain a geometric figure, and the center of mass of this geometric figure is obtained by the centroid algorithm, and the point cloud with the closest Euclidean distance to this center of mass is taken as the center point cloud. The other point clouds are centered on the center point cloud, and the other point clouds are mapped as nodes of the topological structure to obtain the topological structure. Then, the number of hops between the nodes and the center nodes in the topological structure is taken as the degree to obtain 1-degree nodes and n-degree nodes. Due to the natural limitations of EXCEL cells, there are only 8 domain cells outside each cell, 16 domain cells on the second outer side, and so on. Therefore, 8 clusters are required for 1-degree nodes, 16 clusters for 2-degree nodes, and so on. The specific clustering is as follows: for example, fill the central point cloud into the central cell of the EXCEL cell, then fill the first-degree nodes into the outer cells of the central cell, and fill the second-degree nodes into the second-outer cells of the central cell. Clustering is performed according to the structural characteristics of the cell. For example, for a 1-degree node including 10 point clouds, 8 clustering areas are divided according to the eight neighborhoods of the EXCEL cell outside the central point cloud. The average value of all point clouds in each clustering area is calculated to obtain the average parameters, and the 8 average parameters are filled into the 8 neighborhood cells of the central cell. Similarly, and so on.

[0089] In one embodiment of the present invention, the mold quality model includes:

[0090] Within a preset time period, point cloud data of the target mold at M angles are obtained, and a three-dimensional model of the target mold is constructed based on the point cloud data. If there is a problem area inside the three-dimensional model, feature images of the three-dimensional model at M angles are obtained, and the feature images include the problem area;

[0091] Based on the point cloud data of the target mold at M angles, M EXCEL tables of the target mold are obtained through steps 4 to 7, and the feature images of the M angles of the 3D model are fitted with the M EXCEL tables. The problem cells in each EXCEL are determined according to the problem area, and M feature EXCEL tables are obtained;

[0092] The EXCEL table is used as a training sample, and the corresponding feature EXCEL table is used as a sample label to train a mold quality model. The output of the mold quality model represents the predicted problem cell in the EXCEL table;

[0093] The hyperparameters of the hidden layer of the mold quality model are back-propagated and updated through the preset loss function.

[0094] Specifically, mold quality inspection usually relies on three-dimensional analysis of point cloud data and model training, but the existing methods have the problems of high cost and low efficiency in processing complex point cloud data. The present invention proposes a mold quality model based on EXCEL table logic, which predicts and calibrates the quality problem areas of the mold by converting three-dimensional point cloud data into two-dimensional table data, and then combining deep learning model training. The core goal is to achieve the automatic processing of mapping point cloud data to EXCEL tables, identifying problem cells and model prediction. Collection of point cloud data and construction of three-dimensional model. Within a preset time period, point cloud data of M angles of the mold are collected. Each angle corresponds to different surface information of the mold, for example, point cloud data is obtained by multi-view scanning technology or structured light scanning. Based on the point cloud data of M angles, the three-dimensional model of the target mold is reconstructed (the prior art will not be repeated). If there are problem areas in the three-dimensional model (such as cracks inside the mold, surface defects, etc.), these problem areas are marked as characteristic problem areas. Feature images of M angles are extracted from the three-dimensional model, and these images retain the surface features of the mold and its problem areas (such as cracks and holes). According to steps 4 to 7 of the present invention, the point cloud data of M angles are converted into M EXCEL tables. The target plane of the point cloud is constructed, the geometric center is calculated and the central point cloud is generated. The point cloud data is filled into the cells of the EXCEL table by updating the topological structure. The feature images of M angles extracted from the three-dimensional model are fitted one by one with the generated M EXCEL tables. The purpose of the corresponding fitting is to map the problem area calibrated in the feature image to the specific cell of the EXCEL table. The problem cell in each EXCEL table is determined by matching the geometric position and the point cloud coordinates. The problem cell refers to the cell corresponding to the problem area in the EXCEL table, and its value is 1 (the value of other cells is 0). After fitting, MMM feature EXCEL tables are generated, which mark the location and distribution of the problem cells as label data for subsequent model training. The generated EXCEL table (normal point cloud data) is used as the input (training sample) of the model. The corresponding feature EXCEL table (marked problem area) is used as the output (sample label) of the model. The mold quality model is implemented by a deep learning framework and includes several hidden layers. The goal of the model is to learn the mapping relationship from point cloud data (EXCEL table) to problem cell prediction. After the model training is completed, its output is the predicted problem cell location in the EXCEL table, which automatically calibrates the potential problem area in the mold. The loss function measures the difference between the predicted problem cell and the real problem cell. The loss function is used to optimize the hidden layer hyperparameters of the mold quality model, and the weights are updated through the gradient descent algorithm to gradually improve the prediction accuracy of the model.

[0095] In one embodiment of the present invention, the feature images of M angles of the three-dimensional model are fitted correspondingly to M EXCEL tables, including:

[0096] By scaling the feature image at the i-th angle, a scaled image is obtained; wherein the number of rows and columns of the scaled image are equal to the number of rows and columns of the EXCEL table at the i-th angle;

[0097] Overlap the zoomed image of the i-th angle with the EXCEL table to obtain cells that overlap with the problem area in the zoomed image, and set the corresponding cells as problem cells.

[0098] Specifically, the characteristic image of the three-dimensional model is fitted with the EXCEL table so as to accurately identify the problem cells in the EXCEL table and obtain the training sample labels of the mold quality model.

[0099] In one embodiment of the present invention, the mold quality model includes a first hidden layer, a second hidden layer, a third hidden layer and an output layer;

[0100] The calculation formula for the first hidden layer is as follows:

[0101] ;

[0102] in, represents the first vector of the output of the first hidden layer, Represents the vector in the cell at row g and column h in the EXCEL table. , and denote the first learning parameter, the second learning parameter and the third learning parameter respectively, represents element-wise multiplication, represents a multi-layer perceptron;

[0103] The calculation formula for the second hidden layer is as follows:

[0104] ;

[0105] in, The second vector representing the output of the second hidden layer, represents the sliding window between rows, represents the index of the sliding window between rows, represents a sliding window between columns, Represents the index of the sliding window between columns;

[0106] The calculation formula for the third hidden layer is as follows:

[0107] ;

[0108] in, A third vector representing the output of the third hidden layer;

[0109] The calculation formula of the output layer is as follows:

[0110] ;

[0111] in, represents the output of the output layer, Indicates the cosine similarity. represents the standard threshold, Represents the similarity threshold.

[0112] Specifically, through the mapping of multiple hidden layers and output layers, the input EXCEL table is output as a binary matrix of the same size, where the value of the predicted problem cell is 1, otherwise it is 0.

[0113] In one embodiment of the present invention, the loss function of the mold quality model is The details are as follows:

[0114] The mold quality model is used to predict the problem cell of the EXCEL table output of the M angles of the target mold, and the predicted cell is used as a feature point. The feature point is extended according to the angle of the corresponding EXCEL table to obtain Focus, count the number of feature points connected to each focus;

[0115] Loss Function The calculation formula is as follows:

[0116] ;

[0117] in, and denote the first loss coefficient and the second loss coefficient respectively, , express The index of Indicates the number of rows of cells in the EXCEL table. express The index of Indicates the number of columns of cells in the EXCEL table. express The index of The first part of the EXCEL table showing the output of the mold quality model Line Column elements, Indicates the characteristics of the EXCEL table Line Column elements, Indicates the number of focal points, express The index of Indicates The number of feature points connected by each focus.

[0118] Specifically, the loss function consists of two parts, one is the loss function of the prediction problem cell and sample label of the prediction model, and the other is obtained based on physical properties. Since there are cracks inside the target mold, the corresponding cracks will be scanned for each angle (each angle scans the entire surface of the corresponding angle). If the problem cell is regarded as a feature point and extended to obtain a feature line segment, then for the crack, the feature line segment will converge as a focus at the crack, and the number of feature points connected to each focus is M (the same as the number of angles). The loss function of the second part is constructed based on the number of feature points connected to each focus.

[0119] In one embodiment of the present invention, the preset EXCEL command includes

[0120] Based on EXCEL's VBA editor, create macro commands. The macro commands are as follows:

[0121] Declare variables: declare a variable ws of Worksheet type;

[0122] Declare a variable cell of the Range type, which is used to traverse the cells of the EXCEL table in the loop;

[0123] Set the EXCEL table: Set the ws variable to the EXCEL table in "Sheet1" of the current workbook;

[0124] Traverse cells: Use ForEach loop to traverse all cells in the used range of the EXCEL table of ws;

[0125] Conditional judgment and formatting: Inside the loop, conditional judgment is performed on each cell: if the value of the cell is equal to 1, the background color of the cell is set to red (RGB value is 255, 0, 0);

[0126] End loop and macro: When the loop ends, the execution of the macro also ends;

[0127] The predicted problem cells output by the mold quality model are stored in a database, and the predicted problem cells are extracted from the database into EXCEL through SQL query, and macro commands are executed to obtain visual marks of the predicted problem cells in the EXCEL table;

[0128] The predicted problem cells of the mold quality model are stored in a database, and the predicted problem cells are extracted from the database into EXCEL through SQL query, and macro commands are executed to obtain visual marks of the predicted problem cells in the EXCEL table.

[0129] Specifically, by using the VBA editor that comes with EXCEL, after entering the macro command, combined with the automatic capture of the mold quality model output by the database query, the predicted problem cells in M ​​EXCEL tables are automatically marked in red. This method can achieve the following beneficial effects: Through the management tool developed by EXCEL, managers can open the corresponding EXCEL data package of the target mold through the system, and then intuitively view the problem area of ​​the target mold at each angle, so that managers can effectively view and control the mold quality.

[0130] The data processing system based on EXCEL logic is applied to the method described, including:

[0131] The data processing module obtains the point cloud data of M angles of the target mold and allocates M EXCEL tables, where one cell of each EXCEL table corresponds to one point cloud data of one angle;

[0132] The data analysis module analyzes M EXCEL tables based on the preset mold quality model to obtain the problem cells;

[0133] The table marking module visually marks problem cells based on preset EXCEL commands.

[0134] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A data processing method based on EXCEL logic, characterized in that: include: Step 1: Obtain the point cloud data of M angles of the target mold and allocate M EXCEL tables. One cell of each EXCEL table corresponds to a number of point cloud data of one angle, including: Establish a target plane, and obtain the central point cloud of point cloud data at an angle based on the target plane, including: Step 11, obtain the maximum contact surface of the point cloud data at the i-th angle as the target plane; the target plane calculation formula is as follows: ;in, Represents the x-axis component, y-axis component, and z-axis component of the normal vector of the target plane. , represents the intercept of the target plane, represents the contact distance threshold of the target plane, Represents the total number of point clouds at the i-th angle, express The index of Indicates the point cloud data of the i-th angle The x-axis coordinate, y-axis coordinate and z-axis coordinate of the point cloud, Express satisfaction If yes, it is 1, otherwise it is 0; Select the point cloud that does not touch the edge of the target plane , , and vertically project it into the target plane to obtain the mapping point cloud; the projection formula is as follows: ； in, express The x-axis coordinate, y-axis coordinate and z-axis coordinate of the mapped point cloud, express The x-axis coordinate, y-axis coordinate and z-axis coordinate of Connect the edges of the combined point cloud constructed by the mapped point cloud and the non-mapped point cloud on the target plane to obtain a geometric figure, obtain the center of the geometric figure by the centroid method, and use the point cloud with the smallest Euclidean distance to the center of the geometric figure as the central point cloud of the point cloud data at i angles; Based on the central point cloud, the point cloud data of one angle is constructed into a topological structure, clustered based on the topological structure to obtain an updated topological structure, and the point cloud data is filled into the corresponding cells of the EXCEL table according to the updated topological structure, including: Step 12, fill the center point cloud into the center table of the i-th EXCEL table; Taking the central point cloud as the root node, establish the topological structure of the point cloud data at the i-th angle , ,in, represents a 1-degree node, express Degree node, represents a node of degree n, ; Step 13, Degree Node Assignment Cluster centers are found, and clustering is performed according to the position coordinates of the point cloud to obtain the updated topological structure of the point cloud data at the i-th angle; Step 14, filling cells around the central point cloud with nodes of degree 1 to degree n of the updated topological structure in order; Step 2: Analyze M EXCEL tables based on the preset mold quality model to obtain problematic cells; Step 3: Based on the preset EXCEL command, visually mark the problem cells.

2. The data processing method based on EXCEL logic according to claim 1, characterized in that: Obtain the point cloud data of M angles of the target mold, including: By scanning multiple angles of the target mold with radar, point cloud data of each angle of the target mold is obtained; wherein each angle covers the entire surface of the corresponding angle of the target mold.

3. The data processing method based on EXCEL logic according to claim 2 is characterized in that: Mold quality model, including: Within a preset time period, point cloud data of the target mold at M angles are obtained, and a three-dimensional model of the target mold is constructed based on the point cloud data. If there is a problem area inside the three-dimensional model, feature images of the three-dimensional model at M angles are obtained, and the feature images include the problem area; Based on the point cloud data of the target mold at M angles, M EXCEL tables of the target mold are obtained through steps 4 to 7, and the feature images of the M angles of the 3D model are fitted with the M EXCEL tables. The problem cells in each EXCEL are determined according to the problem area, and M feature EXCEL tables are obtained; The EXCEL table is used as a training sample, and the corresponding feature EXCEL table is used as a sample label to train a mold quality model. The output of the mold quality model represents the predicted problem cell in the EXCEL table; The hyperparameters of the hidden layer of the mold quality model are back-propagated and updated through the preset loss function.

4. The data processing method based on EXCEL logic according to claim 3 is characterized in that: The characteristic images of M angles of the 3D model are fitted with M EXCEL tables, including: By scaling the feature image at the i-th angle, a scaled image is obtained; wherein the number of rows and columns of the scaled image are equal to the number of rows and columns of the EXCEL table at the i-th angle; Overlap the zoomed image of the i-th angle with the EXCEL table to obtain cells that overlap with the problem area in the zoomed image, and set the corresponding cells as problem cells.

5. The data processing method based on EXCEL logic according to claim 4 is characterized in that: The mold quality model includes a first hidden layer, a second hidden layer, a third hidden layer and an output layer; The calculation formula for the first hidden layer is as follows: ； in, represents the first vector of the output of the first hidden layer, Represents the vector in the cell at row g and column h in the EXCEL table. , and denote the first learning parameter, the second learning parameter and the third learning parameter respectively, represents element-wise multiplication, represents a multi-layer perceptron; The calculation formula for the second hidden layer is as follows: ； in, The second vector representing the output of the second hidden layer, represents the sliding window between rows, represents the index of the sliding window between rows, represents a sliding window between columns, Represents the index of the sliding window between columns; The calculation formula for the third hidden layer is as follows: ； in, A third vector representing the output of the third hidden layer; The calculation formula of the output layer is as follows: ； in, represents the output of the output layer, Indicates the cosine similarity. represents the standard threshold, Represents the similarity threshold.

6. The data processing method based on EXCEL logic according to claim 5, characterized in that: Loss function of mold quality model The details are as follows: The mold quality model is used to predict the problem cell of the EXCEL table output of the M angles of the target mold, and the predicted cell is used as a feature point. The feature point is extended according to the angle of the corresponding EXCEL table to obtain Focus, count the number of feature points connected to each focus; Loss Function The calculation formula is as follows: ； in, and denote the first loss coefficient and the second loss coefficient respectively, , express The index of Indicates the number of rows of cells in the EXCEL table. express The index of Indicates the number of columns of cells in the EXCEL table. express The index of The first part of the EXCEL table showing the output of the mold quality model Line Column elements, Indicates the characteristics of the EXCEL table Line Column elements, Indicates the number of focal points, express The index of Indicates The number of feature points connected by each focus.

7. The data processing method based on EXCEL logic according to claim 1, characterized in that: Preset EXCEL commands include: Based on EXCEL's VBA editor, create macro commands. The macro commands are as follows: Declare variables: declare a variable ws of Worksheet type; Declare a variable cell of the Range type, which is used to traverse the cells of the EXCEL table in the loop; Set the EXCEL table: Set the ws variable to the EXCEL table in "Sheet1" of the current workbook; Traverse cells: Use ForEach loop to traverse all cells in the used range of the EXCEL table of ws; Conditional judgment and formatting: Inside the loop, conditional judgment is performed on each cell: if the value of the cell is equal to 1, the background color of the cell is set to red (RGB value is 255, 0, 0); End loop: After the loop ends, the execution of the macro command also ends; The predicted problem cells output by the mold quality model are stored in a database, and the predicted problem cells are extracted from the database into EXCEL through SQL query, and macro commands are executed to obtain visual marks of the predicted problem cells in the EXCEL table.

8. A data processing system based on EXCEL logic, applied to the method described in any one of claims 1 to 7, characterized in that: include: The data processing module obtains the point cloud data of M angles of the target mold and allocates M EXCEL tables, where one cell of each EXCEL table corresponds to one point cloud data of one angle; The data analysis module analyzes M EXCEL tables based on the preset mold quality model to obtain the problem cells; The table marking module visually marks problem cells based on preset EXCEL commands.

Images