Method and system for modeling preparation scenes of electrochromic glass based on digitalization

Abstract

The present invention relates to the field of digital modeling technology, and discloses a method for modeling a preparation scene of electrochromic glass based on digitalization, including: collecting a micro-nano structure image data set of the electrochromic glass to be modeled, performing spatial structure matching on the images in the micro-nano structure image data set, and obtaining structure matching points; constructing a spatial point cloud of the electrochromic glass according to the structure matching points, and constructing a physical simulation model of the electrochromic glass; performing physical coupling setting on the physical simulation model, constructing a performance mapping model based on real operating parameters and historical preparation performance data, and analyzing the target working parameters of the electrochromic glass; using the target working parameters, performing data twinning on the target preparation model to analyze the preparation accuracy of the electrochromic glass, and when the preparation accuracy meets the set accuracy, using the twin model as the target preparation model. The present invention can improve the authenticity of the preparation scene modeling of electrochromic glass based on digitalization.

Description

Technical Field

[0001] The invention relates to a method and system for modeling a preparation scene of electrochromic glass based on digitalization, and belongs to the technical field of digital modeling. Background Art

[0002] As a new type of intelligent material, electrochromic glass has broad application prospects in many fields such as architecture, automobiles, aerospace, etc. It can change its own optical properties through the action of an external electric field, adjust the light transmittance, etc., and provide people with many conveniences, such as adjusting indoor lighting and reducing automobile energy consumption. Data modeling plays an indispensable role in the preparation process of electrochromic glass based on digitalization. Building a data modeling platform for the preparation of electrochromic glass based on digitalization can help scientific researchers and production technicians accurately record, track and manage various parameter information in the preparation process, such as key data such as raw material ratio, reaction temperature, and electric field strength, thereby effectively improving the preparation quality and production efficiency of electrochromic glass. At the same time, relevant personnel can use the platform to monitor the changes of various parameters in the preparation process in real time, conduct in-depth data analysis, and optimize and adjust the preparation process in a timely manner to produce electrochromic glass products with better performance and more stable quality. It also facilitates information sharing and communication between different R&D groups, between production departments and R&D departments, and between relevant personnel in the upstream and downstream industrial chains, thereby reducing the risk of product quality problems caused by poor information flow or inaccurate parameter control, and improving the collaborative work efficiency of the entire industrial chain.

[0003] At present, for the data modeling platform in the preparation scenario of electrochromic glass based on digitalization, conventional modeling software is generally used to build a system with a basic operation interface and simple interactive functions to achieve the corresponding data modeling work. However, in this way, the data entry and subsequent management of many key parameters often rely on manual operations by users. Since the preparation process of electrochromic glass involves a large number of complex physical and chemical parameters, there are both structured data such as the content of raw materials and unstructured data such as descriptions of phenomena during the experiment. It is difficult for users to manage data accurately when facing a large amount of data, which leads to the data modeling in the preparation scenario of electrochromic glass based on digitalization being not realistic enough. Summary of the invention

[0004] In order to solve the above problems, the present invention provides a method for modeling a preparation scene of electrochromic glass based on digitalization, which can improve the authenticity of modeling a preparation scene of electrochromic glass based on digitalization.

[0005] To achieve the above-mentioned purpose, the present invention provides a method for modeling a preparation scene of electrochromic glass based on digitalization, comprising:

[0006] Collecting a micro-nano structure image dataset of the electrochromic glass to be modeled, constructing a feature description region of the micro-nano structure image dataset, calculating a feature descriptor of the feature point based on the feature description region, and performing spatial structure matching on images in the micro-nano structure image dataset using the feature descriptor to obtain a structure matching point;

[0007] According to the structural matching points, a spatial point cloud of the electrochromic glass is constructed, a spatial contour of the electrochromic glass is analyzed by using the spatial point cloud, a surface fitting is performed on the electrochromic glass based on the spatial contour, and a physical simulation model of the electrochromic glass is constructed based on the result of the surface fitting;

[0008] Performing physical coupling setting on the physical simulation model to obtain a target preparation model, querying real operating parameters corresponding to the target preparation model and historical preparation performance data of the electrochromic glass, building a performance mapping model of the electrochromic glass based on the real operating parameters and the historical preparation performance data, querying target performance requirements of the electrochromic glass, and analyzing target operating parameters of the electrochromic glass based on the performance mapping model and the target performance requirements;

[0009] The target working parameters are used to set parameters of the target preparation model to obtain a deployment model, and data twinning is performed on the deployment model to obtain a twin model. The twin model is used to perform production simulation on the electrochromic glass, and real-time monitoring data of the production simulation is queried to analyze the preparation accuracy of the electrochromic glass. When the preparation accuracy meets the set accuracy, the twin model is used as the target preparation model of the electrochromic glass.

[0010] Optionally, constructing a feature description area of ​​the micro-nanostructure image data set includes:

[0011] Performing multi-scale Gaussian blur processing on the original image in the micro-nanostructure image data set to obtain a Gaussian blurred image;

[0012] Performing adjacent scale subtraction processing on the Gaussian blurred image to obtain a Gaussian difference image;

[0013] After defining the feature points in the Gaussian difference image, identifying the scale information of the feature points;

[0014] Calculating the Haar wavelet response of the directional interval within the feature point area;

[0015] Determining the characteristic direction of the characteristic point according to the Haar wavelet response;

[0016] A feature description region of the micro-nanostructure image data set is constructed according to the scale information and the feature direction.

[0017] Optionally, using the feature descriptor to perform spatial structure matching on the images in the micro-nano structure image dataset to obtain structure matching points includes:

[0018] Querying the feature descriptor corresponding to the original image in the micro-nanostructure image dataset;

[0019] The similarity between the corresponding descriptors of the original images in the micro-nanostructure image dataset is calculated using the following formula:

[0020] ;

[0021] in represents the similarity, M represents the number of original images, represents the i-th image in the original image, represents the jth image in the original image, Represents the feature descriptor vector of the i-th image feature point p in the original image, Represents the feature descriptor vector of the i-th image feature point Q in the original image, Represents the value of the feature descriptor vector of feature point p in the corresponding dimension w, express Represents the value of the feature descriptor vector of the feature point Q in the corresponding dimension w,

[0022] Based on the similarity, the feature descriptors are preliminarily matched and screened to obtain preliminarily matched feature descriptors;

[0023] Using a preconfigured geometric model to remove mismatches from the preliminary matching feature descriptor to obtain a target matching feature descriptor;

[0024] The target matching feature descriptor is used as a structure matching point.

[0025] Optionally, analyzing the spatial profile of the electrochromic glass by using the spatial point cloud includes:

[0026] Performing filtering on the spatial point cloud to obtain a filtered point cloud;

[0027] Performing edge detection processing on the filtered point cloud to obtain an edge point cloud;

[0028] Using a preconfigured rotation matrix, rotating the edge point cloud to obtain a rotated point cloud;

[0029] Using the rotating point cloud as the spatial contour points of the electrochromic glass;

[0030] Smoothing the spatial contour points to obtain smoothed contour points;

[0031] After the smooth contour points are linked, the spatial contour of the electrochromic glass is determined.

[0032] Optionally, performing surface fitting on the electrochromic glass based on the spatial profile includes:

[0033] an array of point coordinates identifying the contour of the space;

[0034] Querying the fitting expectation of the electrochromic glass curved surface fitting to configure the fitting radial basis function of the electrochromic glass;

[0035] constructing an interpolation matrix of the electrochromic glass based on the fitted radial basis function and the point coordinate array;

[0036] Based on the interpolation matrix, construct a fitting linear equation group of the electrochromic glass;

[0037] Solving the fitting linear equation to obtain fitting surface parameter solution values ​​and surface adjustment solution values;

[0038] Based on the fitting surface parameter solution value and the surface adjustment solution value, the electrochromic glass is subjected to surface fitting.

[0039] Optionally, constructing a fitting linear equation group of the electrochromic glass based on the interpolation matrix includes:

[0040] Based on the interpolation matrix, the fitted linear equation group of the electrochromic glass is constructed using the following formula:

[0041] ;

[0042] in, represents the fitted linear equation, represents the elements in the interpolation matrix, represents the matrix elements of the first row and column, represents the matrix element in the first row and the Nth column, represents the matrix element in the first column of the Nth row, represents the matrix element of the Nth row and the Nth column, b represents the surface adjustment value, represents the fitting surface parameter value of the first row, represents the fitting surface parameter value of the Nth row, represents the first element in the function value array, Represents the Nth element in the function value array.

[0043] Optionally, performing surface fitting on the electrochromic glass based on the fitting surface parameter solution value and the surface adjustment solution value includes:

[0044] Based on the fitting surface parameter solution value and the surface adjustment solution value, the fitting surface function value of the electrochromic glass is calculated using the following formula:

[0045] ;

[0046] in, represents the value of the fitting surface function, Represents the number of spatial contour points of the electrochromic glass, represents the i-th solution value in the solution value of the fitting surface parameter, represents the fitted radial basis function, represents the coordinates of the spatial contour point i of the electrochromic glass, represents the coordinates of the spatial contour point j of the electrochromic glass, represents the Euclidean distance between points i and j, represents the surface adjustment solution value,

[0047] According to the fitting surface function value, a fitting surface of the electrochromic glass is constructed.

[0048] Optionally, performing physical coupling setting on the physical simulation model to obtain a target preparation model includes:

[0049] Analyze the physical fields and coupling relationships of the physical simulation model;

[0050] Based on the physical field and the coupling relationship, configuring a coupling simulation tool of the physical simulation model;

[0051] Query the actual electrical properties of the electrochromic glass corresponding to the physical simulation model;

[0052] According to the actual electrical characteristics, configuring physical field parameters for the coupling simulation tool to obtain a target coupling model;

[0053] Constructing a coupling interface of the target coupling model, and after defining an interaction mechanism of the coupling interface of the target coupling model, performing a model test on the target coupling model with the defined interaction mechanism;

[0054] When the result of the model test is excellent, the target coupling model with the interaction mechanism defined is used as the target preparation model.

[0055] Optionally, performing data twinning on the deployment model to obtain a twin model includes:

[0056] configuring an initial twin model of the deployment model;

[0057] Querying model data of the deployment model;

[0058] Using the model data to perform data twinning on the initial twin model to obtain twin data;

[0059] The multiple similarity loss of the twin data is calculated using the following formula:

[0060] ;

[0061] in, represents multiple similarity loss, H represents the maximum spatial distance between model data and twin data in the corresponding vector space, , , is a hyperparameter, Represents the first similarity between model data and twin data, Represents the second similarity between model data and twin data, A represents model data, B represents twin data, Represents the same part between the model data and the twin data;

[0062] When the multiple similarity loss is not greater than the preset loss, the initial twin model is supplemented with the twin data to obtain a twin model.

[0063] In a second aspect, the present invention provides a system for modeling a production scene of electrochromic glass based on digitalization, the system comprising:

[0064] A structure point matching module is used to collect a micro-nano structure image dataset of the electrochromic glass to be modeled, construct a feature description region of the micro-nano structure image dataset, calculate a feature descriptor of the feature point based on the feature description region, and use the feature descriptor to perform spatial structure matching on the image in the micro-nano structure image dataset to obtain a structure matching point;

[0065] A physical simulation module, used to construct a spatial point cloud of the electrochromic glass according to the structural matching points, analyze the spatial contour of the electrochromic glass using the spatial point cloud, perform surface fitting on the electrochromic glass based on the spatial contour, and construct a physical simulation model of the electrochromic glass based on the result of the surface fitting;

[0066] a working parameter analysis module, used to perform physical coupling setting on the physical simulation model to obtain a target preparation model, query the real operating parameters corresponding to the target preparation model and the historical preparation performance data of the electrochromic glass, construct a performance mapping model of the electrochromic glass based on the real operating parameters and the historical preparation performance data, query the target performance requirements of the electrochromic glass, and analyze the target working parameters of the electrochromic glass based on the performance mapping model and the target performance requirements;

[0067] A production simulation module is used to use the target working parameters to set parameters of the target preparation model to obtain a deployment model, perform data twinning on the deployment model to obtain a twin model, use the twin model to perform production simulation of the electrochromic glass, query the real-time monitoring data of the production simulation to analyze the preparation accuracy of the electrochromic glass, and when the preparation accuracy meets the set accuracy, use the twin model as the target preparation model of the electrochromic glass.

[0068] Compared with the prior art, the embodiment of the present invention has built a relatively systematic and comprehensive modeling, analysis and production control process around electrochromic glass. First, the image data set of the micro-nano structure of the electrochromic glass is collected, and the feature description area is constructed and the feature descriptor is calculated through multiple steps. Based on this, the spatial structure matching is performed to obtain the structure matching points. This process uses microscopic imaging and fine image processing algorithms to lay the foundation for the subsequent three-dimensional model construction, and can accurately restore the shape of its micro-nano structure in space; then, a spatial point cloud is constructed based on the structure matching points, and then the spatial contour is analyzed and the surface is fitted, and finally a physical simulation model is constructed, so that the micro-nano structure of the electrochromic glass can be digitally presented, which is convenient for in-depth exploration of its various physical properties; then, the physical simulation model is physically coupled, and the physical fields and coupling relationships are comprehensively considered. With the help of professional software and other configuration parameters, interfaces and interaction mechanisms are defined, and the target preparation model is obtained through testing; then, a performance mapping model is constructed based on the real operating parameters and historical preparation performance data related to the target preparation model, and the target working parameters are analyzed in combination with the target performance requirements to provide a basis for production control. Finally, the model is deployed using the target working parameters. A twin model is generated through data twins for production simulation. The simulation data is monitored in real time to analyze the preparation accuracy. When the accuracy meets the set requirements, it is determined as the target preparation model, thereby ensuring that the electrochromic glass produced meets high quality standards and achieving a stable and precise production process, which overall improves the authenticity of the electrochromic glass from modeling to production. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 A schematic diagram of a process for realizing a scene modeling method for preparing electrochromic glass based on digitalization provided in one embodiment of the present invention;

[0070] Figure 2 A schematic diagram of a module of a system for digitally realizing a scene modeling system for preparing electrochromic glass according to an embodiment of the present invention.

[0071] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0072] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0073] The embodiment of the present application provides a method for modeling a preparation scenario of electrochromic glass based on digitalization. The execution subject of the method for modeling a preparation scenario of electrochromic glass based on digitalization includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the method for modeling a preparation scenario of electrochromic glass based on digitalization can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0074] Embodiment 1:

[0075] Reference Figure 1 FIG. 1 is a flow chart of a method for modeling a preparation scene of electrochromic glass based on digitalization according to an embodiment of the present invention. In this embodiment, the method for modeling a preparation scene of electrochromic glass based on digitalization includes:

[0076] S1. Collect a micro-nano structure image dataset of the electrochromic glass to be modeled, construct a feature description region of the micro-nano structure image dataset, calculate a feature descriptor of the feature point based on the feature description region, and use the feature descriptor to perform spatial structure matching on the image in the micro-nano structure image dataset to obtain a structure matching point.

[0077] The embodiment of the present invention can provide users with an accurate and comprehensive image dataset by collecting the micro-nano structure image dataset of the electrochromic glass to be modeled in the scene, thereby ensuring that the three-dimensional model finally constructed has high accuracy and authenticity. The electrochromic glass refers to a kind of smart glass, which is often composed of multiple layers of materials, generally including a glass substrate, a transparent conductive layer (such as ITO, etc.), an electrochromic layer (such as tungsten oxide, etc.), an electrolyte layer, and an electrode layer, etc. The micro-nano structure image dataset refers to a collection of a series of images obtained by imaging the microstructure of an electrochromic glass sample using a high-resolution electron microscope (such as a transmission electron microscope TEM, a scanning electron microscope SEM).

[0078] The embodiment of the present invention can ensure that the descriptor can effectively characterize the feature information of the local area around the feature point by constructing the feature description area of ​​the micro-nanostructure image data set, and make it have certain rotation, scale and other invariances, which facilitates subsequent feature matching operations between images of different viewing angles.

[0079] The feature description area refers to a specific area in the image constructed with the feature point as the center, which is used to calculate the feature descriptor of the feature point to characterize the feature information of the local image around the feature point.

[0080] As an embodiment of the present invention, the construction of the feature description area of ​​the micro-nanostructure image data set includes: performing multi-scale Gaussian blur processing on the original image in the micro-nanostructure image data set to obtain a Gaussian blurred image, performing adjacent scale subtraction processing on the Gaussian blurred image to obtain a Gaussian difference image, defining feature points in the Gaussian difference image, identifying scale information of the feature points, calculating the Haar wavelet response of the directional interval within the feature point field, determining the feature direction of the feature point according to the Haar wavelet response, and constructing the feature description area of ​​the micro-nanostructure image data set according to the scale information and the feature direction.

[0081] The Haar wavelet response refers to the result obtained by performing correlation operations between the Haar wavelet template and the image pixel area in image analysis, which reflects the grayscale change characteristics of the image in different directions in the local area. The scale information refers to the scale position correlation of the feature point in the multi-scale space, which corresponds to the different scale parameters used when constructing a multi-scale representation of the image (such as a Gaussian pyramid), and reflects the degree of blur or resolution at which the feature point is detected.

[0082] Optionally, the Gaussian blurred image can be obtained by performing a convolution operation on the original image using Gaussian kernels with different standard deviations (such as standard deviations increasing in sequence), and performing a weighted average on each pixel of the original image according to the weight of the Gaussian function. The Gaussian difference image can be obtained by performing a subtraction operation on the corresponding pixel values ​​of adjacent scales (i.e., images processed by adjacent Gaussian kernels with different standard deviations) in the constructed Gaussian blurred image, and forming a new image after pixel-by-pixel subtraction, that is, the Gaussian difference image. The feature points in the Gaussian difference image can be defined by finding local extreme points in the Gaussian difference image (in three-dimensional space, including the two-dimensional image position The scale information of the feature point can be determined based on the Gaussian difference image layer (corresponding to different scales) where the feature point is located and the scale parameter of the Gaussian blur corresponding to the layer (such as the standard deviation of the Gaussian kernel used). The Haar wavelet response can be obtained by taking a certain neighborhood range (such as a circular area) with the feature point as the center, dividing the direction interval (for example, dividing multiple angle intervals) in the area, and calculating the Haar Wavelet response, that is, the corresponding response value is obtained by accumulating the wavelet template and the pixel value in the horizontal and vertical directions. The characteristic direction can be obtained by statistically summing the amplitudes of the Haar wavelet responses in each direction interval, and the direction with the largest amplitude is determined as the characteristic direction of the feature point. The feature description area of ​​the micro-nanostructure image data set constructed according to the scale information and the characteristic direction can be achieved by taking the feature point as the center and determining the appropriate size according to its scale information (such as the side length is related to the scale factor, which can be 20 times the scale factor, or it can be set in combination with the actual application), and constructing a square area along the characteristic direction as the feature description area operation.

[0083] The embodiment of the present invention can make the image have richer information expression ability by calculating the feature descriptor of the feature point based on the feature description area, and can effectively match under different viewing angles. For example, in a multi-view image, by comparing the distance between feature descriptors (such as Euclidean distance), it is possible to find feature point pairs representing the same physical location, thereby establishing a connection between images of different viewing angles, providing key data for subsequent spatial structure matching, and then realizing the conversion from a two-dimensional image to a three-dimensional model.

[0084] The feature descriptor is a vector generated for each feature point through a specific algorithm (such as the SURF algorithm) and is used to characterize feature information such as local image texture and structure around the feature point.

[0085] Optionally, the feature descriptor for calculating the feature point based on the feature description area can be achieved by dividing the feature description area (such as a 16×16 pixel rectangular area) into 4×4 sub-areas, and then calculating the gradient amplitude and direction of the sub-areas (determined by horizontal and vertical grayscale differences), and then counting the accumulated values ​​of the gradient amplitude of pixels falling into each directional range in the sub-area (such as divided into 8 intervals) to form an 8-dimensional vector, and finally concatenating the 8-dimensional vectors corresponding to the 4×4 sub-areas in sequence to obtain a 128-dimensional feature descriptor (vector).

[0086] The embodiment of the present invention uses the feature descriptor to perform spatial structure matching on the images in the micro-nano structure image data set, and obtains structure matching points. Corresponding feature point pairs can be found between images of different viewing angles according to the feature descriptors, and the positional relationship of these feature points in space is determined, thereby achieving spatial structure matching between images and obtaining structure matching points, so as to restore the true form of the electrochromic glass micro-nano structure in three-dimensional space and provide key position information for building an accurate three-dimensional model.

[0087] The spatial structure refers to the shape and layout of the electrochromic glass micro-nanostructure in three-dimensional space, including the distribution position of the electrochromic material particles in the glass matrix, the arrangement of the crystal structure in space, and the spatial shape formed by the distribution of different elements.

[0088] As an embodiment of the present invention, the use of the feature descriptor to perform spatial structure matching on the images in the micro-nano structure image dataset to obtain structure matching points includes: querying the feature descriptor corresponding to the original image in the micro-nano structure image dataset, and calculating the similarity between the descriptors corresponding to the original images in the micro-nano structure image dataset using the following formula:

[0089] ;

[0090] in, represents the similarity, M represents the number of original images, represents the i-th image in the original image, represents the jth image in the original image, Represents the feature descriptor vector of the i-th image feature point p in the original image, Represents the feature descriptor vector of the i-th image feature point Q in the original image, Represents the value of the feature descriptor vector of feature point p in the corresponding dimension w, Represents the value of the feature descriptor vector of the feature point Q in the corresponding dimension w,

[0091] Based on the similarity, the feature descriptors are preliminarily matched and screened to obtain preliminary matching feature descriptors, and mismatches are removed from the preliminary matching feature descriptors using a preconfigured geometric model to obtain target matching feature descriptors, which are used as structural matching points.

[0092] Optionally, the feature descriptors are preliminarily matched and screened based on the similarity, and a suitable distance threshold is set based on experience or pre-experimental conditions of a specific micro-nanostructure image data set to obtain the preliminary matching feature descriptors. When the calculated similarity between the two feature descriptors is less than the threshold, it is preliminarily considered that the two feature points may be matching point pairs, that is, corresponding points in the spatial structure. For example, if the threshold is set to 0.5, when the Euclidean distance calculation result of the two feature descriptors is 0.3, the two feature points are recorded as candidate matching points to obtain the preliminary matching feature descriptors, and the use of the pre-configured geometric model to remove mismatches of the preliminary matching feature descriptors to obtain the target matching feature descriptors can be realized by the homography matrix model. It should be noted that since the real spatial structure has certain geometric laws, such as in a plane or a plane-like structure, the lines between the matching points should satisfy certain parallel, collinear and other geometric relationships, and there are similar geometric consistency requirements in three-dimensional space. Therefore, it is necessary to calculate the geometric relationship between the matching points, and eliminate those candidate matching points that do not meet the expected geometric constraints, so as to further improve the accuracy of the matching. For example, if, according to the calculated geometric model, the lines connecting some matching point pairs obviously violate geometric laws such as parallel relationships in space, these matching points are considered to be mismatched points and are removed.

[0093] S2. Constructing a spatial point cloud of the electrochromic glass according to the structural matching points, analyzing the spatial contour of the electrochromic glass by using the spatial point cloud, performing surface fitting on the electrochromic glass based on the spatial contour, and constructing a physical simulation model of the electrochromic glass based on the result of the surface fitting.

[0094] The embodiment of the present invention constructs the spatial point cloud of the electrochromic glass according to the structural matching points, which can intuitively show the approximate distribution of the micro-nano structure of the electrochromic glass in space, and provides original and detailed material for further exploring its spatial characteristics.

[0095] The spatial point cloud refers to a collection of a large number of discrete points in three-dimensional space.

[0096] Optionally, the spatial point cloud can determine the three-dimensional coordinates of each point based on the principle of multi-view geometry (such as triangulation), using structural matching points and camera internal and external parameters. Then, these three-dimensional coordinates are organized and stored in a suitable data structure (such as an array or a specific point cloud format) to form a spatial point cloud.

[0097] The embodiment of the present invention utilizes the spatial point cloud to analyze the spatial contour of the electrochromic glass, and can grasp the shape characteristics of the electrochromic glass micro-nano structure as a whole, such as whether it is a regular geometric shape or an irregular complex shape, and then understand the relative position relationship between the various parts.

[0098] As an embodiment of the present invention, the use of the spatial point cloud to analyze the spatial contour of the electrochromic glass includes: filtering the spatial point cloud to obtain a filtered point cloud, performing edge detection processing on the filtered point cloud to obtain an edge point cloud, rotating the edge point cloud using a preconfigured rotation matrix to obtain a rotated point cloud, using the rotated point cloud as the spatial contour points of the electrochromic glass, smoothing the spatial contour points to obtain smoothed contour points, and determining the spatial contour of the electrochromic glass after contour point linking the smoothed contour points.

[0099] Optionally, the filtered point cloud can be obtained by using statistical filtering technology, setting a distance threshold (the specific setting needs to be combined with actual application), counting the average distance between each point and its neighborhood points, and eliminating noise points with obviously larger distances. The edge detection processing of the filtered point cloud to obtain the edge point cloud can be achieved through the Canny algorithm operation. The smoothing processing of the spatial contour points to obtain the smoothed contour points can use a moving average smoothing method, taking the average coordinate value of each point and several points in its neighborhood to replace the current point coordinates, and achieving smoothing through multiple iterations.

[0100] It should be explained that the preconfigured rotation matrix can be constructed using the Rodriguez rotation formula, and the intersection vectors of adjacent point clouds in the edge point cloud can be used as rotation axis related vectors, the intersection angles of adjacent point clouds can be used as rotation angles, and the rotation axis related vectors and the rotation angles can be brought into the Rodriguez rotation formula as rotation parameters to construct the preconfigured rotation matrix.

[0101] The embodiment of the present invention performs surface fitting on the electrochromic glass based on the spatial contour, so that the originally discrete and irregular point cloud data has a relatively concise and continuous mathematical expression, which is convenient for subsequent operations such as various physical performance analysis, simulation and model construction.

[0102] As an embodiment of the present invention, the surface fitting of the electrochromic glass based on the spatial contour includes: identifying the point coordinate array of the spatial contour, querying the fitting expectation of the electrochromic glass surface fitting to configure the fitting radial basis function of the electrochromic glass, constructing the interpolation matrix of the electrochromic glass based on the fitting radial basis function and the point coordinate array, constructing the fitting linear equation group of the electrochromic glass based on the interpolation matrix, solving the fitting linear equations to obtain fitting surface parameter solution values ​​and surface adjustment solution values, and performing surface fitting on the electrochromic glass based on the fitting surface parameter solution values ​​and the surface adjustment solution values.

[0103] Among them, the point coordinate array refers to the coordinate information of each point in the point cloud data corresponding to the spatial contour of the electrochromic glass in the three-dimensional space, and the fitting expectation refers to the effect and characteristics that the final fitted surface should achieve after comprehensively considering the many characteristics of the electrochromic glass itself and the purpose of surface fitting. For example, from the perspective of the shape of the spatial contour of the electrochromic glass, whether it is a regular shape or a complex irregular shape, this affects the selection of the fitting function; from the perspective of the fitting accuracy requirement, it is required to closely fit the contour points with high precision (such as for precision optical analysis scenes), or to allow a certain error in the rough fitting (such as only for appearance display), the fitting radial basis function refers to a class of functions with radial distance as a variable (such as common Gaussian function, thin plate spline function, etc.), in the process of surface fitting, the spatial contour point is taken as the center, and the function is used to measure the degree of correlation between points based on the distance between each point, and then a mathematical tool that can fit the surface of the spatial contour point cloud data is constructed in combination.

[0104] Optionally, the point coordinate array for identifying the spatial contour can be obtained by obtaining processed electrochromic glass spatial contour point cloud data and arranging it into a two-dimensional array. The fitting expectation can be obtained by querying the user's requirements for the shape of the electrochromic glass spatial contour, fitting accuracy, surface smoothness and continuity, and other factors. The fitting radial basis function can select a suitable type from common radial basis functions (such as Gaussian, thin plate spline, etc.) according to the fitting expectation, and set relevant parameters (such as the σ value of the Gaussian function) as needed. The interpolation matrix can be obtained by creating an N×N two-dimensional array, traversing the spatial contour point pairs, and substituting the distance between the points calculated by the selected radial basis function into the calculated matrix element values. The solving of the fitting linear equation to obtain the fitting surface parameter solution and the surface adjustment solution can be achieved by using Gaussian elimination or conjugate gradient and other solution methods, and programming to implement the corresponding algorithm logic to calculate the fitting surface parameters and the surface adjustment solution.

[0105] Exemplarily, constructing the fitted linear equation group of the electrochromic glass based on the interpolation matrix includes: constructing the fitted linear equation group of the electrochromic glass based on the interpolation matrix using the following formula:

[0106] ;

[0107] in, represents the fitted linear equation, represents the elements in the interpolation matrix, represents the matrix elements of the first row and column, represents the matrix element in the first row and the Nth column, represents the matrix element in the first column of the Nth row, represents the matrix element of the Nth row and the Nth column, b represents the surface adjustment value, represents the fitting surface parameter value of the first row, represents the fitting surface parameter value of the Nth row, represents the first element in the function value array, Represents the Nth element in the function value array.

[0108] As an optional embodiment of the present invention, the performing surface fitting on the electrochromic glass based on the fitting surface parameter solution value and the surface adjustment solution value includes: calculating the fitting surface function value of the electrochromic glass based on the fitting surface parameter solution value and the surface adjustment solution value using the following formula:

[0109] ;

[0110] in, represents the value of the fitting surface function, Represents the number of spatial contour points of the electrochromic glass, represents the i-th solution value in the solution value of the fitting surface parameter, represents the fitted radial basis function, represents the coordinates of the spatial contour point i of the electrochromic glass, represents the coordinates of the spatial contour point j of the electrochromic glass, represents the Euclidean distance between points i and j, represents the surface adjustment solution value,

[0111] According to the fitting surface function value, a fitting surface of the electrochromic glass is constructed.

[0112] Optionally, constructing the fitting surface of the electrochromic glass according to the fitting surface function value can be achieved by constructing a fitting surface function using a fitting surface function and a radial basis function, then calculating the spatial area sampling function value of the electrochromic glass, and finally displaying the fitting curve using visualization software.

[0113] The embodiment of the present invention can comprehensively and accurately digitally present the micro-nano structure of the electrochromic glass by constructing the physical simulation model of the electrochromic glass based on the result of the surface fitting, so as to facilitate in-depth research on various properties of the electrochromic glass (such as optical, electrical, thermal properties, etc.).

[0114] Optionally, the physical simulation model can be constructed by using CAD software combined with the results of surface fitting.

[0115] S3. Perform physical coupling setting on the physical simulation model to obtain a target preparation model, query the real operating parameters corresponding to the target preparation model and the historical preparation performance data of the electrochromic glass, construct a performance mapping model of the electrochromic glass based on the real operating parameters and the historical preparation performance data, query the target performance requirements of the electrochromic glass, and analyze the target operating parameters of the electrochromic glass based on the performance mapping model and the target performance requirements.

[0116] The embodiment of the present invention obtains the target preparation model by performing physical coupling setting on the physical simulation model. The coupling relationship between the physical fields can be reasonably set, such as determining how the electric field change affects heat conduction, how thermal stress changes optical properties and other interrelated mechanisms, and the physical fields that were originally analyzed independently can be integrated to form a comprehensive model that conforms to actual working conditions, so as to better analyze the preparation status of the electrochromic glass.

[0117] As an embodiment of the present invention, physical coupling setting is performed on the physical simulation model to obtain a target preparation model, including: analyzing the physical field and coupling relationship of the physical simulation model, configuring a coupling simulation tool of the physical simulation model based on the physical field and the coupling relationship, querying the actual electrical characteristics of the electrochromic glass corresponding to the physical simulation model, configuring physical field parameters for the coupling simulation tool according to the actual electrical characteristics, obtaining a target coupling model, constructing a coupling interface of the target coupling model, and after defining an interaction mechanism of the coupling interface of the target coupling model, performing a model test on the target coupling model with the interaction mechanism defined, and when the result of the model test is excellent, using the target coupling model with the interaction mechanism defined as the target preparation model.

[0118] Among them, the physical field refers to a region in space that has specific physical properties and can exert corresponding physical effects on objects in it, such as electrical fields, mechanical fields, and optical fields. The coupling relationship refers to the correlation mechanism of mutual influence and interaction between different physical fields, such as electrical coupling, electrothermal coupling, and electro-optical coupling.

[0119] Optionally, the analysis of the physical fields and coupling relationships of the physical simulation model can carefully study the working principles, actual usage scenarios and characteristics of existing physical simulation models of electrochromic glass, and sort out the physical fields including electricity, heat, mechanics, optics, etc. from the aspects of the electrical process involved in its color change, the thermal performance of heat generation, the mechanical conditions that may be subjected to external forces and optical related properties, and then determine the coupling relationship between the physical fields based on the physical mechanism of mutual influence in practice (such as the Joule heating effect of electric heat, the thermal expansion and contraction correlation between heat and mechanics, etc.). For example, these relationships can be clarified by referring to relevant physical subject materials and past experimental research results on electrochromic glass; the coupling simulation tool for configuring the physical simulation model based on the physical field and the coupling relationship can select tools such as COMSOL Multiphysics, ANSYS, etc. according to the determined physical field and coupling relationship. This type of professional software that supports multi-physics field coupling simulation, the actual electrical characteristics can be obtained by querying the historical experimental database of the electrochromic glass, and the physical field parameters are configured for the coupling simulation tool based on the actual electrical characteristics to obtain the target coupling model. These actual electrical characteristic parameters can be accurately filled into the corresponding positions according to the function entry of the corresponding parameter settings in the selected coupling simulation tool (such as the electrical module interface of ANSYS software), and the parameter configuration is ensured to comply with the physical principles according to the basic electrical physical laws (such as Ohm's law, Gauss's law, etc.), thereby completing the electrical physical field parameter configuration. The coupling interface can use the interface definition function provided by the selected coupling simulation tool (for example, in COMSOL In the software, through specific coupling interface setting options, the corresponding coupling interface type is selected for the coupling relationship between the determined physical fields (such as different types of preset interfaces for electrothermal coupling, electromechanical coupling, etc.). The interaction mechanism can be based on the physical laws of interaction between physical fields (such as the empirical formula for the influence of thermal on electrical parameters, the geometric optical principle of mechanical deformation on the change of optical path, etc.), and the interaction mechanism at each coupling interface is accurately set in the coupling simulation tool (such as the calculation method of variable transfer, the equation of feedback effect, etc.). It should be noted that when the result of the model test is excellent, it means that the model test is judged to be excellent according to the preset evaluation criteria (such as error range, degree of fit with experimental data, etc., which need to be set in combination with the needs of actual applications).

[0120] The embodiment of the present invention can fully understand the actual situation of each link from preparation to actual operation of the electrochromic glass by querying the actual operating parameters corresponding to the target preparation model and the historical preparation performance data of the electrochromic glass, provide a reliable data basis for the subsequent construction of a performance mapping model, and help to explore the key factors affecting product performance and the intrinsic connections between the factors.

[0121] Optionally, the actual operating parameters and the historical preparation performance data can be obtained by querying the enterprise database and management system of the electrochromic glass.

[0122] The embodiment of the present invention can closely link the many factors affecting the performance of electrochromic glass with the final performance by constructing the performance mapping model of the electrochromic glass based on the actual operating parameters and the historical preparation performance data, thereby realizing the conversion prediction from measurable and controllable input parameters to key performance indicators.

[0123] Optionally, the performance mapping model of the electrochromic glass based on the actual operating parameters and the historical preparation performance data can be constructed by collecting and organizing the actual operating parameters and historical preparation performance data of the electrochromic glass, cleaning and preprocessing them, selecting appropriate features, and using traditional or machine learning methods to build a model, train, optimize and evaluate the verification model; finally, the model is applied and updated as needed to build a model that reflects its performance mapping relationship and assists in product management.

[0124] The embodiment of the present invention can help users measure whether the product is qualified and whether it meets the important reference standards of the market and actual use needs by querying the target performance requirements of the electrochromic glass, so as to ensure that the electrochromic glass finally produced can accurately adapt to the corresponding application scenarios.

[0125] Optionally, the target performance requirements may be learned by querying the corresponding manufacturer or supplier.

[0126] The embodiment of the present invention analyzes the target operating parameters of the electrochromic glass based on the performance mapping model and the target performance requirements. By determining the target operating parameters, the user can adjust the preparation process and optimize the equipment control parameters in a targeted manner to ensure that the produced electrochromic glass can stably achieve the target performance requirements. At the same time, during use, reasonable operation and maintenance can be performed based on these parameters to maximize the functional advantages of the product and improve the overall quality and performance stability of the product.

[0127] Optionally, the process of analyzing the target operating parameters of the electrochromic glass based on the performance mapping model and the target performance requirements is as follows: first, carefully combing through the constructed performance mapping model, clarifying the association mechanism and mathematical logic between the input variables (such as the actual operating parameters of the electrochromic glass related to the electrical, thermal, optical, etc.) and the output variables (such as the various performance indicators such as the color change response time and the transmittance change) in the model, and clarifying how the model converts the input into the output; secondly, for the determined target performance requirements of the electrochromic glass, refining each abstract performance indicator into specific numerical values ​​or value ranges that can be quantified, organizing these clarified target performance indicators into a form that is easy to substitute into the model, so that they meet the format requirements of subsequent calculation and analysis; then, implementing reverse deduction calculations based on the specific type of the performance mapping model to obtain the target operating parameters. If the model can obtain an analytical solution based on simple physical formulas, it can be directly derived through mathematical operations; for complex neural networks, nonlinear regression and other models that are difficult to obtain analytical solutions, numerical calculation methods, such as iterative search algorithms, are used to continuously iterate and update based on the initial guess value; when multiple parameters are involved and there are complex constraints, optimization algorithms such as genetic algorithms and particle swarm optimization algorithms are used to search for the optimal parameter combination so that the target operating parameters can meet the target performance requirements. Finally, the target operating parameters are substituted into the performance mapping model for positive verification to check whether the output performance indicators meet the target requirements. At the same time, the rationality of the parameters in the actual physical scene is considered. If not, the reasons are analyzed and adjusted. If there are multiple sets of parameters that meet the requirements, compare their performance, cost, stability and other aspects, and select the best one, so as to provide a scientific and reliable basis for the parameter setting of electrochromic glass in the actual production and use.

[0128] S4. Utilize the target working parameters to set parameters of the target preparation model to obtain a formulation model, perform data twinning on the formulation model to obtain a twin model, utilize the twin model to perform production simulation on the electrochromic glass, query the real-time monitoring data of the production simulation to analyze the preparation accuracy of the electrochromic glass, and when the preparation accuracy meets the set accuracy, use the twin model as the target preparation model of the electrochromic glass.

[0129] The embodiment of the present invention uses the target working parameters to set the parameters of the target preparation model to obtain a deployment model, which can ensure that the preparation model is closely related to the expected product performance, so that subsequent simulation, production and other links have a basic model that meets actual needs.

[0130] Optionally, the preparation model can be obtained by using a parameter setting interface in the target preparation model according to the target working parameter settings.

[0131] The embodiment of the present invention performs data twinning on the deployment model to obtain a twin model, which can provide a more realistic, comprehensive and dynamic virtual environment for subsequent production simulation, and can test and analyze various possible situations in this virtual environment in advance without interfering with the actual production process or incurring cost consumption.

[0132] As an embodiment of the present invention, performing data twinning on the deployment model to obtain a twin model includes: configuring an initial twin model of the deployment model, querying model data of the deployment model, performing data twinning on the initial twin model using the model data to obtain twin data, and calculating multiple similarity losses of the twin data using the following formula:

[0133] ;

[0134] in, represents multiple similarity loss, H represents the maximum spatial distance between model data and twin data in the corresponding vector space, , , is a hyperparameter, Represents the first similarity between model data and twin data, Represents the second similarity between model data and twin data, A represents model data, B represents twin data, Represents the same part between the model data and the twin data;

[0135] When the multiple similarity loss is not greater than the preset loss, the initial twin model is supplemented with the twin data to obtain a twin model.

[0136] The preset loss is 0.2, which needs to be set in combination with the actual application.

[0137] Optionally, the initial twin model can be configured through a COMSOL Multiphysics tool.

[0138] The embodiment of the present invention can simulate the entire production process of the electrochromic glass from raw material input, execution of various process links to final finished product output by using the twin model to perform production simulation on the electrochromic glass, thereby discovering possible problems in the production process in advance, such as product defects caused by unreasonable process parameters, production delays caused by poor connection between different links, and the impact of physical field changes that do not meet expectations on product performance.

[0139] Optionally, the production simulation of the electrochromic glass using the twin model can be started by simulating the raw material input, processing technology execution and other links in the virtual environment constructed by the digital twin in the sequence of the actual production process.

[0140] The real-time monitoring data of the production simulation is queried to analyze the preparation accuracy of the electrochromic glass. The verification of this link ensures that the target preparation model finally adopted is strictly tested and verified, and its corresponding production process can stably and accurately manufacture electrochromic glass that meets the target performance requirements and quality standards.

[0141] Optionally, the preparation accuracy of the electrochromic glass can be analyzed by comparing the acquired real-time monitoring data with the pre-set accuracy standards of the electrochromic glass in terms of size, performance indicators (such as transmittance, color change response time, etc.). It can be determined whether the set accuracy requirements are met by calculating the error value, deviation range, and other statistical analysis methods.

[0142] It should be explained that when the preparation accuracy meets the set accuracy, it means that the optical performance parameters, electrical performance parameters, mechanical performance parameters and dimensional accuracy parameters of the electrochromic glass after preparation meet the production requirements. For example, the transmittance in the optical performance parameters is 30% - 70% which is excellent, the conductivity in the electrical performance parameters is 10⁻³ - 10³ S / cm which is excellent, the Mohs hardness in the mechanical performance parameters is 5 - 7 which is excellent, and in the dimensional accuracy parameters, the length and width are within ±0.5 mm and the thickness is within ±0.1 mm which is excellent. The specific settings need to be combined with user needs and actual application scenarios.

[0143] Embodiment 2:

[0144] like Figure 2 , which is a system functional module diagram of a method for implementing a digitally-based preparation scene modeling method for electrochromic glass provided by an embodiment of the present invention.

[0145] The system 200 for modeling the preparation scene of electrochromic glass based on digitalization of the present invention can be installed in an electronic device. According to the functions implemented, the system for modeling the preparation scene of electrochromic glass based on digitalization can include a structure point matching module 201, a physical simulation module 202, a working parameter analysis module 203 and a production simulation module 204. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0146] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0147] The structure point matching module 201 is used to collect a micro-nano structure image dataset of the electrochromic glass to be modeled, construct a feature description region of the micro-nano structure image dataset, calculate a feature descriptor of the feature point based on the feature description region, and use the feature descriptor to perform spatial structure matching on the image in the micro-nano structure image dataset to obtain a structure matching point;

[0148] The physical simulation module 202 is used to construct a spatial point cloud of the electrochromic glass according to the structural matching points, analyze the spatial contour of the electrochromic glass using the spatial point cloud, perform surface fitting on the electrochromic glass based on the spatial contour, and construct a physical simulation model of the electrochromic glass based on the result of the surface fitting;

[0149] The working parameter analysis module 203 is used to perform physical coupling setting on the physical simulation model to obtain a target preparation model, query the real operating parameters corresponding to the target preparation model and the historical preparation performance data of the electrochromic glass, build a performance mapping model of the electrochromic glass based on the real operating parameters and the historical preparation performance data, query the target performance requirements of the electrochromic glass, and analyze the target working parameters of the electrochromic glass based on the performance mapping model and the target performance requirements;

[0150] The production simulation module 204 is used to use the target working parameters to set parameters of the target preparation model to obtain a deployment model, perform data twinning on the deployment model to obtain a twin model, use the twin model to perform production simulation of the electrochromic glass, query the real-time monitoring data of the production simulation to analyze the preparation accuracy of the electrochromic glass, and when the preparation accuracy meets the set accuracy, use the twin model as the target preparation model of the electrochromic glass.

[0151] In detail, the modules in the digital electrochromic glass preparation scene modeling system 200 described in the embodiment of the present invention are used in the same manner as described above. Figure 1 The technical means described in the method for digitally realizing the preparation scene modeling of electrochromic glass are the same and can produce the same technical effects, so they will not be repeated here.

[0152] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0153] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0154] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0155] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0156] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for modeling a preparation scene of electrochromic glass based on digitalization, characterized in that: The method comprises: Collecting a micro-nano structure image dataset of the electrochromic glass to be modeled, constructing a feature description region of the micro-nano structure image dataset, calculating a feature descriptor of a feature point corresponding to the feature description region, and using the feature descriptor to perform spatial structure matching on images in the micro-nano structure image dataset to obtain a structure matching point; According to the structural matching points, a spatial point cloud of the electrochromic glass is constructed, the spatial contour of the electrochromic glass is analyzed by using the spatial point cloud, a surface fitting is performed on the electrochromic glass based on the spatial contour, and a physical simulation model of the electrochromic glass is constructed based on the result of the surface fitting, wherein the surface fitting of the electrochromic glass based on the spatial contour comprises: identifying an array of point coordinates of the spatial contour, querying a fitting expectation of the surface fitting of the electrochromic glass to configure a fitting radial basis function of the electrochromic glass, and performing surface fitting on the electrochromic glass based on the fitting radial basis function and the point coordinates. The interpolation matrix of the electrochromic glass is constructed by an array, and a fitting linear equation group of the electrochromic glass is constructed based on the interpolation matrix. The fitting linear equation is solved to obtain a fitting surface parameter solution value and a surface adjustment solution value. The electrochromic glass is subjected to surface fitting based on the fitting surface parameter solution value and the surface adjustment solution value. Further, the electrochromic glass is subjected to surface fitting based on the fitting surface parameter solution value and the surface adjustment solution value, including: calculating the fitting surface function value of the electrochromic glass based on the fitting surface parameter solution value and the surface adjustment solution value using the following formula: ； in, represents the value of the fitting surface function, Represents the number of spatial contour points of the electrochromic glass, represents the i-th solution value in the solution value of the fitting surface parameter, represents the fitted radial basis function, represents the coordinates of the spatial contour point i of the electrochromic glass, represents the coordinates of the spatial contour point j of the electrochromic glass, represents the Euclidean distance between points i and j, represents the surface adjustment solution value, Constructing a fitting surface of the electrochromic glass according to the fitting surface function value; Performing physical coupling setting on the physical simulation model to obtain a target preparation model, querying real operating parameters corresponding to the target preparation model and historical preparation performance data of the electrochromic glass, building a performance mapping model of the electrochromic glass based on the real operating parameters and the historical preparation performance data, querying target performance requirements of the electrochromic glass, and analyzing target operating parameters of the electrochromic glass based on the performance mapping model and the target performance requirements; The target working parameters are used to set parameters of the target preparation model to obtain a deployment model, and data twinning is performed on the deployment model to obtain a twin model. The twin model is used to perform production simulation on the electrochromic glass, and real-time monitoring data of the production simulation is queried to analyze the preparation accuracy of the electrochromic glass. When the preparation accuracy meets the set accuracy, the twin model is used as the target preparation model of the electrochromic glass.

2. The method for modeling a preparation scene of electrochromic glass based on digitalization according to claim 1, characterized in that: The step of constructing the feature description area of ​​the micro-nanostructure image data set includes: Performing multi-scale Gaussian blur processing on the original image in the micro-nanostructure image data set to obtain a Gaussian blurred image; Performing adjacent scale subtraction processing on the Gaussian blurred image to obtain a Gaussian difference image; After defining the feature points in the Gaussian difference image, identifying the scale information of the feature points; Calculating the Haar wavelet response of the directional interval within the feature point area; Determining a characteristic direction of the characteristic point according to the Haar wavelet response; A feature description region of the micro-nanostructure image data set is constructed according to the scale information and the feature direction.

3. The method for modeling a preparation scene of electrochromic glass based on digitalization according to claim 1, characterized in that: The using the feature descriptor to perform spatial structure matching on the images in the micro-nano structure image data set to obtain structure matching points includes: Querying the feature descriptor corresponding to the original image in the micro-nanostructure image dataset; The similarity between the corresponding descriptors of the original images in the micro-nanostructure image dataset is calculated using the following formula: ; in, represents the similarity, M represents the number of original images, represents the i-th image in the original image, represents the jth image in the original image, Represents the feature descriptor vector of the i-th image feature point p in the original image, Represents the feature descriptor vector of the j-th image feature point Q in the original image, Represents the value of the feature descriptor vector of feature point p in the corresponding dimension w, Represents the value of the feature descriptor vector of the feature point Q in the corresponding dimension w, Based on the similarity, the feature descriptors are preliminarily matched and screened to obtain preliminarily matched feature descriptors; Using a preconfigured geometric model to remove mismatches from the preliminary matching feature descriptor to obtain a target matching feature descriptor; The target matching feature descriptor is used as a structure matching point.

4. The method for modeling a preparation scene of electrochromic glass based on digitalization according to claim 1, characterized in that: The analyzing the spatial profile of the electrochromic glass by using the spatial point cloud includes: Performing filtering on the spatial point cloud to obtain a filtered point cloud; Performing edge detection processing on the filtered point cloud to obtain an edge point cloud; Using a preconfigured rotation matrix, rotating the edge point cloud to obtain a rotated point cloud; Using the rotating point cloud as the spatial contour points of the electrochromic glass; Smoothing the spatial contour points to obtain smoothed contour points; After the smooth contour points are linked, the spatial contour of the electrochromic glass is determined.

5. The method for modeling a manufacturing scene of electrochromic glass based on digitalization according to claim 1, characterized in that: The step of constructing a fitting linear equation group of the electrochromic glass based on the interpolation matrix comprises: Based on the interpolation matrix, the fitted linear equation group of the electrochromic glass is constructed using the following formula: ； in, represents the fitted linear equation, represents the elements in the interpolation matrix, represents the matrix element in the first row and first column, represents the matrix element in the first row and the Nth column, represents the matrix element in the first column of the Nth row, represents the matrix element of the Nth row and the Nth column, b represents the surface adjustment value, represents the fitting surface parameter value of the first row, represents the fitting surface parameter value of the Nth row, represents the first element in the function value array, Represents the Nth element in the function value array.

6. The method for modeling a manufacturing scene of electrochromic glass based on digitalization according to claim 1, characterized in that: Performing physical coupling setting on the physical simulation model to obtain a target preparation model includes: Analyzing the physical fields and coupling relationships of the physical simulation model; Based on the physical field and the coupling relationship, configuring a coupling simulation tool of the physical simulation model; Query the actual electrical properties of the electrochromic glass corresponding to the physical simulation model; According to the actual electrical characteristics, configuring physical field parameters for the coupling simulation tool to obtain a target coupling model; Constructing a coupling interface of the target coupling model, and after defining an interaction mechanism of the coupling interface of the target coupling model, performing a model test on the target coupling model with the defined interaction mechanism; When the result of the model test is excellent, the target coupling model with the interaction mechanism defined is used as the target preparation model.

7. The method for modeling a manufacturing scene of electrochromic glass based on digitalization according to claim 1, characterized in that: The step of performing data twinning on the deployment model to obtain a twin model includes: configuring an initial twin model of the deployment model; Querying model data of the deployment model; Using the model data to perform data twinning on the initial twin model to obtain twin data; The multiple similarity loss of the twin data is calculated using the following formula: ； in, represents multiple similarity loss, H represents the maximum spatial distance between model data and twin data in the corresponding vector space, , , is a hyperparameter, Represents the first similarity between model data and twin data, Represents the second similarity between model data and twin data, A represents model data, B represents twin data, Represents the same part between the model data and the twin data; When the multiple similarity loss is not greater than the preset loss, the initial twin model is supplemented with the twin data to obtain a twin model.

8. A scene modeling system for realizing the preparation of electrochromic glass based on digitalization, characterized in that: The system comprises: A structure point matching module is used to collect a micro-nano structure image dataset of the electrochromic glass to be modeled, construct a feature description region of the micro-nano structure image dataset, calculate a feature descriptor corresponding to a feature point of the feature description region, and use the feature descriptor to perform spatial structure matching on images in the micro-nano structure image dataset to obtain a structure matching point; The physical simulation module is used to construct a spatial point cloud of the electrochromic glass according to the structural matching points, analyze the spatial contour of the electrochromic glass by using the spatial point cloud, perform surface fitting on the electrochromic glass based on the spatial contour, and construct a physical simulation model of the electrochromic glass based on the result of the surface fitting, wherein the surface fitting on the electrochromic glass based on the spatial contour comprises: identifying a point coordinate array of the spatial contour, querying a fitting expectation of the electrochromic glass surface fitting to configure a fitting radial basis function of the electrochromic glass, and performing surface fitting on the electrochromic glass based on the fitting radial basis function and the surface fitting result. The interpolation matrix of the electrochromic glass is constructed based on the point coordinate array, and a fitting linear equation group of the electrochromic glass is constructed based on the interpolation matrix. The fitting linear equation is solved to obtain a fitting surface parameter solution value and a surface adjustment solution value, and the electrochromic glass is subjected to surface fitting based on the fitting surface parameter solution value and the surface adjustment solution value. Further, the surface fitting of the electrochromic glass based on the fitting surface parameter solution value and the surface adjustment solution value includes: based on the fitting surface parameter solution value and the surface adjustment solution value, using the following formula to calculate the fitting surface function value of the electrochromic glass: in, ; represents the value of the fitting surface function, Represents the number of spatial contour points of the electrochromic glass, represents the i-th solution value in the solution value of the fitting surface parameter, represents the fitted radial basis function, represents the coordinates of the spatial contour point i of the electrochromic glass, represents the coordinates of the spatial contour point j of the electrochromic glass, represents the Euclidean distance between points i and j, represents the surface adjustment solution value, Constructing a fitting surface of the electrochromic glass according to the fitting surface function value; a working parameter analysis module, used to perform physical coupling setting on the physical simulation model to obtain a target preparation model, query the real operating parameters corresponding to the target preparation model and the historical preparation performance data of the electrochromic glass, construct a performance mapping model of the electrochromic glass based on the real operating parameters and the historical preparation performance data, query the target performance requirements of the electrochromic glass, and analyze the target working parameters of the electrochromic glass based on the performance mapping model and the target performance requirements; A production simulation module is used to use the target working parameters to set parameters of the target preparation model to obtain a deployment model, perform data twinning on the deployment model to obtain a twin model, use the twin model to perform production simulation of the electrochromic glass, query the real-time monitoring data of the production simulation to analyze the preparation accuracy of the electrochromic glass, and when the preparation accuracy meets the set accuracy, use the twin model as the target preparation model of the electrochromic glass.

Images