Size analysis method and system based on engine mold data acquisition

Abstract

The application discloses a size analysis method and system based on engine mold collected data, and relates to the field of data analysis. The method comprises the following steps: obtaining a design model and three-dimensional point cloud data of an engine mold, and dividing the design model into multiple functional areas based on design functions; registering the point cloud data with the design model, and calculating an initial deviation value of each data point; for each functional area, calculating an area deformation aggregation value; for each two functional areas, calculating a deformation propagation coefficient; for each functional area, calculating a systematic influence deviation according to the area deformation aggregation values and the deformation propagation coefficients of other areas; for each data point, calculating a correlation correction deviation according to the initial deviation value and the systematic influence deviation of the belonging area; and finally outputting a diagnosis result based on the correlation correction deviation and the systematic influence deviation. The application can effectively distinguish local abnormalities from overall deformation, accurately locate the wear source, and improve the detection accuracy and maintenance efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, and particularly relates to a size analysis method based on engine mold collected data. BACKGROUND

[0002] As the core process equipment in the automobile industry, the manufacturing precision and durability of the engine mold directly determine the size quality and service performance of key components such as engine cylinder body and cylinder cover. During the manufacturing and in-service use of the mold, due to mechanical stress, thermal fatigue and wear and other factors, the key working surfaces such as the cavity and core of the mold will inevitably change in size and degrade in appearance.

[0003] Although the existing method can intuitively display the out-of-tolerance position, it regards the mold as a geometric set without internal association, fails to deeply explore the internal relationship and synergistic effect of deformation between different functional areas, only focuses on the geometric deviation of discrete points, and completely ignores the thermal coupling and stress transmission effect between different functional areas such as the mold closing surface, cooling water channel and slider mechanism of the engine mold under actual high temperature and high pressure cyclic working conditions. SUMMARY

[0004] To solve the defect problem of isolated analysis and inability to identify the correlation and synergistic deformation between areas in the existing engine mold size analysis method, the present application provides a size analysis method based on engine mold collected data.

[0005] In a first aspect, the present application provides a size analysis method based on engine mold collected data, characterized in that it comprises the following steps:

[0006] Obtaining a design model and three-dimensional point cloud data of an engine mold, and dividing a plurality of functional areas in the design model based on design functions;

[0007] Registering the three-dimensional point cloud data with the design model, and calculating the initial deviation value of each data point;

[0008] For each functional area, calculating the regional deformation aggregation value of the functional area according to the initial deviation value of all data points in the area and the geometric saliency weight of each data point, wherein the geometric saliency weight is determined based on the curvature of the position of the data point;

[0009] For each two functional areas, calculating the deformation propagation coefficient from one functional area to another functional area according to the geodesic distance and structural coupling factor between them;

[0010] For each functional area, calculating the systematic influence deviation of the functional area according to the regional deformation aggregation value of other functional areas and the deformation propagation coefficient from other functional areas to the functional area;

[0011] For each data point, a relevance correction bias of the data point is calculated according to the initial bias value of the data point and the systematic influence bias of the functional area to which the data point belongs;

[0012] Based on the relevance correction bias, the size analysis of the engine mold collected data is completed.

[0013] Preferably, the functional areas include a moving mold cavity area, a fixed mold cavity area, a slider area, an insert area, and a clamping surface area.

[0014] Preferably, the specific step of calculating the area deformation aggregation value includes multiplying the initial bias value of each data point in the functional area by the geometric significance weight of the data point, then summing, and then dividing by the sum of the geometric significance weights of all data points.

[0015] Preferably, the calculation of the deformation propagation coefficient is proportional to a structural coupling factor between two functional areas and inversely proportional to the square of the geodesic distance between the two functional areas.

[0016] Preferably, the structural coupling factor is determined in the following manner: if the two functional areas share a boundary, the structural coupling factor is the length of the shared boundary; if the two functional areas are not directly adjacent, the structural coupling factor is a constant set based on the overall structural size of the mold.

[0017] Preferably, the specific step of calculating the systematic influence bias includes multiplying the area deformation aggregation value of each other functional area by the deformation propagation coefficient from the other functional area to the current functional area, and then summing the product results of all other functional areas.

[0018] Preferably, the specific step of calculating the relevance correction bias includes subtracting the systematic influence bias of the functional area to which the data point belongs from the initial bias value of the data point.

[0019] Preferably, the completion of the size analysis of the engine mold collected data includes establishing a rule base based on the relevance correction bias, comparing the relevance correction bias with a preset threshold value, determining "local severe wear" or "systematic deformation" according to the rule base, and outputting the corresponding diagnostic results and maintenance recommendations to complete the size analysis of the engine mold collected data.

[0020] In a second aspect, the present application provides a size analysis system based on engine mold collected data, which adopts the following technical solution:

[0021] The size analysis system based on engine mold collected data includes a processor and a memory.

[0022] The present application has the following beneficial effects:

[0023] By introducing the geometric saliency weight, the different importance of high-curvature key features (such as fillets and edges) and flat areas in part forming and stress concentration is distinguished when calculating the overall deformation of the region, so that the deformation evaluation is not accurate, the region deformation aggregation value can more sensitively capture and reflect the deformation trend on the key geometric features, and the engineering relevance and accuracy of the deformation quantification are improved, and important signals are avoided from being overwhelmed by a large amount of non-key area noise.

[0024] By constructing the deformation propagation coefficient, the influence strength of the deformation of one region on another region is quantified, so that the propagation of the deformation in the mold structure network can be simulated, which is beneficial to the deformation diagnosis of the size abnormality of the engine mold.

[0025] By calculating the correlation correction deviation, the problem that the local true wear signal is overwhelmed when there is obvious overall deformation (such as thermal expansion), resulting in either missing the true defect or generating a large number of false positives, is solved, the signal-to-noise ratio and positioning accuracy of the detection of real damage such as local erosion and cracks are enhanced, and the reliability and maintenance efficiency of the detection are improved. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a method flowchart of steps S1-S6 in the size analysis method based on engine mold collected data according to the embodiment of the application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments of the application.

[0028] Referring to Figure 1 The size analysis method based on engine mold collected data includes steps S1-S6, which are specifically as follows:

[0029] S1: Obtain the design model and three-dimensional point cloud data of the engine mold, and divide a plurality of functional areas in the design model based on the design function; register the three-dimensional point cloud data with the design model, and calculate the initial deviation value of each data point.

[0030] The three-dimensional point cloud data of the engine mold can be scanned by a three-dimensional point cloud data acquisition device. Since there is a preset design model during engine design, the three-dimensional point cloud data scanned and the design model are in different coordinate systems, which will cause the deviation calculation reference to be inconsistent, so a high-precision spatial alignment of the three-dimensional point cloud data and the design model is needed to be realized by an optimization algorithm, and the three-dimensional point cloud data and the design model are converted to the same coordinate system.

[0031] Specifically, the point cloud data is pre-processed by downsampling and denoising, and then coarse registration is realized by principal component analysis method, and finally the rigid transformation matrix of the three-dimensional point cloud data is iteratively calculated, the objective function is the sum of the nearest point distance of the three-dimensional point cloud data and the surface of the design model, and the rigid transformation matrix of the three-dimensional point cloud data is obtained by iteratively minimizing the objective function by using the optimization group intelligent algorithm, and the three-dimensional point cloud data and the design model are converted to the same coordinate system by using the rigid transformation matrix of the three-dimensional point cloud data.

[0032] Among them, the point cloud data is pre-processed by downsampling and denoising, and the principal component analysis method is a known technical means, and the optimization group intelligent algorithm can be selected as a genetic algorithm, which is a known technical means.

[0033] For each data point in the three-dimensional point cloud data , the shortest Euclidean distance from the data point to the triangular mesh surface of the design model is calculated as the initial deviation value , and the distance symbol is determined by the direction relationship of the normal vector of the design model at the projection point of the point , wherein the positive direction of the normal vector is the positive deviation, and vice versa, the shortest Euclidean distance from the point to the surface is a known technical means, and will not be described here.

[0034] S2: For each functional area, the regional deformation aggregation value of the functional area is calculated according to the initial deviation value of all data points in the area and the geometric saliency weight of each data point.

[0035] In the design model, according to the design drawing and process knowledge, the surface of the design model is divided into movable mold cavity area, fixed mold cavity area, slider area, insert block area and mold closing surface area by personnel with relevant work experience, each area is assigned a unique identifier, and geometric boundary information is stored. Among them, the mapping relationship table of point cloud data points and functional areas is established by nearest neighbor search to ensure that each point belongs to a unique area, and a point-area mapping relationship table is obtained.

[0036] After obtaining the three-dimensional point cloud data and the design model in the same coordinate system, the size analysis of the engine mold can be completed by the deviation value of the three-dimensional point cloud data and the design model, but due to the large noise and strong randomness of the deviation value of a single data point, it cannot represent the overall deformation state of the functional area; the traditional arithmetic average method cannot distinguish the difference between high-curvature key features (such as fillets and edges) and flat areas in stress concentration and forming importance, resulting in insufficient deformation evaluation sensitivity and important signals being diluted by a large number of non-key area data.

[0037] In order to comprehensively reflect the deviation trend of all points in the area and highlight the contribution of key geometric features, the regional deformation aggregation value is constructed.

[0038] Since curvature is a core parameter characterizing the geometric significance of a surface, high curvature regions are usually associated with stress concentration and material flow critical points, and small deformations may have an amplified impact on the quality of parts.

[0039] By incorporating geometric significance weights, the weighted average method can effectively reduce random error interference, enabling differential geometric analysis of the design model surface and calculation of each data point. average curvature and Gaussian curvature The calculation processes for mean curvature and Gaussian curvature are well-known techniques and will not be elaborated further.

[0040] Furthermore, outliers (such as scanning noise) can distort the weighted results. To reduce the impact of outliers on error interference, the median absolute deviation method is used to identify outliers. For the deviation value... Beyond the regional median Data points with a multiple of absolute deviation are temporarily set as the first in the 3D point cloud data. Weight values ​​of each data point This eliminates its interference with aggregation calculations.

[0041] For non-anomaly point calculation in 3D point cloud data, the first... Weight values ​​of each data point The formula can be:

[0042]

[0043] in, and The weighting coefficient is adjustable, and an empirical value is set according to the mold type. , It can be adjusted by the implementer according to the specific implementation scenario.

[0044] For each data point The average curvature.

[0045] For each data point Gaussian curvature.

[0046] For functional areas Through formula Calculate the weighted average deformation as the functional area. Comprehensive deformation index The summation operation is completed by traversing all data points within the region.

[0047] S3: For every two functional areas, calculate the deformation propagation coefficient from one functional area to another based on the geodesic distance between them and the structural coupling factor.

[0048] In actual working conditions, stress and heat are transferred between functional areas through structural paths in engine molds, resulting in deformation correlation effects. However, traditional methods cannot quantify the intensity of the impact of deformation in one area on another. Traditional methods rely solely on Euclidean straight-line distances and ignore structural topological relationships, leading to distorted assessment of propagation intensity and an inability to explain systematic deformation patterns.

[0049] The intensity of deformation propagation is positively correlated with the tightness of the structural connection between regions and negatively correlated with the effective distance. Therefore, the geodesic distance can accurately reflect the actual path length of stress propagation along the surface of the mold entity, which is more in line with the structural topology requirements than the Euclidean distance.

[0050] The geodesic distance calculation process is as follows: On the triangular mesh surface of the design model, the Dijkstra algorithm is used to calculate the functional area. and Shortest path length between geometric center points This distance characterizes the actual propagation path of deformation in the structural network.

[0051] After obtaining the geodesic distance, in order to determine the functional area and The coupling strength can be determined by judging the functional area. and Whether the boundary is shared or not: if shared, the coupling factor is the total length of the shared boundary; if not shared, due to the conduction effect of the engine mold substrate, the coupling factor is a constant. , which represents the coupling strength through the matrix material.

[0052] Among them, functional areas can be used when determining shared boundaries. and The minimum Euclidean distance between boundary data points is used to determine shared boundaries. and If the minimum Euclidean distance between boundary data points is 1, it indicates that the boundary data point is a shared boundary data point, and the total number of shared boundary data points is equal to the total length of the shared boundary. The minimum Euclidean distance is 1 because each data point has its own region, and data points between two regions do not overlap. Therefore, if they are shared boundary data points, the minimum Euclidean distance is 1. and Data points belonging to the boundary can be obtained through the boundary in the design model.

[0053] Calculation from functional area arrive Deformation propagation coefficient:

[0054]

[0055] wherein S is a coupling factor, is a functional area with the shortest path length between the geometric center points, is a material factor, taking an empirical value , is a very small constant to prevent the denominator from being zero, and The value size can be adjusted by the implementer according to the specific implementation scene.

[0056] is a functional area with the shortest path length distance between the geometric center points. The smaller the value is, the closer the distance is, and the more obvious the deformation propagation from the functional area to is, the larger the value is.

[0057] is a coupling factor between the area and . The larger the value is, the higher the coupling strength is, and the more obvious the deformation propagation from the functional area to is, the larger the value is.

[0058] The larger the value is, the stronger the influence relationship of the deformation propagation from the functional area to is.

[0059] S4: For each functional area, calculate the systematic influence deviation of the functional area according to the area deformation aggregation value of other functional areas and the deformation propagation coefficient from other functional areas to the functional area.

[0060] The observed area deviation is a mixture of local factors and systematic influences, and it is impossible to distinguish the deformation source, that is, whether the deviation of a region is caused by its own wear or is related to the deformation of other regions. The lack of such distinction leads to blind maintenance decisions, which may misjudge the systematic deformation as a local defect, waste resources and miss optimization opportunities.

[0061] The systematic influence deviation is the deformation component transmitted to the target area by the external area through physical association, which conforms to linear superposition, and the contribution of each source area is determined by its own comprehensive deformation index ( ) and deformation propagation coefficient ( ).

[0062] Calculate the contribution of the source: for the target functional area and each other functional area ( ), calculate the contribution value , Quantify the individual impact of .

[0063] The superimposed calculation result of the systematic influence deviation of the target functional area is:

[0064]

[0065] Calculate the systematic influence deviation of , sum all areas except , and ensure linear synthesis of each source contribution.

[0066] In the calculation of the systematic influence deviation of all areas, the case can be automatically skipped by comparing the area identifiers , ensuring that only external influences are represented.

[0067] S5: For each data point, calculate the correlation correction deviation of the data point according to the initial deviation value of the data point and the systematic influence deviation of the functional area to which the data point belongs.

[0068] For each data point, calculate the correlation correction deviation of the first data point in the non-anomalous three-dimensional point cloud data according to the initial deviation value of the data point and the systematic influence deviation of the functional area to which the data point belongs.

[0069]

[0070] wherein is the initial deviation value, which includes local wear and systematic variation components. In strong systematic background (such as thermal expansion), the true local damage (such as cracks, erosion) signal is submerged, leading to missed detection or false positives. Traditional threshold methods cannot separate mixed signals, reducing the reliability and positioning accuracy of detection.

[0071] is the systematic influence deviation, and the correlation correction is achieved by subtracting the systematic influence estimate from the original deviation to obtain the correlation correction deviation , which realizes signal separation, and the residual error after correction mainly reflects the deviation caused by local factors.

[0072] is the correction intensity coefficient, which is an empirical value​​ .

[0073] Based on the pre-established point-area mapping relationship table, for each data point , the function area to which it belongs is determined through spatial position query , which provides the basis for point-level correction.

[0074] S6: Correct the deviation based on correlation, complete the size analysis of engine mold collected data.

[0075] Based on the correlation correction deviation, a rule base is established, when and , it is determined that "local serious wear and tear", and "priority repair" is recommended; when and involves multiple areas, it is determined that "systematic deformation", and "process parameter adjustment" is recommended; when , the systematic influence deviation threshold , the systematic influence deviation threshold , the rule base can be adjusted by the implementer according to the specific implementation scene.

[0076] The embodiment of the application also discloses a size analysis system based on engine mold collected data, comprising a processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, the size analysis method based on engine mold collected data according to the application is realized.

[0077] The above system also includes a communication bus and a communication interface and other components familiar to those skilled in the art, the setting and function of which are known in the art, therefore, will not be repeated here.

[0078] In the present application, the aforementioned memory can be any tangible medium containing or storing a program, which can be used or combined with an instruction execution system, device or instrument. For example, the computer readable storage medium can be any appropriate magnetic storage medium or magneto-optical storage medium, such as resistive memory, dynamic random access memory, static random access memory, etc., or any other medium that can be used to store the required information and can be accessed by an application, module or both. Any such computer storage medium can be part of the device or accessible or connectable to the device.

[0079] It should be pointed out that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the protection scope of the present application patent should be subject to the appended claims.

Claims

1. A dimensional analysis method based on engine mold acquisition data, characterized in that, Includes the following steps: Obtain the design model and 3D point cloud data of the engine mold, and divide the design model into multiple functional areas based on the design functions; The 3D point cloud data is registered with the design model, and the initial deviation value of each data point is calculated. For each functional region, the regional deformation aggregate value of the functional region is calculated based on the initial deviation value of all data points in the region and the geometric salience weight of each data point. The geometric salience weight is determined based on the curvature of the location of the data point. When calculating the overall deformation of the region through the geometric salience weight, high curvature key features are distinguished from flat regions. For every two functional zones, the deformation propagation coefficient from one functional zone to another is calculated based on the geodesic distance between them and the structural coupling factor. The structural coupling factor is determined as follows: if two functional areas share a boundary, the structural coupling factor is the total length of the shared boundary; if the two functional areas are not directly adjacent, the structural coupling factor is a constant set based on the overall structural dimensions of the mold. For each functional region, the systematic influence deviation of that functional region is calculated based on the regional deformation aggregation value of other functional regions and the deformation propagation coefficient from other functional regions to that functional region; the deformation trend on key geometric features is captured and reflected by calculating the regional deformation aggregation value. The influence of deformation in one region on another region is quantified by constructing a deformation propagation coefficient. For each data point, the correlation correction deviation is calculated based on the initial deviation value of the data point and the systematic influence deviation of the functional area to which the data point belongs; Based on the correlation correction deviation, the dimensional analysis of the engine mold acquisition data is completed.

2. The dimensional analysis method based on engine mold acquisition data according to claim 1, characterized in that, The functional areas include the moving model cavity area, the fixed model cavity area, the slider area, the insert area, and the mold closing surface area.

3. The dimensional analysis method based on engine mold acquisition data according to claim 1, characterized in that, The specific steps for calculating the regional deformation aggregation value include: multiplying the initial deviation value of each data point within the functional area by the geometric significance weight of that data point, then summing the results, and finally dividing by the sum of the geometric significance weights of all data points.

4. The dimensional analysis method based on engine mold acquisition data according to claim 1, characterized in that, The deformation propagation coefficient is calculated to be directly proportional to the structural coupling factor between the two functional regions and inversely proportional to the square of the geodesic distance between the two functional regions.

5. The dimensional analysis method based on engine mold acquisition data according to claim 1, characterized in that, The specific steps for calculating the systematic impact deviation are as follows: multiply the aggregated value of the regional deformation of each other functional region by the deformation propagation coefficient from other functional regions to the current functional region, and then sum the product results of all other functional regions.

6. The dimensional analysis method based on engine mold acquisition data according to claim 1, characterized in that, The specific steps for calculating the correlation correction deviation are as follows: subtract the systematic influence deviation of the functional region to which the data point belongs from the initial deviation value of the data point.

7. The dimensional analysis method based on engine mold acquisition data according to claim 1, characterized in that, The dimensional analysis of the engine mold acquisition data includes: establishing a rule base based on correlation correction deviation, comparing the correlation correction deviation with a preset threshold, determining "local severe wear" or "systematic deformation" according to the rule base, and outputting the corresponding diagnostic results and maintenance suggestions to complete the dimensional analysis of the engine mold acquisition data.

8. A dimensional analysis system based on engine mold acquisition data, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the dimensional analysis method based on engine mold acquisition data according to any one of claims 1-7.

Images