A door and window installation dynamic wind pressure detection method and system based on intelligent sensors

Abstract

The application provides a door and window installation dynamic wind pressure detection method and system based on an intelligent sensor, relates to the technical field of door and window wind pressure detection, and comprises the following steps: based on a plurality of strain sensors embedded in door and window glass, collecting strain data of the door and window glass, combining point cloud data of the door and window glass, and constructing a three-dimensional deformation model of the door and window glass by using a deformation mapping method; based on the three-dimensional deformation model of the door and window glass, combining material parameters of the door and window glass, and calculating dynamic wind pressure distribution characteristics of a surface of the door and window glass by using a finite element inverse analysis method and an inversion compensation method; and according to the dynamic wind pressure distribution characteristics of the surface of the door and window glass, evaluating installation performance of the door and window to obtain a wind pressure performance evaluation detection result. The application is helpful for discovering installation problems in time and taking improvement measures, and improves installation quality and safety of the door and window.

Description

Technical Field

[0001] This invention relates to the field of wind pressure detection technology for doors and windows, and more specifically, to a method and system for dynamic wind pressure detection of door and window installations based on intelligent sensors. Background Technology

[0002] Glass doors and windows are an important channel for interaction between the home environment and the natural environment, and also one of the core elements of architectural design. With the continuous development of architectural design concepts, the area of ​​doors and windows in a single connected area (same opening) is showing a trend of increase. This change brings many impacts, the most significant being the proportional increase in the overall weight of doors and windows. This increased weight makes the mechanical environment experienced by doors and windows more complex when subjected to external forces such as wind pressure. Wind pressure, as one of the key factors affecting door and window performance, places higher demands on the installation quality of doors and windows due to its dynamic changes.

[0003] Improper installation of doors and windows, such as insecure connections or inadequate sealing, can cause abnormal deformation under wind pressure. This deformation not only affects the aesthetics and lifespan of the doors and windows, but more seriously, it can reduce their wind pressure resistance, thus threatening the overall safety of the building. For example, in strong winds, loosely installed doors and windows may sway, deform, or even fall off, causing damage to people and property inside.

[0004] However, current door and window installation defects can cause wind pressure to concentrate in local areas. Traditional detection methods mostly rely on manual visual inspection or single-point pressure testing, lacking quantitative analysis of the correlation between the overall wind pressure distribution and installation defects, making it difficult to detect potential installation problems.

[0005] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0006] In view of this, the present invention provides a method and system for dynamic wind pressure detection of door and window installation based on intelligent sensors, so as to solve the above-mentioned problems.

[0007] To solve the above problems, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, a method for dynamic wind pressure detection of door and window installations based on intelligent sensors is provided, comprising the following steps:

[0009] S1. Based on several strain sensors embedded in the door and window glass, the strain data of the door and window glass is collected, and combined with the point cloud data of the door and window glass, a three-dimensional deformation model of the door and window glass is constructed using the deformation mapping method.

[0010] S2. Based on the three-dimensional deformation model of the window glass, combined with the material parameters of the window glass, the dynamic wind pressure distribution characteristics on the surface of the window glass are calculated using the finite element inverse analysis method and the inversion compensation method.

[0011] S3. Based on the dynamic wind pressure distribution characteristics of the glass surface of doors and windows, evaluate the installation performance of doors and windows and obtain the wind pressure performance evaluation test results.

[0012] Preferably, the step of collecting strain data of the window glass based on several strain sensors embedded in the window glass, and constructing a three-dimensional deformation model of the window glass using the deformation mapping method in combination with the point cloud data of the window glass includes the following steps:

[0013] S11. Use strain sensors to collect strain data of door and window glass under wind pressure, and use the gross error elimination method to preprocess the strain data to obtain standard strain data.

[0014] S12. Combining the preset spatial coordinates of the strain sensor, and based on the generalized Hooke's law and the parameters of the door and window glass materials, and in conjunction with the structural deformation theory, the standard strain data is converted into a three-dimensional displacement field.

[0015] S13. Collect initial point cloud data of door and window glass under no wind pressure conditions, use the Poisson reconstruction algorithm to fit the initial point cloud data, and construct the initial three-dimensional geometric model of door and window glass.

[0016] S14. Map the three-dimensional displacement field onto the initial three-dimensional model, and adjust the deformation of the initial three-dimensional model to obtain the three-dimensional deformation model under wind pressure.

[0017] Preferably, the step of collecting strain data of door and window glass under wind pressure using strain sensors and preprocessing the strain data using a gross error elimination method to obtain standard strain data includes the following steps:

[0018] S111. Obtain the monitoring data stream of each strain sensor and calculate the first derivative sequence of each monitoring data stream;

[0019] S112. Using the wavelet thresholding method, calculate the outlier threshold for each first derivative sequence and remove gross errors in the first derivative sequence that exceed the outlier threshold.

[0020] S113. Perform wavelet decomposition on the monitoring data stream after removing data points, extract the low-frequency trend term, and use the Laida criterion to identify and remove secondary gross errors to obtain the removed data stream.

[0021] S114. Use linear interpolation to fill in the data points that were removed from the data stream to obtain standard strain data.

[0022] Preferably, the conversion of standard strain data into a three-dimensional displacement field by combining the preset spatial coordinates of the strain sensor, based on the generalized Hooke's law and the material parameters of the window and door glass, and in conjunction with structural deformation theory, includes the following steps:

[0023] S121. Based on standard strain data, determine the strain direction of the door and window glass, and construct the mapping relationship between the spatial coordinates and the strain direction according to the preset spatial coordinates of the strain sensor to form spatial strain distribution data.

[0024] S122. Obtain the elastic modulus and Poisson's ratio of the door and window glass material, and use the generalized Hooke's law to convert the spatial strain distribution data into stress tensor distribution data.

[0025] S123. Combining the structural boundary conditions and deformation theory of door and window glass, establish the governing equations between stress and displacement, and solve the governing equations by numerical methods to obtain the three-dimensional displacement vectors of each spatial point, forming a three-dimensional displacement field.

[0026] Preferably, the step of mapping the three-dimensional displacement field onto the initial three-dimensional model and adjusting the deformation of the initial three-dimensional model to obtain the three-dimensional deformation model under wind pressure includes the following steps:

[0027] S141. Based on the coordinate information of each spatial point in the three-dimensional displacement field, construct the correspondence between the point cloud nodes in the initial three-dimensional geometric model and form a matching dataset of model nodes and three-dimensional displacement.

[0028] S142. Based on the matching dataset of model nodes and 3D displacements, the 3D displacement vectors are superimposed on the corresponding point cloud nodes to realize the spatial position update of model node coordinates and form a set of deformed node coordinates.

[0029] S143. Reconstruct the model of the deformed node coordinate set to generate the initial three-dimensional deformation geometric model under wind pressure, and smooth the initial three-dimensional deformation geometric model to obtain the final three-dimensional deformation geometric model.

[0030] Preferably, the calculation of the dynamic wind pressure distribution characteristics on the surface of the window glass based on the three-dimensional deformation model of the window glass, combined with the material parameters of the window glass, and using the finite element inverse analysis method and inversion compensation method includes the following steps:

[0031] S21. Perform time-series discretization processing on the three-dimensional deformation model under wind pressure, and extract the displacement field information of the door and window glass surface at each time point.

[0032] S22. Based on the displacement field information of the glass surface of doors and windows, combined with the material parameters of the glass, the mechanical mapping relationship between deformation and wind pressure is established using the finite element inverse analysis method, and the preliminary dynamic wind pressure distribution of the glass is initially solved.

[0033] S23. Correct the initial dynamic wind pressure distribution by inversion compensation method, and extract the wind pressure peak, distribution area and time change characteristics based on the correction results to form the dynamic wind pressure distribution characteristics of the window glass surface.

[0034] Preferably, the step of performing time-series discretization processing on the three-dimensional deformation model under wind pressure to extract the displacement field information of the door and window glass surface at each time step includes the following steps:

[0035] S211. Determine the time interval of the time-series discrete based on the sampling frequency of the strain sensor;

[0036] S212. Extract the three-dimensional deformation model corresponding to each time step based on the time-series discrete time interval, and establish a spatial matching relationship between all model nodes and the initial three-dimensional model.

[0037] S213. Based on spatial matching relationships, calculate the three-dimensional displacement of the nodes of the three-dimensional deformation model relative to the initial three-dimensional model nodes at each moment, and complete the global displacement field information of the glass surface of the doors and windows at each moment through interpolation.

[0038] Preferably, the step of establishing the mechanical mapping relationship between deformation and wind pressure based on the displacement field information of the window glass surface, combined with the window glass material parameters, and using the finite element inverse analysis method to preliminarily solve the preliminary dynamic wind pressure distribution of the window glass includes the following steps:

[0039] S221. Based on the material parameters of window and door glass, construct a finite element inverse analysis model of window and door glass;

[0040] S222. The displacement field information of the glass surface of the doors and windows is used as a known boundary input to the finite element inverse analysis model. The mechanical mapping relationship between the deformation of the doors and windows and the wind pressure is determined by the structural mechanical equilibrium equation.

[0041] S223. Use finite element inverse analysis to solve the mapping relationship, calculate the wind pressure load value, and form a preliminary wind pressure spatial distribution;

[0042] S224. Combining the temporal displacement field information, the spatial distribution of wind pressure at each moment is integrated to obtain the dynamic wind pressure distribution on the surface of the door and window glass as time changes.

[0043] Preferably, the step of correcting the initial dynamic wind pressure distribution using the inversion compensation method, and extracting the wind pressure peak value, distribution area, and time variation characteristics based on the correction results to form the dynamic wind pressure distribution characteristics of the window glass surface includes the following steps:

[0044] S231. Based on the preliminary dynamic wind pressure distribution results, error analysis and local fitting correction are performed to obtain the correction results;

[0045] S232. Based on the corrected preliminary dynamic wind pressure distribution results, extract the wind pressure peak value, and extract the distribution area and time change characteristics through spatial clustering and time series statistics.

[0046] S233. Integrate the peak wind pressure, distribution area and time variation characteristics to obtain the dynamic wind pressure distribution characteristics on the window glass surface.

[0047] According to another aspect of the present invention, a dynamic wind pressure detection system for door and window installation based on intelligent sensors is provided, the system comprising:

[0048] The model building module is used to collect strain data of the door and window glass based on several strain sensors embedded in the glass, and combine the point cloud data of the door and window glass to construct a three-dimensional deformation model of the door and window glass using the deformation mapping method.

[0049] The feature calculation module is used to calculate the dynamic wind pressure distribution characteristics on the surface of the window glass based on the three-dimensional deformation model of the window glass, combined with the material parameters of the window glass, using the finite element inverse analysis method and the inversion compensation method.

[0050] The wind pressure performance evaluation module is used to evaluate the installation performance of doors and windows based on the dynamic wind pressure distribution characteristics on the glass surface of doors and windows, and obtain the wind pressure performance evaluation test results.

[0051] The beneficial effects of this invention are as follows: This invention collects strain data by embedding strain sensors and combines it with point cloud data to construct a three-dimensional deformation model, which can accurately reflect the deformation of window and door glass under wind pressure. Moreover, by screening and eliminating gross errors and filling in data through various methods, the accuracy of the data can be guaranteed. At the same time, this invention also uses the finite element inverse analysis method and the inversion compensation method to calculate the dynamic wind pressure distribution characteristics. From time series discretization to model construction, solution and correction, it can comprehensively and accurately obtain the peak wind pressure, distribution area and time change characteristics, providing a reliable basis for scientifically evaluating the installation performance of windows and doors, helping to discover installation problems in a timely manner and take improvement measures, thereby improving the installation quality and safety of windows and doors. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0053] Figure 1 This is a flowchart of a dynamic wind pressure detection method for door and window installation based on intelligent sensors according to an embodiment of the present invention;

[0054] Figure 2This is a schematic diagram of a dynamic wind pressure detection system for door and window installation based on intelligent sensors, according to an embodiment of the present invention.

[0055] In the picture:

[0056] 1. Model building module; 2. Feature calculation module; 3. Wind pressure performance evaluation module. Detailed Implementation

[0057] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0058] According to an embodiment of the present invention, a method and system for dynamic wind pressure detection of door and window installation based on intelligent sensors are provided.

[0059] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, a method for dynamic wind pressure detection of door and window installations based on intelligent sensors is provided, comprising the following steps:

[0060] S1. Based on several strain sensors embedded in the door and window glass, the strain data of the door and window glass is collected, and combined with the point cloud data of the door and window glass, a three-dimensional deformation model of the door and window glass is constructed using the deformation mapping method.

[0061] In a preferred embodiment, the step of collecting strain data of the window glass based on several strain sensors embedded in the window glass, and combining this data with the point cloud data of the window glass to construct a three-dimensional deformation model of the window glass using the deformation mapping method includes the following steps:

[0062] S11. Use strain sensors to collect strain data of door and window glass under wind pressure, and use the gross error elimination method to preprocess the strain data to obtain standard strain data.

[0063] In a preferred embodiment, the step of collecting strain data of window glass under wind pressure using a strain sensor and preprocessing the strain data using a gross error elimination method to obtain standard strain data includes the following steps:

[0064] S111. Obtain the monitoring data stream of each strain sensor and calculate the first derivative sequence of each monitoring data stream;

[0065] It should be noted that the monitoring data stream consists of the strain signal (such as voltage value or digital quantity) that each strain sensor outputs in real time and changes over time, forming a discrete time series. Then, the first derivative sequence is calculated using the forward difference method. The first derivative is used to characterize the rate of change of strain over time. The forward difference method approximates the derivative by calculating the difference in strain between adjacent time points.

[0066] S112. Using the wavelet thresholding method, calculate the outlier threshold for each first derivative sequence and remove gross errors in the first derivative sequence that exceed the outlier threshold.

[0067] It should be noted that for abnormal peaks (gross errors) in the first derivative sequence caused by sensor noise and transient interference (such as vibration and electromagnetic interference), a reasonable threshold is set using wavelet thresholding to preserve effective signals and remove abnormal gross data, thus avoiding interference with subsequent strain data processing. Specifically, the abnormal threshold can be calculated using the minimaxi criterion in wavelet analysis; then, abnormal data points are identified for each data point in the first derivative sequence.

[0068] S113. Perform wavelet decomposition on the monitoring data stream after removing data points, extract the low-frequency trend term, and use the Laida criterion to identify and remove secondary gross errors to obtain the removed data stream.

[0069] It should be noted that after the initial filtering, a small number of hidden gross errors may still remain in the monitoring data stream (such as outliers with small amplitudes but deviating from the trend), and the data may be affected by high-frequency noise. Wavelet decomposition can separate the signal into low-frequency trend terms and high-frequency detail terms. Then, the residual between the original monitoring data stream and the low-frequency trend term is calculated. Statistical analysis is performed on the residual sequence to calculate its mean and standard deviation. Then, the Laida criterion (3σ criterion) is used to identify secondary gross errors and remove them.

[0070] S114. Use linear interpolation to fill in the data points that were removed from the data stream to obtain standard strain data.

[0071] It should be noted that when using linear interpolation, the strain value at the missing position needs to be calculated and filled based on the valid data points before and after the missing segment. In this way, the continuity of the preprocessed strain data is restored, and finally standard strain data that can accurately reflect the strain state of the window glass under wind pressure is formed.

[0072] S12. Combining the preset spatial coordinates of the strain sensor, and based on the generalized Hooke's law and the parameters of the door and window glass materials, and in conjunction with the structural deformation theory, the standard strain data is converted into a three-dimensional displacement field.

[0073] As a preferred embodiment, the step of converting standard strain data into a three-dimensional displacement field by combining the preset spatial coordinates of the strain sensor and based on the generalized Hooke's law and the material parameters of the window and door glass, in conjunction with structural deformation theory, includes the following steps:

[0074] S121. Based on standard strain data, determine the strain direction of the door and window glass, and construct the mapping relationship between the spatial coordinates and the strain direction according to the preset spatial coordinates of the strain sensor to form spatial strain distribution data.

[0075] It should be noted that the standard strain data includes strain values ​​measured by each sensor (such as normal strain or shear strain). c xy Its direction is directly related to the sensor's mounting angle. For example, if the sensor is along the glass surface... x Paste along the axis, corresponding to the measured values x Directional strain e x ;along y The corresponding axis is pasted. e y .

[0076] S122. Obtain the elastic modulus and Poisson's ratio of the door and window glass material, and use the generalized Hooke's law to convert the spatial strain distribution data into stress tensor distribution data.

[0077] It should be noted that the elastic modulus of window and door glass (such as float glass and tempered glass) E The stress (approximately 70-72 GPa) and Poisson's ratio ν (approximately 0.22-0.23) can be determined from its factory specifications. As an isotropic linear elastic material, the stress-strain relationship of glass obeys the generalized Hooke's law, expressed as:

[0078] ;

[0079] In the formula, s Represents stress components, s xx , s yy , s zz They represent x , y , z Normal stress in the direction, e For normal strain components, c For shear strain components, s xy , s yz , s xz They represent x -y , yz , xz In-plane shear stress e xx , e yy , e zz They represent x , y , z The positive strain in the direction; c xy , c yz , c xz They represent x - y , yz , xz Shear strain in a plane. By combining stress components into a stress tensor, stress tensor distribution data over the entire domain is formed.

[0080] S123. Combining the structural boundary conditions and deformation theory of door and window glass, establish the governing equations between stress and displacement, and solve the governing equations by numerical methods to obtain the three-dimensional displacement vectors of each spatial point, forming a three-dimensional displacement field.

[0081] It should be noted that the boundary conditions of window and door glass are determined by their installation method with the frame. Common types include: fixed constraints with rigid connections between the glass edges and the frame; hinged constraints allowing slight rotation but zero displacement; and elastic constraints allowing minute displacement. Based on the boundary conditions and combining deformation theory from elasticity (such as geometric equations describing the derivative relationship between strain and displacement, and equilibrium equations ensuring stress distribution satisfies force balance), governing equations with three-dimensional displacement as unknowns can be established. Then, numerical methods are used to solve these governing equations, and the three-dimensional displacement vectors of each spatial point are obtained through inversion calculations, thereby constructing a three-dimensional displacement field that comprehensively reflects the displacement of various parts of the window and door glass.

[0082] Specifically, in engineering, numerical methods such as the finite element method are used to solve the problem. First, the glass is discretized into tiny elements, the element stiffness matrix is ​​constructed and assembled into a global matrix, and then the boundary conditions are substituted to solve the system of equations, thus deriving the value of each spatial point. x , y , z The displacement components in the direction ultimately form a three-dimensional displacement field covering the entire glass area.

[0083] S13. Collect initial point cloud data of door and window glass under no wind pressure conditions, use the Poisson reconstruction algorithm to fit the initial point cloud data, and construct the initial three-dimensional geometric model of door and window glass.

[0084] It should be noted that when collecting initial point cloud data of door and window glass under wind-free conditions, a high-precision 3D scanning device is used. Then, in an ideal environment without the influence of wind pressure, the door and window glass is scanned comprehensively and meticulously to obtain its initial point cloud data. This point cloud data is a set of 3D coordinates of numerous discrete points on the surface of the door and window glass, which can initially present the appearance outline of the glass.

[0085] The specific steps for fitting the initial point cloud data using the Poisson reconstruction algorithm to construct the initial three-dimensional geometric model of the window glass include: estimating the normal vector of each point in the point cloud data using algorithms such as Principal Component Analysis (PCA) to obtain the normal vector direction of each point; constructing an octree structure based on the distribution range and density of the point cloud data, which recursively divides the three-dimensional space into multiple small cubes, adapting to point cloud data of different densities through continuous subdivision; constructing an indicator function based on the point cloud data and normal vector information, and substituting it into the Poisson equation for solution; solving the Poisson equation using numerical calculation methods (such as the finite difference method and the finite element method) to obtain a continuous implicit function; and finally, extracting the isosurface from the implicit function using algorithms such as the moving cube algorithm, which is the surface of the initial three-dimensional geometric model of the window glass.

[0086] S14. Map the three-dimensional displacement field onto the initial three-dimensional model, and adjust the deformation of the initial three-dimensional model to obtain the three-dimensional deformation model under wind pressure.

[0087] In a preferred embodiment, mapping the three-dimensional displacement field onto the initial three-dimensional model and adjusting the deformation of the initial three-dimensional model to obtain a three-dimensional deformation model under wind pressure includes the following steps:

[0088] S141. Based on the coordinate information of each spatial point in the three-dimensional displacement field, construct the correspondence between the point cloud nodes in the initial three-dimensional geometric model and form a matching dataset of model nodes and three-dimensional displacement.

[0089] It should be noted that when constructing the correspondence between the point cloud nodes and the initial 3D geometric model, a spatial proximity search algorithm (such as KD tree search) is used. For each spatial point in the 3D displacement field, the nearest node is found in the point cloud nodes of the initial model. If the point density of the displacement field is higher than that of the initial model point cloud, a many-to-one matching is used (multiple displacement points correspond to the same model node), and the average displacement is taken as the representative displacement of that node. Finally, a matching dataset containing "initial model node ID, initial coordinates, and corresponding 3D displacement vector" is formed.

[0090] S142. Based on the matching dataset of model nodes and 3D displacements, the 3D displacement vectors are superimposed on the corresponding point cloud nodes to realize the spatial position update of model node coordinates and form a set of deformed node coordinates.

[0091] It should be noted that the point cloud node coordinates of the initial three-dimensional geometric model reflect the position when there is no wind pressure, while the three-dimensional displacement vector reflects the position change under the action of wind pressure. The initial deformation of the model can be achieved by superimposing the coordinates. Specifically, for each matched model node, the coordinates after deformation are obtained by adding the corresponding displacement vector to the initial coordinates.

[0092] Furthermore, for the boundary nodes connecting the glass and the frame (such as fixed constraint nodes), their displacement vectors should theoretically be (0,0,0), and the coordinates remain unchanged after superposition to conform to the physical characteristics of the boundary constraints; after all nodes are updated, a set of deformed node coordinates is formed.

[0093] S143. Reconstruct the model of the deformed node coordinate set to generate the initial three-dimensional deformation geometric model under wind pressure, and smooth the initial three-dimensional deformation geometric model to obtain the final three-dimensional deformation geometric model.

[0094] It should be noted that when reconstructing the model from the deformed set of node coordinates, a triangulation algorithm (such as Delaunay triangulation) is used. Based on the spatial coordinates of the deformed nodes, adjacent nodes are connected into triangular patches, which are then pieced together to form a complete glass surface mesh. For the glass thickness direction, a three-dimensional solid model is generated by stretching the surface mesh (keeping the thickness parameters constant), thus forming the initial three-dimensional deformed geometric model.

[0095] Furthermore, the initial reconstructed model may exhibit surface wrinkles or sharp edges due to differences in node displacements (such as abrupt changes in displacement between adjacent nodes), requiring smoothing adjustments. Smoothing adjustments typically employ filtering or optimization algorithms. Filtering algorithms, such as Gaussian filtering and median filtering, remove noise and sharp edges by calculating a weighted average or median of the vertex coordinates on the model surface. Optimization algorithms, on the other hand, define an objective function (such as the smoothness of the model surface or the uniformity of curvature) and use iterative optimization to adjust the positions of the model vertices until the objective function reaches its optimal value, thereby achieving smooth model adjustment.

[0096] S2. Based on the three-dimensional deformation model of the window glass, combined with the material parameters of the window glass, the dynamic wind pressure distribution characteristics on the surface of the window glass are calculated using the finite element inverse analysis method and the inversion compensation method.

[0097] As a preferred embodiment, the calculation of the dynamic wind pressure distribution characteristics on the surface of the window glass based on the three-dimensional deformation model of the window glass, combined with the material parameters of the window glass, and using the finite element inverse analysis method and inversion compensation method includes the following steps:

[0098] S21. Perform time-series discretization processing on the three-dimensional deformation model under wind pressure, and extract the displacement field information of the door and window glass surface at each time point.

[0099] As a preferred embodiment, the step of performing time-series discretization processing on the three-dimensional deformation model under wind pressure and extracting the displacement field information of the door and window glass surface at each time point includes the following steps:

[0100] S211. Determine the time interval of the time-series discrete based on the sampling frequency of the strain sensor;

[0101] It should be noted that the sampling frequency refers to the number of times the sensor collects data per unit time, usually expressed in Hertz (Hz). The time interval is the length of time between two consecutive samples, and it is the reciprocal of the sampling frequency. For example, if the sampling frequency of a strain sensor is... f Hz, then the time interval Δ t =1 / f Second.

[0102] S212. Extract the three-dimensional deformation model corresponding to each time step based on the time-series discrete time interval, and establish a spatial matching relationship between all model nodes and the initial three-dimensional model.

[0103] It should be noted that the nodes of the initial 3D model and the 3D deformation model at each time step must correspond one-to-one. The nearest neighbor search algorithm can be used to find the node in the initial 3D model that is spatially closest (≤0.05mm) to the node of the 3D deformation model as the matching object.

[0104] S213. Based on spatial matching relationships, calculate the three-dimensional displacement of the nodes of the three-dimensional deformation model relative to the initial three-dimensional model nodes at each moment, and complete the global displacement field information of the glass surface of the doors and windows at each moment through interpolation.

[0105] It should be noted that for the initial model node with established spatial matching relationship and the three-dimensional deformation model node at each time step, the displacement components of the node in the three coordinate axis directions can be obtained by calculating the difference between their coordinates, and thus the three-dimensional displacement can be obtained.

[0106] However, due to limitations such as sensor measurements, it is impossible to directly obtain the 3D deformation model node information for all locations on the surface of window and door glass. Interpolation methods can be used to estimate the displacement of unknown locations based on the displacement information of known nodes using a specific mathematical model, thereby completing the global displacement field information.

[0107] Specifically, commonly used interpolation methods include linear interpolation, cubic spline interpolation, and radial basis function interpolation. For example, when using linear interpolation, for two known nodes... P 1( x 1, y 1, z 1) and P 2( x 2,y 2, z 2) and its displacement d 1 and d 2. Any point on the line segment between them P ( x , y , z The displacement can be calculated using a linear interpolation formula.

[0108] S22. Based on the displacement field information of the glass surface of doors and windows, combined with the material parameters of the glass, the mechanical mapping relationship between deformation and wind pressure is established using the finite element inverse analysis method, and the preliminary dynamic wind pressure distribution of the glass is initially solved.

[0109] As a preferred embodiment, the step of establishing the mechanical mapping relationship between deformation and wind pressure based on the displacement field information of the window glass surface, combined with the window glass material parameters, and using the finite element inverse analysis method, and preliminarily solving the preliminary dynamic wind pressure distribution of the window glass includes the following steps:

[0110] S221. Based on the material parameters of window and door glass, construct a finite element inverse analysis model of window and door glass;

[0111] It should be noted that the finite element inverse analysis method is developed based on the traditional finite element forward analysis. Forward analysis solves for the displacement, stress, and other responses of a structure based on known material parameters, geometric models, and load conditions; while inverse analysis, given the structural responses (such as the displacement field information in this problem), uses a suitable mathematical model and algorithm to inversely deduce the loads (wind pressure) on the structure. When constructing the finite element inverse analysis model, it is necessary to consider parameters such as the constitutive relation, elastic modulus, and Poisson's ratio of the material, as well as the geometric characteristics such as the actual size, shape, and installation method of the window and door glass.

[0112] Specifically, an accurate geometric model is established in finite element software based on the actual size and shape of the window glass. Then, the geometric model is divided into a finite number of elements, and the boundary conditions of the model are defined. Finally, in the finite element software, the collected material parameters are assigned to each element in the geometric model so that the model can accurately reflect the material properties of the window glass.

[0113] S222. The displacement field information of the glass surface of the doors and windows is used as a known boundary input to the finite element inverse analysis model. The mechanical mapping relationship between the deformation of the doors and windows and the wind pressure is determined by the structural mechanical equilibrium equation.

[0114] It should be noted that the structural mechanics equilibrium equations describe the equilibrium state of a structure under load. For window and door glass under wind pressure, these equations essentially reflect the mechanical principle that the internal stress and external force of this structural system reach a dynamic equilibrium when subjected to the external load of wind pressure. When wind pressure acts on the surface of window and door glass, the glass will undergo corresponding deformation, which will lead to stress distribution within the glass.

[0115] The structural mechanics equilibrium equations mathematically express the balance between stress and wind pressure: the resultant force of the stress components at each point inside the glass in the three coordinate directions, the equivalent force generated by the wind pressure in the corresponding direction, and the possible volume forces within the glass itself sum to zero. By inputting the displacement field information of the window glass surface as known boundary conditions into the finite element inverse analysis model, the model can, based on the structural mechanics equilibrium equations and starting from the known deformation (displacement) conditions, gradually derive and determine the precise mechanical mapping relationship between the window glass deformation and wind pressure.

[0116] S223. Use finite element inverse analysis to solve the mapping relationship, calculate the wind pressure load value, and form a preliminary wind pressure spatial distribution;

[0117] It should be noted that the process of solving the mechanical mapping relationship through finite element inverse analysis is an iterative optimization process. By continuously adjusting the initial guess value of the wind pressure load, the error between the calculated displacement field and the actual measured displacement field is minimized. When the error meets a certain accuracy requirement, the current wind pressure load value can be considered reasonable, thus obtaining the spatial distribution of wind pressure.

[0118] Specifically, in each iteration, the software calculates the displacement field of the window glass based on the current wind pressure load value and compares it with the actual measured displacement field to calculate the error. Then, it adjusts the wind pressure load value based on the error and proceeds to the next iteration until the error meets the preset accuracy requirements. After the iteration is complete, the calculated wind pressure load value is mapped onto each node on the window glass surface, forming a preliminary spatial distribution of wind pressure. The post-processing function of the finite element software can be used to visually display the wind pressure distribution in the form of cloud maps, contour lines, etc.

[0119] S224. Combining the temporal displacement field information, the spatial distribution of wind pressure at each moment is integrated to obtain the dynamic wind pressure distribution on the surface of the door and window glass as time changes.

[0120] S23. Correct the initial dynamic wind pressure distribution by inversion compensation method, and extract the wind pressure peak, distribution area and time change characteristics based on the correction results to form the dynamic wind pressure distribution characteristics of the window glass surface.

[0121] In a preferred embodiment, the step of correcting the initial dynamic wind pressure distribution using the inversion compensation method, and extracting the wind pressure peak value, distribution area, and time variation characteristics based on the correction results to form the dynamic wind pressure distribution characteristics of the window glass surface includes the following steps:

[0122] S231. Based on the preliminary dynamic wind pressure distribution results, error analysis and local fitting correction are performed to obtain the correction results;

[0123] It should be noted that when measuring the displacement field information of the glass surface of doors and windows, the measuring equipment used (such as laser displacement sensors) itself has accuracy limitations and will produce measurement errors. One is random error (such as instantaneous fluctuations caused by environmental vibration and light interference), which manifests as small oscillations of the displacement value around the true value; the other is systematic deviation (such as sensor calibration deviation and installation angle deviation), which manifests as the displacement measurement value deviating from the true value as a whole (such as always being too large or too small).

[0124] Specifically, the error analysis stage requires two steps to locate the problem: First, calculate the deviation between the preliminary wind pressure distribution and the "measured values ​​of a small number of high-precision pressure sensors" (e.g., select 3-5 calibration points at the center and edge of the glass, calculate the mean square error, and identify system deviations (e.g., a large overall mean square error indicates the existence of systematic displacement measurement errors); Second, locate the areas of concentrated error through spatial error distribution cloud maps (e.g., glass corners near sensor measurement blind spots, where large data interpolation errors can easily lead to abnormal wind pressure).

[0125] Furthermore, for regions dominated by random errors, the Local Weighted Regression (LOESS) algorithm is used, with nearby reliable data points (such as actual pressure sensor measurement points) as weight centers to smooth local wind pressure fluctuations; for regions dominated by systematic biases (such as the overall offset zone caused by sensor calibration deviations), an error compensation function is established through polynomial fitting, such as: p 修 = p 初 + k ( x , y ), k ( x , y ) represents the compensation coefficient that varies with position. p 修 The original wind pressure calculation results obtained from displacement field inversion before correction. p 初 The wind pressure value after local fitting correction is used to back-calculate the compensation coefficient by combining the measured value of the verification point to achieve local correction. At the same time, the correction process needs to constrain the smoothness of the wind pressure gradient (such as the wind pressure change rate of adjacent units does not exceed the threshold) to avoid introducing new physical irrationality due to local adjustment (such as sudden wind pressure changes in invariant areas).

[0126] S232. Based on the corrected preliminary dynamic wind pressure distribution results, extract the wind pressure peak value, and extract the distribution area and time change characteristics through spatial clustering and time series statistics.

[0127] It should be noted that peak wind pressure refers to the maximum wind pressure value that occurs on the surface of window or door glass within a certain period of time. By scanning the corrected preliminary dynamic wind pressure distribution results over time, recording the wind pressure values ​​at each spatial location at different times, and then finding the maximum value, the peak wind pressure at that spatial location can be obtained.

[0128] Spatial clustering is a method of classifying adjacent or similar data points in space into the same class. When extracting wind pressure distribution areas, the surface of door and window glass can be divided into different areas according to the magnitude of the wind pressure peak or the variation characteristics of the wind pressure value. Statistical analysis of the wind pressure data in each wind pressure distribution area is performed on the time series to calculate the average value, standard deviation, and frequency of change of wind pressure and other time variation characteristic indicators.

[0129] S233. Integrate the peak wind pressure, distribution area and time variation characteristics to obtain the dynamic wind pressure distribution characteristics on the window glass surface.

[0130] S3. Based on the dynamic wind pressure distribution characteristics of the glass surface of doors and windows, evaluate the installation performance of doors and windows and obtain the wind pressure performance evaluation test results.

[0131] Specifically, during the evaluation, the measured dynamic wind pressure distribution characteristics of the window and door glass surface can be compared with relevant standards (such as the wind pressure resistance performance indicators of windows and doors specified in industry standards or design specifications); if the measured value exceeds the standard value, it indicates that the installation performance of the windows and doors does not meet the requirements.

[0132] like Figure 2 As shown, according to an embodiment of the present invention, a dynamic wind pressure detection system for door and window installation based on intelligent sensors is provided, the system comprising:

[0133] Model building module 1 is used to collect strain data of the door and window glass based on several strain sensors embedded in the door and window glass, and combine the point cloud data of the door and window glass to construct a three-dimensional deformation model of the door and window glass using the deformation mapping method.

[0134] Feature calculation module 2 is used to calculate the dynamic wind pressure distribution characteristics on the surface of the window glass based on the three-dimensional deformation model of the window glass and the material parameters of the window glass, using the finite element inverse analysis method and the inversion compensation method.

[0135] The wind pressure performance evaluation module 3 is used to evaluate the installation performance of doors and windows based on the dynamic wind pressure distribution characteristics on the glass surface of doors and windows, and obtain the wind pressure performance evaluation test results.

[0136] In summary, by utilizing the above-mentioned technical solutions of this invention, the present invention collects strain data by embedding strain sensors and constructs a three-dimensional deformation model by combining it with point cloud data. This model can accurately reflect the deformation of window and door glass under wind pressure. Furthermore, by using various methods to screen and eliminate gross errors and fill in data, the accuracy of the data can be guaranteed. At the same time, this invention also uses the finite element inverse analysis method and the inversion compensation method to calculate the dynamic wind pressure distribution characteristics. From time-series discretization to model construction, solution, and correction, it can comprehensively and accurately obtain the peak wind pressure, distribution area, and time variation characteristics. This provides a reliable basis for scientifically evaluating the installation performance of windows and doors, helps to promptly identify installation problems and take improvement measures, and improves the quality and safety of window and door installation.

[0137] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0138] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for dynamic wind pressure detection of door and window installations based on intelligent sensors, characterized in that, Includes the following steps: S1. Based on several strain sensors embedded in the door and window glass, the strain data of the door and window glass is collected, and combined with the point cloud data of the door and window glass, a three-dimensional deformation model of the door and window glass is constructed using the deformation mapping method. S2. Based on the sampling frequency of the strain sensor, determine the time interval of the time-series discrete step; extract the three-dimensional deformation model corresponding to each time step according to the time-series discrete step, and establish a spatial matching relationship between all model nodes and the initial three-dimensional model; Based on spatial matching relationships, the three-dimensional displacement of the nodes of the three-dimensional deformation model relative to the initial three-dimensional model nodes at each time moment is calculated, and the global displacement field information of the glass surface of the door and window at each time moment is formed by interpolation. Based on the displacement field information of the window glass surface and combined with the window glass material parameters, the mechanical mapping relationship between deformation and wind pressure is established using the finite element inverse analysis method, and the preliminary dynamic wind pressure distribution of the window glass is initially solved. The preliminary dynamic wind pressure distribution is corrected by the inversion compensation method, and the wind pressure peak, distribution area and time change characteristics are extracted according to the correction results to form the dynamic wind pressure distribution characteristics of the window glass surface. S3. Based on the dynamic wind pressure distribution characteristics of the glass surface of doors and windows, evaluate the installation performance of doors and windows and obtain the wind pressure performance evaluation test results.

2. The method for dynamic wind pressure detection of door and window installation based on intelligent sensors according to claim 1, characterized in that, The process of collecting strain data from window and door glass using several strain sensors embedded in the glass, and then combining this data with point cloud data to construct a three-dimensional deformation model of the window and door glass using a deformation mapping method includes the following steps: S11. Use strain sensors to collect strain data of door and window glass under wind pressure, and use the gross error elimination method to preprocess the strain data to obtain standard strain data. S12. Combining the preset spatial coordinates of the strain sensor, and based on the generalized Hooke's law and the parameters of the door and window glass materials, and in conjunction with the structural deformation theory, the standard strain data is converted into a three-dimensional displacement field. S13. Collect initial point cloud data of door and window glass under no wind pressure conditions, use the Poisson reconstruction algorithm to fit the initial point cloud data, and construct the initial three-dimensional geometric model of door and window glass. S14. Map the three-dimensional displacement field onto the initial three-dimensional model, and adjust the deformation of the initial three-dimensional model to obtain the three-dimensional deformation model under wind pressure.

3. The method for dynamic wind pressure detection of door and window installation based on intelligent sensors according to claim 2, characterized in that, The process of collecting strain data of window glass under wind pressure using strain sensors and preprocessing the strain data using gross error elimination to obtain standard strain data includes the following steps: S111. Obtain the monitoring data stream of each strain sensor and calculate the first derivative sequence of each monitoring data stream; S112. Using the wavelet thresholding method, calculate the outlier threshold for each first derivative sequence and remove gross errors in the first derivative sequence that exceed the outlier threshold. S113. Perform wavelet decomposition on the monitoring data stream after removing data points, extract the low-frequency trend term, and use the Laida criterion to identify and remove secondary gross errors to obtain the removed data stream. S114. Use linear interpolation to fill in the data points that were removed from the data stream to obtain standard strain data.

4. The method for dynamic wind pressure detection of door and window installation based on intelligent sensors according to claim 2, characterized in that, The process of converting standard strain data into a three-dimensional displacement field by combining the preset spatial coordinates of the strain sensor, based on the generalized Hooke's law and the material parameters of the window and door glass, and in conjunction with structural deformation theory, includes the following steps: S121. Based on standard strain data, determine the strain direction of the door and window glass, and construct the mapping relationship between the spatial coordinates and the strain direction according to the preset spatial coordinates of the strain sensor to form spatial strain distribution data. S122. Obtain the elastic modulus and Poisson's ratio of the door and window glass material, and use the generalized Hooke's law to convert the spatial strain distribution data into stress tensor distribution data. S123. Combining the structural boundary conditions and deformation theory of door and window glass, establish the governing equations between stress and displacement, and solve the governing equations by numerical methods to obtain the three-dimensional displacement vectors of each spatial point, forming a three-dimensional displacement field.

5. A method for dynamic wind pressure detection of door and window installation based on intelligent sensors according to claim 2, characterized in that, The process of mapping the three-dimensional displacement field onto the initial three-dimensional model and adjusting the deformation of the initial three-dimensional model to obtain the three-dimensional deformation model under wind pressure includes the following steps: S141. Based on the coordinate information of each spatial point in the three-dimensional displacement field, construct the correspondence between the point cloud nodes in the initial three-dimensional geometric model and form a matching dataset of model nodes and three-dimensional displacement. S142. Based on the matching dataset of model nodes and 3D displacements, the 3D displacement vectors are superimposed on the corresponding point cloud nodes to realize the spatial position update of model node coordinates and form a set of deformed node coordinates. S143. Reconstruct the model of the deformed node coordinate set to generate the initial three-dimensional deformation geometric model under wind pressure, and smooth the initial three-dimensional deformation geometric model to obtain the final three-dimensional deformation geometric model.

6. The method for dynamic wind pressure detection of door and window installation based on intelligent sensors according to claim 5, characterized in that, The process of establishing a mechanical mapping relationship between deformation and wind pressure based on the displacement field information of the window glass surface, combined with the window glass material parameters, and using the finite element inverse analysis method to preliminarily solve the preliminary dynamic wind pressure distribution of the window glass includes the following steps: S221. Based on the material parameters of window and door glass, construct a finite element inverse analysis model of window and door glass; S222. The displacement field information of the glass surface of the doors and windows is used as a known boundary input to the finite element inverse analysis model. The mechanical mapping relationship between the deformation of the doors and windows and the wind pressure is determined by the structural mechanical equilibrium equation. S223. Use finite element inverse analysis to solve the mapping relationship, calculate the wind pressure load value, and form a preliminary wind pressure spatial distribution; S224. Combining the temporal displacement field information, the spatial distribution of wind pressure at each moment is integrated to obtain the dynamic wind pressure distribution on the surface of the door and window glass as time changes.

7. The method for dynamic wind pressure detection of door and window installation based on intelligent sensors according to claim 6, characterized in that, The process of correcting the initial dynamic wind pressure distribution using the inversion compensation method, and extracting the wind pressure peak value, distribution area, and time variation characteristics based on the correction results to form the dynamic wind pressure distribution characteristics of the window glass surface includes the following steps: S231. Based on the preliminary dynamic wind pressure distribution results, error analysis and local fitting correction are performed to obtain the correction results; S232. Based on the corrected preliminary dynamic wind pressure distribution results, extract the wind pressure peak value, and extract the distribution area and time change characteristics through spatial clustering and time series statistics. S233. Integrate the peak wind pressure, distribution area and time variation characteristics to obtain the dynamic wind pressure distribution characteristics on the window glass surface.

8. A dynamic wind pressure detection system for door and window installation based on intelligent sensors, used to implement the dynamic wind pressure detection method for door and window installation based on intelligent sensors as described in any one of claims 1-7, characterized in that, The system includes: The model building module is used to collect strain data of the door and window glass based on several strain sensors embedded in the glass, and combine the point cloud data of the door and window glass to construct a three-dimensional deformation model of the door and window glass using the deformation mapping method. The feature calculation module is used to calculate the dynamic wind pressure distribution characteristics on the surface of the window glass based on the three-dimensional deformation model of the window glass, combined with the material parameters of the window glass, using the finite element inverse analysis method and the inversion compensation method. The wind pressure performance evaluation module is used to evaluate the installation performance of doors and windows based on the dynamic wind pressure distribution characteristics on the glass surface of doors and windows, and obtain the wind pressure performance evaluation test results.

Images