A method and system for controlling the construction quality of doors and windows in building engineering

Abstract

The present invention relates to the technical field of construction engineering, specifically a construction engineering door and window construction quality control method and system, including: obtaining door and window design data, determining each characterization information during door and window installation, and setting test parameters of the door and window under each characterization information according to the characterization information during door and window installation; using the test parameters to test the door and window to obtain basic calibration data and gap influence data; calibrating the target quality data under the current test; obtaining the characteristic attributes calibrated during door and window construction according to the target quality data, and using the characteristic attributes to construct a characteristic association network, calculating the association influence path of the characteristic association network under each characteristic attribute; using the association influence path to determine the main influence range during door and window construction, and evaluating the current door and window construction quality according to the data within the main influence range; achieving the overall construction quality and efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of construction engineering, and specifically to a method and system for controlling the construction quality of doors and windows in construction engineering. Background Art

[0002] In construction engineering, the construction quality of doors and windows directly affects the overall performance of the building, including but not limited to aspects such as waterproofing, sound insulation, heat insulation, and safety. However, in the actual construction process, due to the lack of precise quality control methods and effective monitoring means, deviations often occur in the installation of doors and windows, affecting the final use effect. Therefore, how to effectively evaluate and control the construction quality of doors and windows has become an important research topic.

[0003] For example, the Chinese patent application with the publication number CN118070607A discloses a method, system, and device for predicting the risks of doors and windows based on stress tracking. In view of the problem that it is difficult to accurately predict the breakage risks that are prone to occur in doors and windows in the prior art, the following solutions are proposed, including the following steps: Step 1: Obtain the structural attributes of the doors and windows, where the structural attributes of the doors and windows at least include environmental information and structural information; Step 2: Establish a finite element model of the door and window structure; The prior art shows that the dangerous situations during the construction of doors and windows are determined by stress analysis of the door and window structure.

[0004] For example, the Chinese patent application with the publication number CN111177837A discloses a data processing method, device, and electronic device based on a three-dimensional building model. By processing the obtained three-dimensional building model, a structured spatial plane model is obtained, the spatial adjacency state between the spatial units of the structured spatial plane model and the spatial adjacency type of adjacent spatial units are determined, and a business space connection relationship network is generated to meet the needs of different services and reduce the overall engineering cost of obtaining and managing building space relationship information.

[0005] The prior art describes the relative situation of doors and windows in management by measuring the stress of doors and windows and the relative spatial adjacency of doors and windows. However, these situations tend to view the basic attributes of doors and windows and cannot quickly view the current quality situation of doors and windows. Multiple indicators are required to describe the actual situation of the current doors and windows, resulting in the inability to comprehensively identify the main elements affected when there are corresponding gaps and deformations in the doors and windows, thus making the evaluation effect of the doors and windows poor. Summary of the Invention

[0006] To solve the above technical problems, the technical solution adopted by the present invention is: A method for controlling the construction quality of doors and windows in construction engineering, including: S1, obtaining the door and window design data, determining each characterization information during the installation of the doors and windows, identifying the fitting situation of the doors and windows at each edge position according to each characterization information during the installation of the doors and windows, and setting the test parameters of the doors and windows under each characterization information.

[0007] S2. Use the test parameters to test the doors and windows, and obtain the basic calibration data and gap influence data; obtain the edge integrity and edge fitting indexes during the construction of the doors and windows through the basic calibration data.

[0008] S3. Through the gap influence data, obtain the magnitude of the gap change and the gap change level during the construction of the doors and windows, and use the edge integrity, edge fitting indexes, magnitude of the gap change, and gap change level as input features to calibrate the target quality data under the current test.

[0009] S4. According to the target quality data, obtain the characteristic attributes calibrated during the construction of the doors and windows, and use the characteristic attributes to construct a characteristic association network, and calculate the association influence path of the characteristic association network under each characteristic attribute.

[0010] S5. Use the association influence path to determine the main influence range during the construction of the doors and windows, and evaluate the current construction quality of the doors and windows according to the data within the main influence range.

[0011] The implementation method of the characterization information in step S1 includes: extracting multiple construction nodes during the construction of the doors and windows according to the door and window design data, where the construction nodes represent each installation position during the installation of the doors and windows.

[0012] Generate the characterization information during the installation of the doors and windows according to the completion degree of each construction node, and after sorting the construction nodes in the order of completion of each construction node and labeling each construction node, obtain the characterization information during the installation of the doors and windows.

[0013] The implementation method of step S1 also includes: classifying the characterization information according to the characterization information during the installation of the doors and windows, and determining the corresponding test parameter sequence under each characterization information.

[0014] According to the association relationship between the test parameter sequences, segment and space the corresponding content of each characterization information to obtain the standardized sequence corresponding to the test parameters, and form the test parameters under each characterization information under multiple tests.

[0015] The implementation method of step S2 includes: S21. Obtain the basic coordinates of each component of the doors and windows from the basic calibration data, compare the basic coordinates with the coordinates of each component corresponding to the current construction of the doors and windows, and sequentially determine the coordinate deviation of each component.

[0016] S22. Check the coordinate deviation of each component, and respectively identify the adjacency relationship between the coordinate deviations of each component on adjacent components.

[0017] S23. In response to the adjacency relationship between the coordinate deviations of each component, determine whether each adjacent component is a window installation structure, whether it includes doors, holes, dividing lines, and windows, and set the edge integrity according to the content described under the adjacency relationship.

[0018] S24. Extract the errors of the sealing material attached to the doors and windows under the coordinate deviations of each component, and set the edge fitting index according to the relative positions of the coordinate deviations of each component and the sealing material.

[0019] The implementation method of step S3 includes: S31. Take the edge integrity, edge fitting index, gap change magnitude, and gap change level as input features, set a probability deviation sequence according to the occurrence probabilities of the input features, sequentially judge the conditional probabilities of the input features after combination, and perform principal component analysis on the input features according to the values of the conditional probabilities. Use the data output by the principal component analysis as the target features.

[0020] S32. Perform an association mapping on the target features to determine the approaching degree of the target features, and sort the data obtained by the association mapping of the target features according to the approaching degree of the target features and output them as target quality data.

[0021] The implementation method of step S32 also includes: Process the data obtained by the association mapping of the target features, identify the numerical interval that the target features approach, and set the approaching degree for the target features according to the upper limit value, lower limit value, and average value of the numerical interval that the target features approach.

[0022] The implementation method of step S4 includes: S41. Sequentially extract the feature attributes from the target quality data according to the working sequence of door and window installation, and determine the node types corresponding to the feature attributes.

[0023] S42. Take the feature attributes as nodes according to the node types corresponding to the feature attributes, calculate the weights between the feature attributes under each node type, and construct a feature association network.

[0024] S43. Calculate the path coefficients of each path in the feature association network, and select the shortest path with the largest path coefficient value as the association influence path according to the path coefficients.

[0025] The implementation method of the feature association network also includes: Obtain the door and window structure relationships corresponding to the feature attributes, identify the optimal connected subsets corresponding to the feature attributes, and connect the feature attributes according to the optimal connected subsets to obtain the feature association network.

[0026] The implementation method of step S5 includes: Randomly select a first sample point from the association influence path, perform linear regression on the first sample point to determine the calculation result of the first sample point; and select a second sample point from the association influence path according to the calculation result of the first sample point.

[0027] Determine the comprehensive weight of the first sample point and the second sample point under the associated influence path, perform information matching on the associated influence path according to the comprehensive weight, determine the weight distribution of the associated influence path after information matching, and take the area with the largest weight distribution as the main influence range under the door and window construction.

[0028] Compare the feature descriptions of each node within the main influence range with the preset database to determine the priority of the door and window construction assessment, and obtain the assessment result of the door and window construction.

[0029] A quality control system for door and window construction in a construction project, comprising: a data acquisition module for collecting various original data related to door and window construction from the site, including but not limited to design drawings and construction records, and combining the various original data into door and window design data.

[0030] A test analysis module for extracting and performing specific test operations from the door and window design data to determine the executed test parameters and the test data corresponding to the test parameters.

[0031] A feature classification module for extracting key features from the test data. The key features include edge integrity, edge fitting index, gap change size, and gap change level, and for classifying and managing the key features.

[0032] An associated network construction module for constructing a feature association network based on the extracted key features and calculating the associated influence paths between different key features.

[0033] A quality assessment module for comprehensively assessing the quality of door and window construction according to the associated influence path and providing specific improvement suggestions to the user.

[0034] The beneficial effects of the present invention are as follows: First, the present invention obtains relevant data during door and window installation, selects test parameters for door and window testing from these relevant data, and realizes parametric modeling of construction quality, which can avoid subjectivity during manual inspection. It tends to consider the fitting and corresponding gap conditions at various positions under door and window construction to describe the specific situation of the current door and window construction, improving the controllability and observability of door and window quality construction.

[0035] Second, the present invention uses multiple indicators measured by test parameters, constructs a feature association network using these indicators, which helps to identify the main factors affecting the quality of door and window construction and their interaction mechanisms; describes door and window construction according to the association between multiple feature factors to further describe the actual content measured by the test parameters, so as to realize the process of quantitative analysis; at the same time, according to the connection conditions under the feature association network, the relative weights in the network can be identified, which is convenient for subsequent expression of the influence degree of each construction node on the overall quality for the main test content in the feature association network, improving the efficiency of overall construction quality testing and analysis.

[0036] III. The present invention extracts the associated influence path through the feature association network to determine the main influence range of the door and window construction; finally, based on the data within the main influence range, the current door and window construction quality is evaluated; it describes the main influencing factors during the door and window construction and evaluates the overall door and window construction according to these factors to improve the overall construction quality and efficiency, reduce the influence between construction problems during construction, and reduce the problem of the decline in the door and window construction quality caused by the data affected by the construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The present invention will be further described below in conjunction with the drawings and embodiments.

[0038] Figure 1 It is a schematic flowchart of a method for controlling the construction quality of doors and windows in a construction project.

[0039] Figure 2 It is a schematic flowchart of step S2 of a method for controlling the construction quality of doors and windows in a construction project.

[0040] Figure 3 It is a schematic flowchart of step S3 of a method for controlling the construction quality of doors and windows in a construction project.

[0041] Figure 4 It is a schematic flowchart of step S4 of a method for controlling the construction quality of doors and windows in a construction project.

[0042] Figure 5 It is a schematic diagram of a system for controlling the construction quality of doors and windows in a construction project. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] The embodiments of the present invention will be described in detail below. The following described embodiments are exemplary and are only used to explain the present invention and should not be construed as a limitation to the present invention. For those not specified in the embodiments regarding specific technologies or conditions, they shall be carried out according to the technologies or conditions described in the literature in this field or according to the product specifications.

[0044] Refer to Figure 1 , a method for controlling the construction quality of doors and windows in a construction project, including: S1, obtaining the door and window design data, determining each characterization information during the door and window installation, identifying the fitting conditions of the doors and windows at each edge position according to each characterization information during the door and window installation, and setting the test parameters of the doors and windows under each characterization information.

[0045] S2, testing the doors and windows using the test parameters to obtain the basic calibration data and the gap influence data; obtaining the edge integrity and edge fitting index of the doors and windows during the construction through the basic calibration data.

[0046] S3. Affect data through the gap, obtain the magnitude of the gap change and the gap change level during the construction of doors and windows, and use the edge integrity, edge fitting index, magnitude of the gap change, and gap change level as input features to calibrate the target quality data under the current test.

[0047] S4. According to the target quality data, obtain the characteristic attributes calibrated during the construction of doors and windows, use the characteristic attributes to construct a characteristic association network, and calculate the associated influence paths of the characteristic association network under each characteristic attribute.

[0048] S5. Use the associated influence paths to determine the main influence range during the construction of doors and windows, and evaluate the current construction quality of doors and windows according to the data within the main influence range.

[0049] In an embodiment of the present invention, the implementation method of the characterization information in step S1 includes: extracting multiple construction nodes during the construction of doors and windows according to the door and window design data, where the construction nodes represent each installation position during the installation of doors and windows, such as the construction of the outer window and the lower opening of the outer door, and each position where the doors and windows will be constructed is regarded as a construction node.

[0050] Generate the characterization information during the installation of doors and windows according to the completion degree of each construction node. After sorting the construction nodes in the order of completion of each construction node, label each construction node to obtain the characterization information during the installation of doors and windows.

[0051] The completion degree of the construction node indicates whether the work at the corresponding position is completed during the construction of doors and windows, and label the construction node according to the completion situation. At the same time, images or other forms of coordinate data after completion can be collected to describe the corresponding characterization information.

[0052] The characterization information will represent information such as the fitting dimensions, levelness and perpendicularity, and diagonal error between the door and window frame and the opening at this time, and is used to describe the relevant position coordinates and corresponding image information during the installation of doors and windows, as the description information for comparing the installation situation of doors and windows with the actual quality of door and window installation.

[0053] At the same time, the characterization information can also represent information on multiple data such as airtight performance, watertight performance, sealing strip state, wind pressure resistance performance, and surface flatness preset for each position in the door and window design data; these parameters will be extracted as the main part representing the current performance of the doors and windows in the door and window design data at this time, and then the doors and windows corresponding to these parameters will be tested, or a part of the parameters will be selected as the test parameters for subsequent use according to the actual fitting situation of the doors and windows.

[0054] Air tightness: The air infiltration volume when the doors and windows are closed, reflecting their heat insulation and energy-saving effects. Watertightness: The ability of doors and windows to resist rainwater leakage under simulated wind and rain conditions. Status of the sealing strip: Check whether the sealing strip is installed in place and there is no aging or cracking. Wind pressure resistance: The deformation and load-bearing capacity of doors and windows under a certain wind pressure. These four parameters can be used to judge their corresponding performances through the gaps of doors and windows during testing.

[0055] Surface flatness means that there should be no obvious scratches, depressions or deformations on the surface of the doors and windows. This part can be analyzed by collecting images of the installed doors and windows and modeling to find out the integrity of each part of the door and window installation, so as to identify whether the whole meets the installation requirements.

[0056] When collecting and processing the data of door and window installation, a feeler gauge and a vernier caliper can be used to detect the gaps between the door and window sashes and frames, and between the frames and the walls, and compare with the allowable deviation values.

[0057] Then use a laser level and a straightedge to check the verticality and horizontality of the door and window frames to find out whether there are deviations in these positions. At the same time, these data can be randomly sampled in the area after the building construction is completed, and three-dimensional laser scanning, infrared thermal imaging and BIM models can be used for verification. Generate a three-dimensional point cloud model for the information of each door and window installed on site, and calculate the deviation between the positions of these models and the door and window design data at each actual construction position to find out whether there are weak points during the door and window installation. The characterization information corresponding to these weak points will be used as the main content for subsequent processing and identification.

[0058] At the same time, a pressure box can also be used for the pressure test of doors and windows to check the leakage and anti-deformation conditions of doors and windows. At this time, mainly use each position where the doors and windows are installed to check the connectivity deviation to determine whether there are deviations in the current door and window construction. The content with deviations will be used as the test parameters for subsequent use, and the overall edge integrity and deviation of the door and window will be tested.

[0059] When setting the test parameters of doors and windows under various characterization information, mainly determine the current main test parameters according to the data sampled on site, such as the geometric dimensions, gaps, and edge positions of the door and window installation and the deviation from the door and window design data. The test parameters described here do not mean that the construction quality of the current door and window only adopts the selected several test parameters. Under normal door and window measurement, the waterproof, windproof, heat insulation, sound insulation and other dimensions will be measured for testing. The present invention focuses on processing the parts with obvious problems at this time and displays the associated influence path in a visual network, such as "sealant failure → rainwater leakage → wall mildew → circuit safety hazard", to describe what impacts the currently identified test parameters may cause to evaluate the construction quality of the doors and windows.

[0060] Therefore, the implementation method of step S1 further includes: classifying each characterization information during the installation of the doors and windows, and determining the corresponding test parameter sequence under each characterization information. At this time, the classified test parameter sequence will represent the installation accuracy, that is, the parameters corresponding to the dimensions and the parameters corresponding to the performance, and then a sequence representing multiple test parameters is obtained.

[0061] According to the correlation relationship between each test parameter sequence, segment and space the corresponding content of each characterization information to obtain the standardized sequence corresponding to the test parameters, and form the test parameters under each characterization information for multiple tests. The correlation relationship between each test parameter sequence is to couple each test parameter. For example, the error of the opening size can be associated with information such as airtightness (air leakage caused by too large a gap) and watertightness (rainwater infiltration), and the deviation of the level / verticality is associated with the reduction of wind pressure resistance performance (easy deformation due to uneven stress). These contents will be used as the correlation relationship between each test parameter sequence. At the same time, this correlation relationship will be set in the database in advance to assist in subsequent processing of the combination of each test parameter sequence.

[0062] After that, the segmentation and spacing are to utilize the correlation relationship between each test parameter, segment these related parameters, mainly based on the test parameters that can cause the correlation, to segment the parameters in multiple test parameter sequences, and to describe a standardized sequence with mutual influence. There are multiple test parameters in the standardized sequence, and then the test parameters in the standardized sequence are output to realize the main component analysis of the test parameters.

[0063] Under different characterization information, the test parameters will be somewhat different. For example, when each characterization information represents the dimensions and coordinates of the current doors and windows, the width / height deviation, installation surface deviation, and diagonal length difference of the window frame can be used for setting as the main recognized test parameters.

[0064] For example, when identifying the opening matching size and geometric accuracy, the test parameters can be expressed as: opening size error: width / height deviation ≤ ±3mm (according to the "Architectural Construction Drawing Design" specification); level / verticality: installation surface deviation ≤ ±2mm (measured using a laser theodolite); diagonal error: diagonal length difference of the window frame ≤ 5mm (reflecting the degree of rectangular deformation).

[0065] The test parameters described here are only part of the parameters extracted for the doors and windows during their construction. The actual measured and selected parameters will be adjusted in other forms according to the measurement requirements and the actual design method of the doors and windows.

[0066] After that, the test parameters can also include the deviation between the actual coordinates and the design values, the width of the gap between the window frame and the wall, and the results of firmness tests, etc.; Installation position deviation: the deviation between the actual coordinates and the design values ≤ ±5 mm (compared using the BIM model); Gap control: the width of the gap between the window frame and the wall ≤ 3 mm (filled with polyurethane foam); Firmness test: no looseness of the window frame when hammered (impact force ≥ 0.5 kN).

[0067] In a certain project, due to the uneven wall, the window frame was tilted. The thickness of the wooden shims was adjusted in layers (each layer ≤ 2 mm) and refixed, reducing the perpendicularity deviation from 4 mm to 1.5 mm. The tightness was confirmed by using a feeler gauge.

[0068] When the test parameters are performance - type test parameters, other forms of parameters are mainly used for description, such as the test parameters in the cases of airtightness and watertightness, wind pressure resistance and surface quality, etc.

[0069] Airtightness grade: air infiltration rate per unit seam length (Q1 - Q8, Q8 is the best); Watertightness grade: simulated wind and rain pressure difference (ΔP1 - ΔP6, ΔP6 corresponds to the rainstorm level); Wind pressure resistance grade: P3 - P8 (P8 can resist a typhoon of level 12); Surface flatness: the gap detected by a 2 - m straightedge ≤ 2 mm (to avoid scratches and deformation).

[0070] Extract from the current test parameters according to the values represented by these parameters to find out the mainly used test parameters.

[0071] In an embodiment of the present invention, the basic calibration data is the performance reference value reflecting the doors and windows in the ideal installation state. These data will reflect the normal standard values of the test parameters during the test of the doors and windows, and use this data to compare the integrity at the edges of the doors and windows and the actual fitting conditions of the adhesives, sealing strips, etc. attached at the edges, and verify the content of the gaps generated by the doors and windows in each case, so as to verify whether it can be represented as complete at the edges, and use these calculated indicators as the constraint conditions for their use to complete the inspection of the construction quality of the doors and windows.

[0072] The edge integrity represents the geometric matching degree between the doors and windows and the edge of the opening. The measurement parameters are mainly the error values representing the installation accuracy of the doors and windows, such as the minimum size deviation, diagonal error, and surface flatness, etc., to determine that the window frame has no collision and deformation, and the errors in the positions of each component in its installation all meet the construction specifications.

[0073] The edge fitting index represents the contact quality between the sealing material and the window frame / wall. The test parameters are mainly the compression amount of the sealing strip and the uniformity of the fitting gap. These values can measure the width and depth of the gaps at multiple points using a caliper, and then judge whether the errors in the relative positions of the window form meet the installation requirements after applying the sealing material.

[0074] Such asFigure 2 As shown, the implementation of step S2 includes: S21, obtaining the basic coordinates of each component of the door and window from the basic calibration data, comparing the basic coordinates with the coordinates of each component corresponding to the current door and window construction, and sequentially determining the coordinate deviation of each component.

[0075] S22, checking the coordinate deviation of each component, respectively identifying the adjacency relationship between the coordinate deviations of each component on adjacent components. At this time, it shows whether these deviations appear together on adjacent components when there is a deviation in the door and window components. If they appear together, it means that both adjacent components have problems, or due to a problem with one component, multiple deviations occur.

[0076] S23, in response to the adjacency relationship between the coordinate deviations of each component, determining whether each adjacent component is a window installation structure, whether it includes doors, holes, dividing lines, and windows, and setting the edge integrity according to the content described in the adjacency relationship. At this time, the edge integrity is to combine data such as coordinate errors existing according to the adjacency relationship to describe whether its edge is complete. This index value will represent the error situation at the corresponding position.

[0077] The described window sill installation structure can be components with direct physical contact between the door and window frames, walls, etc. Then, the doors, holes, dividing lines, and windows indicate whether the current coordinate deviation of each component is located on the window, and according to the position where it is located, the edge integrity at this time is described. The edge integrity mainly describes the generated deviation.

[0078] S24, extracting the error of the sealing material attached to the door and window under the coordinate deviation of each component, and setting the edge fitting index according to the relative position of the coordinate deviation of each component and the sealing material. The edge fitting index finally describes the error that occurs when attaching the sealing material to the corresponding component, and combines this error with the corresponding position on the door and window.

[0079] At this time, it is to process the coordinate errors generated on the door and window, and explain these coordinate errors according to the position of the component and the relative content of the sealing strip to represent the corresponding position when the coordinate error occurs. For the related errors of the sealing material, the relative coordinates can also be identified by means of image recognition and compared with the basic calibration data to obtain the related errors.

[0080] In an embodiment of the present invention, the gap influence data describes the gaps generated during the performance test of the door and window construction quality. These gaps will affect the waterproof, windproof and other performances of the door and window during normal use. For example, if the gaps set in the window frame change due to wind force test, heating test, then the size and level of change of the gap lock change represent the interval, which will indicate the decline of the current window itself performance and the reduction of the door and window construction quality.

[0081] For example, the minimum gap between the door and window and the edge of the opening, the gap at the normal sliding groove and other positions during the installation of the door and window. Measure these gaps to determine the overall stability of the current door and window under multiple simulated test environments. The gap change level will set multiple intervals according to the numerical value of the gap change size. For example, according to the gap width specified in the airtightness Q4 standard, describe the gap change level at this time. The intervals corresponding to the gap change level will be preset in the database according to the data requirements for each level under construction standardization, and describe the main situation during the construction of this door and window.

[0082] The target quality data reflects the key indicators of the door and window installation quality, including the main part of these test parameters, to find out the main problems that may exist during the current door and window construction, and use the data corresponding to these problems as the target quality data to analyze the problems in the door and window construction quality.

[0083] For example Figure 3 As shown, the implementation method of step S3 includes: S31, taking the edge integrity, edge fitting index, gap change size and gap change level as input features, setting a probability deviation sequence according to the occurrence probability of each input feature, successively judging the conditional probability of each input feature after combination, and performing principal component analysis on the input features according to the value of each conditional probability, and taking the data output by the principal component analysis as the target feature.

[0084] When using principal component analysis, first standardize the numerical values of the edge integrity, edge fitting index, gap change size and gap change level corresponding to the input features, convert them into a value range of 0 - 1, then calculate the occurrence probability of each input feature, that is, the distribution of this numerical value based on historical data or experimental data, and then obtain the relative weight of each input feature according to the proportion of each occurrence probability in the sum of the occurrence probabilities of the total data. Then convert this probability value into a deviation value. For example, subtract the average value of the occurrence probability from the occurrence probability of the input feature to describe the deviation that each input feature can generate. After that, use the deviations of these probabilities to form a probability deviation sequence.

[0085] After that, the input features are combined in pairs to determine the conditional probability that multiple input features are satisfied simultaneously after combination. When performing principal component analysis later, after obtaining the covariance matrix of the probability deviation of the input features, the weights of two corresponding elements in the covariance matrix are set to the difference between the conditional probability of the two corresponding input features and the occurrence probability of the input feature corresponding to a single element. For example, if there are two elements A and B in the covariance matrix, when considering the weights of A and B, the weight can be P(B|A) - P(B) to determine the relative weight of element B when element A occurs, and to describe the adjustment of the content of this principal component analysis under the influence of conditional probability. When performing eigenvalue decomposition later, the existing technology will be used to implement the form of eigenvalue extraction. The eigenvalues extracted from the covariance matrix will be calculated in this case, and when the cumulative variance contribution rate of the data output by the principal component reaches 0.95, the corresponding data will be used as the current target feature.

[0086] S32. Perform an association mapping on the target feature to determine the degree of approximation of the target feature, and sort the data associated with the target feature according to the degree of approximation of the target feature and then output it as the target quality data.

[0087] The association mapping of the target feature is to find the data corresponding to the current target feature, including the positions, coordinates on the doors and windows, and the description information recorded by these features. Then, the degree of approximation of the target feature will represent the numerical interval to which the corresponding input feature approximates. For example, the upper limit value, lower limit value, and average value under normal conditions. The data corresponding to the target feature is represented by multiple values to describe the situation of the approximation numerical interval. Then, after setting multiple values according to the numerical interval to which the target feature approximates, sorting is performed, and three forms of target quality data can be obtained.

[0088] Therefore, the implementation method of step S32 also includes: processing the data associated with the target feature, identifying the numerical interval to which the target feature approximates, and setting the degree of approximation of the target feature according to the upper limit value, lower limit value, and average value of the numerical interval to which the target feature approximates.

[0089] For the trend degree of the target feature, the TOPSIS method can be used to set this degree of approximation, that is, dividing the absolute value difference between this value and the negative ideal solution by the sum of the absolute value differences between this value and the negative ideal solution and the positive ideal solution. When setting the degree of approximation, select the part with the maximum calculated trend degree among the upper limit value, lower limit value, and average value of the numerical interval of this feature for setting, and then perform sorting to obtain the target quality data.

[0090] In an embodiment of the present invention, in step S4, the features included in the target quality data are used as the feature attributes to be used. Then, the feature attributes are used as nodes, and the association relationships between the feature attributes are used as edges to construct a feature association network. For the association relationships, the Pearson correlation coefficient or the Spearman rank correlation coefficient can be used to calculate the feature attributes. The calculation is mainly performed on the corresponding values of pairwise feature attributes. It should be noted that the values represented by the feature attributes are all normalized data, and there is a mapping relationship between the normalized data and the original data corresponding to the feature attributes, which is convenient for directly expressing the specific values represented by each node in the feature association network later.

[0091] As Figure 4 shown, the implementation method of step S4 includes: S41, extracting the feature attributes from the target quality data in sequence according to the working sequence of window and door installation, and determining the node types corresponding to the feature attributes. The node types include input feature nodes corresponding to the edge integrity, edge fitting index, gap change size, and gap change level, the intermediate variable node type describing the test parameters, and the target quality node corresponding to the selection of the target quality data. These nodes represent the content represented by the test parameters mainly included in the target quality attributes.

[0092] S42, taking each feature attribute as a node according to the node type corresponding to the feature attribute, calculating the weights between the feature attributes under each node type, and constructing a feature association network. As shown in Table 1, the extracted feature attributes are divided into continuously associated node forms, and these nodes are combined to obtain a feature association network, which represents the situation when the feature attributes are associated. Then, these associations and problems are combined to describe what should be represented when relevant failure problems occur.

[0093] Table 1. Schematic table of node types of feature attributes

[0094]

[0095] Table 1 shows the content that can be included when the feature attributes are described using node types, and the quantized values corresponding to the nodes of the feature attributes under various node types. For example, the input feature node represents the probability deviation relative to the ideal situation, and the intermediate variable node represents the occurrence probability of the intermediate variable node on the condition of the input feature node, to describe the relationship between the nodes at each level or type under the network connection. Then, the target quality node uses the same dimension as the input feature node to judge, and uses the probability distribution and discrete form to describe the relative situation of the target quality node. As shown in Table 2, the weights between various types of direct nodes in the feature association network can be described in the following way.

[0096] Table 2. Schematic table of weights

[0097]

[0098] The association situations adopted by each of these three types are described in Table 2, that is, the weights between input feature nodes will be implemented in the way of calculating the correlation coefficient according to the corresponding data, and the correlation coefficient can be represented by the Pearson correlation coefficient; for the input feature nodes and the target quality nodes, a linear regression analysis will be adopted, taking the input features as independent variables and the target quality as the dependent variable, calculating the regression coefficient, and using the regression coefficient as the weight between the input feature nodes and the target quality nodes. When connecting the input feature nodes, intermediate variable nodes, and target quality nodes, the product of the weights on this path will be used to represent the corresponding path coefficient; for the weight from the intermediate variable to the target variable, a linear regression analysis will also be adopted, taking the input features as independent variables and the target quality as the dependent variable to obtain the corresponding regression coefficient, and this regression coefficient is the weight from the intermediate variable to the target variable.

[0099] After that, the paths under these weights are connected to obtain an association influence path related to the currently collected data.

[0100] The implementation method of the feature association network also includes: obtaining the door and window structure relationships corresponding to each feature attribute, identifying the optimal connected subsets corresponding to each feature attribute, and connecting each feature attribute according to the optimal connected subsets to obtain the feature association network.

[0101] At this time, it is further described whether the corresponding features on the doors and windows can have structural connections. After connecting the nodes corresponding to the feature attributes, a recursive feature elimination method is used for processing to find the optimal subset in the corresponding situation.

[0102] When extracting the optimal connected subsets, they are extracted according to the correlation coefficients corresponding to each feature attribute. After that, the input feature nodes with a correlation coefficient greater than 0.5 are selected, and when the regression coefficients between the input feature nodes and the target quality nodes, and between the intermediate variable nodes and the target quality nodes are greater than 0.6, the nodes that can be connected by the feature attributes at this time are used as the optimal subset. Then, each feature attribute is connected to obtain the feature association network.

[0103] S43. Calculate the path coefficients of each path in the feature association network, and select the shortest path with the largest path coefficient value as the association influence path according to the path coefficients.

[0104] At this time, after selecting the connection of feature attributes under different input features and corresponding node types, the relevant problems existing during the installation of doors and windows are mainly represented in this network, and the features that may correspond to these problems are combined. Finally, the association of different feature attributes is obtained, and further described according to the errors that appear in these feature attributes to judge the construction quality represented as a whole.

[0105] In an embodiment of the present invention, in step S5, mainly according to the node types included in the obtained associated influence path and the connection method of these nodes, the content that will be mainly affected by the current door and window construction is found, and the content affected by this part is marked as the main influence range. Then, the feature information described by the main influence range is used to describe the quality of the current door and window construction, which is convenient for subsequent engineering personnel to analyze according to the description of the corresponding door frame and judge whether the construction is completed normally.

[0106] Therefore, the implementation method of step S5 includes: randomly selecting a first sample point from the associated influence path, performing linear regression on the first sample point to determine the calculation result of the first sample point; and selecting a second sample point from the associated influence path according to the calculation result of the first sample point. The first sample point represents any one of the input feature node, the intermediate variable node, and the target quality node. Then, the value of this node under the parameters calculated by each node described in the entire associated influence path is calculated. For example, the change trend or relative association strength of this node is quantified, that is, the relative weight corresponding to this first sample point and other nodes is obtained. Then, a second sample point is selected to represent the comprehensive weight of the two points under the associated influence path. Then, by selecting the group of the first sample point and the second sample point with the largest value according to the comprehensive weight, the most obvious groups of parameters measured under the current door and window construction can be known, and thus it can be determined that this problem may exist in the current door and window construction.

[0107] Determine the comprehensive weight of the first sample point and the second sample point under the associated influence path, and perform information matching on the associated influence path according to the comprehensive weight to determine the weight distribution of the associated influence path after information matching. The area with the largest weight distribution is used as the main influence range under the door and window construction.

[0108] For the implementation method of information matching, it is to calculate the comprehensive weight of the corresponding data of the first sample point and the second sample point under the current association path, compare the corresponding values of each group of the first sample point and the second sample point, and then determine the distribution and value of the comprehensive weight. Then, select the area corresponding to the part with the largest value in the weight distribution sequence of the comprehensive weight combination, and identify this area as the main problem area under the current door and window construction quality test. The content measured and the information described by the corresponding nodes in this area will indicate the main problem points in the door and window construction, such as the edge integrity and the gap change level measured at this node. It is found that the gap change level has increased by 30% compared to the content measured under normal conditions, which means that there are some small gaps in the installation of this door frame that have not been properly handled, or the gap of this door and window changes significantly under tests such as wind force and water force for the corresponding materials used for this door and window, and it is necessary to adjust the door and window materials used during construction to illustrate the main problems existing during the current door and window construction.

[0109] Compare the feature descriptions of each node within the main influence range with the preset database to determine the priority of the door and window construction assessment and obtain the assessment result of the door and window construction.

[0110] Compare the content expressed by the data of each node within the main influence range with the preset database to explain what quality problems exist in the current door and window construction for the data measured at this time. Then, according to the positions of these quality problems on the associated influence path, determine the priority of the current problems in the door and window construction. The priority can be represented according to the relative weights of the nodes represented by this data and other nodes. Then, output the connection situation, position, etc. of the nodes represented by these quality problems. This output is the assessment result, which is convenient for subsequent engineering personnel to check the construction problems existing at this time.

[0111] As Figure 5 shown, the present invention also provides a quality control system for building engineering door and window construction, including: a data acquisition module, a test analysis module, a feature classification module, an association network construction module, and a quality assessment module; wherein, the output end of the data acquisition module is connected to the test analysis module, the output end of the test analysis module is connected to the feature classification module, the output end of the feature classification module is connected to the association network construction module, and the output end of the association network construction module is connected to the quality assessment module.

[0112] The data acquisition module is used to collect various original data related to door and window construction from the site, including but not limited to design drawings and construction records, and combine the various original data into door and window design data.

[0113] The test analysis module is used to extract and perform specific test operations from the door and window design data to determine the test parameters to be executed and the test data corresponding to the test parameters.

[0114] A feature classification module, configured to extract key features from test data. The key features include edge integrity, edge fitting index, gap change magnitude, and gap change level, and classify and manage the key features.

[0115] An association network construction module, configured to construct a feature association network based on the extracted key features and calculate the association influence paths between different key features.

[0116] A quality assessment module, configured to comprehensively evaluate the construction quality of doors and windows according to the association influence paths and provide specific improvement suggestions to users.

[0117] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention, and still be covered by the protection scope of the present invention.

Claims

1. A method for controlling the construction quality of doors and windows in construction engineering, characterized in that, Including: S1. Obtain the door and window design data, determine each characterization information during door and window installation, identify the fitting conditions of the door and window at each edge position according to each characterization information during door and window installation, and set the test parameters of the door and window under each characterization information; S2. Use the test parameters to test the door and window to obtain the basic calibration data and the gap influence data; obtain the edge integrity and edge fitting index of the door and window during construction through the basic calibration data; S3. Through the gap influence data, obtain the size and level of gap change of the door and window during construction, and use the edge integrity, edge fitting index, size of gap change and level of gap change as input features to calibrate the target quality data under the current test; The implementation method of step S3 includes: S31. Use the edge integrity, edge fitting index, size of gap change and level of gap change as input features, set the probability deviation sequence according to the occurrence probability of each input feature, sequentially judge the conditional probability of each input feature after combination, and perform principal component analysis on the input features according to the value of each conditional probability, and use the data output by the principal component analysis as the target feature; S32. Perform an association mapping on the target feature to determine the approximation degree of the target feature, and sort the data obtained by the association mapping of the target feature according to the approximation degree of the target feature and output it as the target quality data; S4. According to the target quality data, obtain the characteristic attributes calibrated during door and window construction, use the characteristic attributes to construct a characteristic association network, and calculate the association influence path of the characteristic association network under each characteristic attribute; The implementation method of step S4 includes: S41. Extract the characteristic attributes from the target quality data in sequence according to the working order of door and window installation, and determine the node type corresponding to each characteristic attribute; S42. Use each characteristic attribute as a node according to the node type corresponding to each characteristic attribute, calculate the weights between the characteristic attributes under each node type, and construct a characteristic association network; S43. Calculate the path coefficient of each path in the characteristic association network, and select the shortest path with the largest path coefficient value as the association influence path according to the path coefficient; S5. Use the association influence path to determine the main influence range during door and window construction, and evaluate the current door and window construction quality according to the data within the main influence range.

2. The quality control method for window and door construction in a building project according to claim 1, characterized in that The implementation method of the characterization information in step S1 includes: Extract multiple construction nodes during door and window construction according to the door and window design data, where the construction nodes represent each installation position during door and window installation; Generate the characterization information during door and window installation according to the completion degree of each construction node, sort the construction nodes in the order of completion of each construction node, and label each construction node to obtain each characterization information during door and window installation.

3. A construction project door and window construction quality control method according to claim 1, characterized in that, The implementation method of step S1 also includes: Classify each characterization information according to the characterization information during door and window installation, and determine the corresponding test parameter sequence under each characterization information; Segment and space the corresponding content of each characterization information according to the association relationship between each test parameter sequence to obtain the standardized sequence corresponding to the test parameter, and form the test parameters under each characterization information for multiple tests.

4. A construction project door and window construction quality control method according to claim 1, characterized in that, The implementation method of step S2 includes: S21. Obtain the basic coordinates of each component of the door and window from the basic calibration data, compare the basic coordinates with the coordinates of each component corresponding to the current door and window construction, and sequentially determine the coordinate deviation of each component. S22. Check the coordinate deviation of each component, and respectively identify the adjacency relationship between the coordinate deviations of each component on adjacent components. S23. In response to the adjacency relationship between the coordinate deviations of each component, determine whether each adjacent component is a window installation structure, whether it includes doors, holes, dividing lines, and windows, and set the edge integrity according to the content described under the adjacency relationship. S24. Extract the error of the sealing material attached to the door and window under the coordinate deviation of each component, and set the edge fitting index according to the relative position of the coordinate deviation of each component and the sealing material.

5. A construction project door and window construction quality control method according to claim 1, characterized in that, The implementation method of step S32 also includes: Process the data for the associated mapping of the target feature, identify the numerical range that the target feature approaches, and set the approaching degree of the target feature according to the upper limit value, lower limit value, and average value of the numerical range that the target feature approaches.

6. The construction quality control method for doors and windows in construction projects according to claim 1, characterized in that, The implementation method of the feature association network also includes: Obtain the door and window structure relationship corresponding to each feature attribute, identify the optimal connected subset corresponding to each feature attribute, and connect each feature attribute according to the optimal connected subset to obtain the feature association network.

7. A construction project door and window construction quality control method according to claim 1, characterized in that, The implementation method of step S5 includes: Randomly select a first sample point from the associated influence path, perform linear regression on the first sample point to determine the calculation result of the first sample point; and select a second sample point from the associated influence path according to the calculation result of the first sample point. Determine the comprehensive weight of the first sample point and the second sample point under the associated influence path, perform information matching on the associated influence path according to the comprehensive weight, determine the weight distribution of the associated influence path after information matching, and take the area with the largest weight distribution as the main influence range under the door and window construction. Compare the feature descriptions of each node in the main influence range with the preset database to determine the priority of the door and window construction assessment and obtain the assessment result of the door and window construction.

8. A construction project door and window construction quality control system, characterized in that Including: A data acquisition module, which is used to collect various original data related to door and window construction from the site, including but not limited to design drawings and construction records, and combine various original data into door and window design data. A test analysis module, which is used to extract and execute specific test operations from the door and window design data to determine the executed test parameters and the test data corresponding to the test parameters. A feature classification module, which is used to extract key features from the test data. The key features include edge integrity, edge fitting index, gap change size, and gap change level, and classify and manage the key features. Take the edge integrity, edge fitting index, gap change size, and gap change level as input features, set the probability deviation sequence according to the occurrence probability of each input feature, sequentially judge the conditional probability of each input feature after combination, and perform principal component analysis on the input features according to the value of each conditional probability. Take the data output by the principal component analysis as the target feature. Perform associated mapping on the target features, determine the approaching degree of the target features, and sort the data associated with the target features according to the approaching degree of the target features, and then output it as target quality data; An associated network construction module, which is used to construct a feature association network based on the extracted key features and calculate the associated influence paths between different key features; Extract the feature attributes from the target quality data in sequence according to the working sequence of door and window installation, and determine the node types corresponding to the respective feature attributes; Take each feature attribute as a node according to the node type corresponding to each feature attribute, calculate the weights between the feature attributes under each node type, and construct a feature association network; Calculate the path coefficients of each path in the feature association network, and select the shortest path with the largest path coefficient value as the associated influence path according to the path coefficients; Extract the feature attributes from the target quality data in sequence according to the working sequence of door and window installation, and determine the node types corresponding to the respective feature attributes; Take each feature attribute as a node according to the node type corresponding to each feature attribute, calculate the weights between the feature attributes under each node type, and construct a feature association network; Calculate the path coefficients of each path in the feature association network, and select the shortest path with the largest path coefficient value as the associated influence path according to the path coefficients; A quality evaluation module, which is used to comprehensively evaluate the construction quality of doors and windows according to the associated influence paths and provide specific improvement suggestions to the user.

Images