Virtual scene construction quality training and assessment system building method based on BIM technology

Abstract

The invention provides a virtual scene construction quality training and assessment system building method based on a BIM technology. The method comprises the steps of knowledge base classification; classifying quality problems; building a branch engineering BIM scene learning model; the model is lightweight; binding a corresponding quality problem component and a corresponding problem viewport in the branch engineering BIM scene misedition model; the trainee carries out training and learning by clicking the branch project; building BIM scene models of various different stages and regions in the construction process; according to the quality problem classification, separately establishing wrong version models with quality problems in different BIM construction scenes; strongly quantifying the model; setting examination questions; configuring different assessment scores and setting a score assignment rule; creating examination paper; and the student clicks on the examination module to start the examination. According to the invention, a construction quality common fault problem learning library can be provided, and project field management personnel can be helped to learn and master reasons and prevention and control measures of project construction quality problems of all sub-parts at any time.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of BIM, and particularly relates to a virtual scene construction quality training and examination system construction method based on BIM technology. BACKGROUND

[0002] The construction quality has always been a difficult problem for engineering personnel, and the high-quality development of the construction industry is particularly important for the construction quality; and with the digital transformation of the construction industry, BIM is a key technology to realize digitalization, and BIM technology is increasingly widely used in the field of building construction and is increasingly mature, and in the aspect of engineering personnel capability improvement, the three-dimensional visualization and informatization characteristics of BIM technology provide construction quality learning training and examination for project managers, and improve the quality consciousness of construction managers and the level of on-site quality management.

[0003] How to improve the construction quality management consciousness of project site engineering personnel, how to improve the quality problem identification of each sub-item engineering and the cause analysis of quality problems, how to effectively rectify after the quality problem occurs and how to effectively prevent similar quality problems in the future, these are the focus of the current construction quality training and examination. At present, the construction quality examination based on revit and navisworks software environment realizes BIM, and this mode has the following defects: firstly, computer equipment is needed; secondly, the local end of the computer equipment needs to download and install BIM related software revit or navis works, and the configuration requirement of the equipment is high; in addition, the current BIM quality examination, the examination personnel only selects the quality problem by selecting the problem; finally, the current BIM quality examination, the scene is simple, and each scene is only set up for an independent problem. SUMMARY

[0004] In view of the defects in the prior art, the application provides a virtual scene construction quality training and examination system construction method based on BIM technology, which can provide a quality problem learning library to help project site managers master the causes and solutions of construction quality problems of each sub-item engineering at any time.

[0005] The application achieves the above technical purpose through the following technical means.

[0006] A virtual scene construction quality training and examination system construction method based on BIM technology comprises the following processes:

[0007] Step 1: classify the training module knowledge base;

[0008] Step 2: classify construction quality problems;

[0009] Step 3: establish a sub-contract engineering BIM learning model;

[0010] Step 4: Upload the BIM learning model of each sub-project to the existing system, perform lightweight model processing, and then bind it to the ten sub-project categories obtained in Step 1.

[0011] Step 5: Bind each quality issue to the corresponding issue component and issue viewport in the BIM learning model of the sub-project;

[0012] Step 6: Relevant personnel can directly click on a specific sub-project on the trainee's end to display the BIM scene model of that sub-project. Clicking on a quality issue will directly zoom in on the corresponding issue area and display it in the viewport for training and learning.

[0013] Step 7: Based on multiple actual engineering projects, establish multiple BIM scene models for different stages and areas during the construction process;

[0014] Step 8: Based on the classification of quality issues, create misprinted models of potential quality problem areas or points in different BIM construction scenarios;

[0015] Step 9: On the management side, upload the local BIM construction scene model through the model management module. After being lightweighted, it will be classified into the "Assessment Model" category.

[0016] Step 10: Set each BIM construction scene as an assessment question; set n assessment options, including m quality issue options, and fix the BIM scene model components and display viewports for each of the n assessment options;

[0017] Step 11: Based on the number m of quality issues in each BIM construction scenario, configure different assessment scores and set specific answer keywords for scoring;

[0018] Step 12: Create an assessment paper, select the personnel to be assessed, the assessment opening time, the assessment duration, and other factors, and set the total score for each assessment;

[0019] Step 13: Relevant personnel click "Start" in the assessment module to enter the exam. They will then enter different construction building BIM scene models and, through immersive interactive browsing, identify potential quality issues, mark them, and record in detail the causes of the issues and the rectification and corrective measures.

[0020] Step 14: After the exam, analyze the exam results, comparing your answers to the reference answers for each question to help you identify areas for improvement.

[0021] Furthermore, step 2 specifically includes: classifying the construction quality problems encountered during the project construction process into each sub-project, and for each quality problem, compiling the construction phenomenon of the quality problem, the cause analysis, the quality remedial measures for the problem, and the methods and measures to avoid such problems in subsequent construction processes.

[0022] Furthermore, in step 3, BIM learning models for each sub-project are established. In each sub-project BIM learning model, different quality problem manifestations from step 2 are set, forming ten types of sub-project BIM learning models that summarize the various quality problems.

[0023] Furthermore, the specific methods for model lightweighting in steps 4 and 9 are as follows:

[0024] Upload and parsing: Upload the created RVT format model to the cloud platform. The cloud platform parses the file and extracts the model's collection, material, and texture information.

[0025] Geometric optimization: Optimize the geometric data of the RVT format model, including vertex reduction, edge simplification, and patch merging. The specific algorithm formula is as follows:

[0026]

[0027] Where E is the total error, v k It is the original vertex, v. k It is the vertex after the movement, Q k It is related to the original vertex v k The relevant quadratic error matrix, W k It is a weighting factor;

[0028] Texture and Material Processing: Texture and material data optimization, including texture downsampling, color quantization, and material merging operations, using a bilinear interpolation texture downsampling algorithm.

[0029] The color value of the target pixel is calculated by selecting four adjacent pixels from the original texture image and taking a weighted average based on the color values ​​of the four pixels and their relative distance to the target pixel.

[0030]

[0031] Where G is the color value of the target pixel, G 01 G 02 G 03 G 04 is the color value of four adjacent pixels in the original texture image, and a and b are the relative positions of the target pixel in the original texture image;

[0032] Generate SVF format: Through the above optimizations, the data is reorganized into Streaming View Format, a 3D data format optimized for web rendering;

[0033] The optimized SVF format model is provided to users for viewing and interaction on the platform through the View & Data API;

[0034] WebGL rendering: Rendering algorithm optimization using Level of Detail (LOD) technology.

[0035]

[0036] Where M is the total number of LOD levels, a is the scaling factor (usually greater than 1), and B... max It is the maximum viewing distance (at which the model with the highest level of detail is used), B observer It is the distance from the observer to the model;

[0037] Implement lightweight display and interactive operations of the model on the web.

[0038] Furthermore, in step 11, test questions with scores of 5, 10, 20, and 30 points are set respectively. The specific scoring method is as follows: the score is assigned according to the number of keywords, c = {5 (d < 7.5), 10 (7.5 ≤ d < 15), 20 (15 ≤ d < 25), 30 (25 ≤ d < 35)}, where c is the score of each question and d is the number of keywords.

[0039] Furthermore, in step 12, there are two methods for creating the paper: random paper creation and manual paper creation.

[0040] Randomized test questions: Automatically selected from a database of x BIM scene models. A test paper is composed of y scene models, and the total score of y scene models is A;

[0041] y = d1 + d2 + d3 + d4

[0042] A = 5*d1 + 10*d2 + 20*d3 + 30*d4

[0043] in:

[0044] x1 represents the number of questions in the question bank with a score of 5; d1 represents the number of questions extracted from the question bank with a score of 5.

[0045] x2 represents the number of questions in the question bank with a score of 10; d2 represents the number of questions extracted from the question bank with a score of 10.

[0046] x3 represents the number of questions in the question bank with a score of 20; d3 represents the number of questions extracted from the question bank with a score of 20.

[0047] x4 represents the number of questions in the question bank with a score of 30; d4 represents the number of questions extracted from the question bank with a score of 30.

[0048] Manual test paper generation: Based on the specific profession or special personnel being assessed, select y models from the same professional direction to form a test paper.

[0049] The present invention has the following beneficial effects:

[0050] (1) Three-dimensional interactive assessment based on BIM scene model. Using the quality model of a real project scene as the test questions, trainees interact with the scene model in real time, which helps to stimulate learning interest, improve learning initiative and participation, and avoid the problems of traditional training methods being monotonous and lacking diversity; at the same time, the training and assessment methods are not constrained by venue, time and equipment performance, and the time can be freely chosen.

[0051] (2) By statistically analyzing the system's exam data, we can identify the quality problems that students find difficult to identify in real project quality inspections, or the reasons and solutions for quality problems that they do not clearly understand. This will help strengthen their learning of targeted quality problems and the selection of the focus of actual project quality inspections.

[0052] (3) This invention adopts the HTML protocol and adopts the online learning, training and assessment method on the web page, which is not restricted by computer equipment and configuration requirements; This invention uses the architectural scene model of real projects, and identifies quality problems in the architectural scene by means of "finding problems" in the immersive visualization form, and analyzes the causes of the problems, and proposes specific rectification and prevention measures; This invention can continuously add and supplement the real project scene model according to the actual project, and expand the database of the system. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating the construction process of the virtual scene construction quality training and assessment system based on BIM technology as described in this invention.

[0054] Figure 2 This is a flowchart illustrating the operation of the BIM-based virtual scene construction quality training and assessment system described in this invention. Detailed Implementation

[0055] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the scope of protection of the present invention is not limited thereto.

[0056] Reference Figure 1 , 2The present invention discloses a virtual scene construction quality training and assessment system based on BIM technology, comprising a training module and an assessment module. The system construction method is generally divided into two parts according to different modules, specifically including the following processes:

[0057] Step 1: Knowledge base classification;

[0058] The training database is classified into secondary categories based on the division of sub-projects and sub-items in the "Unified Standard for Acceptance of Construction Quality of Building Engineering GB50300-2013", as follows:

[0059] Table 1. Knowledge Base Classification Table

[0060]

[0061]

[0062] Step 2: Categorize quality issues;

[0063] Based on the construction quality problems encountered during the project construction process, they are categorized into each sub-project. For each quality problem, a description of the construction phenomenon, the cause analysis, the quality remedial measures, and the methods and measures to avoid such problems in subsequent construction processes are prepared.

[0064] Step 3: Establish a BIM learning model for the sub-project;

[0065] Establish BIM scene models for each sub-project. In each sub-project BIM scene model, set the different quality problem manifestations in step 2 to form ten types of sub-project BIM scene models that summarize the various quality problems.

[0066] Step 4: Upload the BIM learning model of each sub-project to the existing system;

[0067] On the system management side, the local BIM scene model is uploaded through the model management module. After the BIM scene model is lightweighted, it is bound to ten types of sub-projects.

[0068] The lightweight processing method for BIM models is as follows:

[0069] ① Upload and parsing

[0070] The created RVT format BIM model is uploaded to the cloud platform, which parses the file and extracts the model's set, material, and texture information.

[0071] ②Geometric optimization

[0072] The MODLE API optimizes the geometric data of RVT format models, including vertex reduction, edge simplification, and patch merging, to reduce the complexity and data volume of RVT format models. The specific algorithm is as follows:

[0073]

[0074] Where E is the total error, v k It is the original vertex, v. k It is the vertex after the movement, Q k It is related to the original vertex v k The relevant quadratic error matrix, W k It is a weighting factor.

[0075] ③Texture and Material Processing

[0076] Texture and material data optimization includes operations such as texture downsampling, color quantization, and material merging to reduce the amount of texture and material data while maintaining visual acceptability. A bilinear interpolation texture downsampling algorithm is used.

[0077] The color value of the target pixel is calculated by selecting four adjacent pixels from the original texture image and taking a weighted average based on the color values ​​of the four pixels and their relative distance to the target pixel.

[0078]

[0079] Where G is the color value of the target pixel, G 01 G 02 G 03 G 04 is the color value of four adjacent pixels in the original texture image, and a and b are the relative positions of the target pixel in the original texture image.

[0080] ④ Generate SVF format

[0081] Step ③ optimizes the data by reorganizing it into a Streaming View Format, a 3D data format optimized for web rendering.

[0082] ⑤ The optimized SVF format model is provided to users for viewing and interaction on the platform through the View & Data API.

[0083] ⑥ WEBGL rendering

[0084] The View&Data API is based on WebGL technology, enabling efficient rendering of 3D models in the browser. WebGL is a JavaScript API used in HTML5. <canvas>Elemental rendering of high-performance 2D and 3D graphics.

[0085] It uses the level of detail (LOD) technique to optimize the rendering algorithm:

[0086]

[0087] Where M is the total number of LOD levels, a is the scaling factor (usually greater than 1), B max is the maximum viewing distance (at which the highest level of detail model is used), B observer is the distance from the observer to the model.

[0088] ⑦Realize the lightweight display and interactive operation of BIM scene model on the web side.

[0089] Step 5: Under the quality problem of the sub-contract engineering classification, bind the corresponding quality problem component in the sub-contract engineering BIM scene model and the corresponding viewport.

[0090] Step 6: Click on a certain sub-contract engineering on the student side to display the BIM scene model of the sub-contract engineering, click on a certain quality problem to directly point to the corresponding problem area and display it in the display viewport, so that students can learn about specific quality problems, possible parts and problem phenomena in actual projects, and display the causes and prevention, rectification methods and measures for the quality problem.

[0091] Step 7: According to multiple actual engineering projects, establish n BIM scene models of various stages and different areas in the construction process; for example, establish a certain floor construction scene model of the main structure, a 89 house type scene model of a certain residential area, and a certain basement refrigeration room scene model.

[0092] Step 8: According to the quality problem classification, establish the wrong version model of the quality problem area or problem point that may occur in different BIM construction scenes; for example, in the "main structure of a certain floor" construction scene, set up "concrete structure honeycomb pockmarks" on the structure wall, "no anti-kang in bathroom" in the bathroom area, and "insufficient pouring at the top of the secondary structure wall" in the secondary structure wall construction.

[0093] Step 9: In the management end, upload the local BIM construction scene model through the model management module, and after lightening, belong to the "assessment model" classification; the specific implementation method is consistent with the above training module.

[0094] Step 10: After uploading all BIM construction scene models, set each BIM construction scene as an assessment question; set n assessment options, of which m are quality problem options, and fix the BIM scene model components and display viewports for n assessment options.

[0095] Step 11: According to the number m of quality problems contained in each BIM construction scene, different examination scores are configured, and specific answer keywords are set to score, and examination questions with scores of 5, 10, 20 and 30 are set, and the specific scoring method is as follows: scoring according to the number of keywords, c={5(d<7.5), 10(7.5≤d<15), 20(15≤d<25), 30(25≤d<35)}, c is the score of each question, and d is the number of keywords.

[0096] Step 12: Create an examination paper, select the examined personnel, examination opening time, examination duration and other factors, and set the total score A according to each examination; there are two ways to set the paper, as follows:

[0097] The first kind: random paper setting, automatically extracting y scene models from x BIM scene model examination library to form a paper, and the total score of y scene models is A;

[0098] y=d1+d2+d3+d4

[0099] A=5*d1+10*d2+20*d3+30*d4

[0100] Among them:

[0101] x1 is the number of "5-point" question bank questions; d1 is the number of "5-point" questions in the question bank;

[0102] x2 is the number of "10-point" question bank questions; d2 is the number of "10-point" questions in the question bank;

[0103] x3 is the number of "20-point" question bank questions; d3 is the number of "20-point" questions in the question bank;

[0104] x4 is the number of "30-point" question bank questions; d4 is the number of "30-point" questions in the question bank.

[0105] The second kind: manual paper setting, according to the specific professional or special personnel of the examination, select y models of the same professional direction to form a paper; realize the difference and efficiency of each examination paper.

[0106] Step 13: The relevant personnel click start in the examination module and enter the examination, a total of y questions, enter the construction building BIM scene model respectively, find the possible quality problems through immersive interaction, and mark and record the causes of the problems and rectification and correction measures. The present application simulates the actual engineering site inspection, identifies the quality problems in the quality scene model in the virtual simulation of the real project scene. ​

[0107] Step 14: After the end of the examination, the analysis of this examination is carried out, and the comparison between the answers and the reference answers of each question is carried out to assist the students to understand the deficiencies of the examination in time.

[0108] Step 15: According to the actual engineering project, the examination module question bank is continuously established and supplemented.

[0109] The embodiments are preferred embodiments of the present application, but the present application is not limited to the above embodiments, and any obvious improvement, replacement or modification made by those skilled in the art without departing from the essential content of the present application shall belong to the protection scope of the present application.< / canvas>

Claims

1. A method for building a virtual scene construction quality training and assessment system based on BIM technology, characterized in that, The process includes the following: Step 1: Categorize the knowledge base for the training modules; Step 2: Categorize construction quality issues; Step 3: Establish a BIM learning model for the sub-project; Step 4: Upload the BIM learning model of each sub-project to the existing system, perform lightweight model processing, and then bind it to the ten sub-project categories obtained in Step 1. Step 5: Bind each quality issue to the corresponding issue component and issue viewport in the BIM learning model of the sub-project; Step 6: Relevant personnel can directly click on a specific sub-project on the trainee's end to display the BIM scene model of that sub-project. Clicking on a quality issue will directly zoom in on the corresponding issue area and display it in the viewport for training and learning. Step 7: Based on multiple actual engineering projects, establish multiple BIM scene models for different stages and areas during the construction process; Step 8: Based on the classification of quality issues, create misprinted models of potential quality problem areas or points in different BIM construction scenarios; Step 9: On the management side, upload the local BIM construction scene model through the model management module. After being lightweighted, it will be classified into the "Assessment Model" category. Step 10: Set each BIM construction scene as an assessment question; set n assessment options, including m quality issue options, and fix the BIM scene model components and display viewports for each of the n assessment options; Step 11: Based on the number m of quality issues in each BIM construction scenario, configure different assessment scores and set specific answer keywords for scoring; Step 12: Create an assessment paper, select the personnel to be assessed, the assessment opening time, the assessment duration, and other factors, and set the total score for each assessment; Step 13: Relevant personnel click "Start" in the assessment module to enter the exam. They will then enter different construction building BIM scene models and, through immersive interactive browsing, identify potential quality issues, mark them, and record in detail the causes of the issues and the rectification and corrective measures. Step 14: After the exam, analyze the exam results, comparing your answers to the reference answers for each question to help you identify areas for improvement.

2. The method for building a virtual scene construction quality training and assessment system based on BIM technology according to claim 1, characterized in that, Step 2 specifically includes: classifying the construction quality problems encountered during the project construction process into each sub-project, and for each quality problem, compiling the construction phenomenon of the quality problem, the cause analysis, the quality remedial measures for the problem, and the methods and measures to avoid such problems in subsequent construction processes.

3. The method for building a virtual scene construction quality training and assessment system based on BIM technology according to claim 1, characterized in that, In step 3, BIM learning models for each sub-project are established. In each sub-project BIM learning model, different quality problem manifestations from step 2 are set, forming ten types of sub-project BIM learning models that summarize the various quality problems.

4. The method for building a virtual scene construction quality training and assessment system based on BIM technology according to claim 1, characterized in that, The specific methods for model lightweighting in steps 4 and 9 are as follows: Upload and parsing: Upload the created RVT format model to the cloud platform. The cloud platform parses the file and extracts the model's collection, material, and texture information. Geometric optimization: Optimize the geometric data of the RVT format model, including vertex reduction, edge simplification, and patch merging. The specific algorithm formula is as follows: Where E is the total error, v k It is the original vertex, v. k It is the vertex after the movement, Q k It is related to the original vertex v k The relevant quadratic error matrix, W k It is a weighting factor; Texture and Material Processing: Texture and material data optimization, including texture downsampling, color quantization, and material merging operations, using a bilinear interpolation texture downsampling algorithm. The color value of the target pixel is calculated by selecting four adjacent pixels from the original texture image and taking a weighted average based on the color values ​​of the four pixels and their relative distance to the target pixel. Where G is the color value of the target pixel, G 01 G 02 G 03 G 04 is the color value of four adjacent pixels in the original texture image, and a and b are the relative positions of the target pixel in the original texture image; Generate SVF format: Through the above optimizations, the data is reorganized into Streaming View Format, a 3D data format optimized for web rendering; The optimized SVF format model is provided to users for viewing and interaction on the platform through the View & Data API; WebGL rendering: Rendering algorithm optimization using Level of Detail (LOD) technology. Where M is the total number of LOD levels, a is the scaling factor (usually greater than 1), and B... max It is the maximum viewing distance (at which the model with the highest level of detail is used), B observer It is the distance from the observer to the model; Implement lightweight display and interactive operations of the model on the web.

5. The method for building a virtual scene construction quality training and assessment system based on BIM technology according to claim 1, characterized in that, In step 11, questions worth 5, 10, 20, and 30 points are set respectively. The specific scoring method is as follows: the score is assigned according to the number of keywords, c = {5 (d < 7.5), 10 (7.5 ≤ d < 15), 20 (15 ≤ d < 25), 30 (25 ≤ d < 35)}, where c is the score of each question and d is the number of keywords.

6. The method for building a virtual scene construction quality training and assessment system based on BIM technology according to claim 1, characterized in that, In step 12, there are two methods for creating the volumes: random volume creation and manual volume creation. Randomized test questions: Automatically selected from a database of x BIM scene models. A test paper is composed of y scene models, and the total score of y scene models is A; y = d1 + d2 + d3 + d4 A = 5*d1 + 10*d2 + 20*d3 + 30*d4 in: x1 represents the number of questions in the question bank with a score of 5; d1 represents the number of questions extracted from the question bank with a score of 5. x2 represents the number of questions in the question bank with a score of 10; d2 represents the number of questions extracted from the question bank with a score of 10. x3 represents the number of questions in the question bank with a score of 20; d3 represents the number of questions extracted from the question bank with a score of 20. x4 represents the number of questions in the question bank with a score of 30; d4 represents the number of questions extracted from the question bank with a score of 30. Manual test paper generation: Based on the specific profession or special personnel being assessed, select y models from the same professional direction to form a test paper.

Images