A bayesian-based analysis method and device for BIM model quality fineness

Abstract

The application relates to the technical field of BIM model, and discloses a BIM model quality fineness analysis method and device based on Bayes, aiming to solve the problem of poor accuracy of the existing parameter analysis method, and mainly comprising the following steps: classifying the quality fineness of a BIM model, and determining the parameter factors corresponding to various quality fineness; determining the prior probability of each parameter factor and each quality fineness, determining the posterior probability of each parameter factor and each quality fineness according to the prior probability and based on a Bayes network; and analyzing the influence degree of each parameter factor and each quality fineness on the BIM model quality fineness according to the posterior probability. The application avoids the influence of subjectivity and one-sidedness on analysis, improves the accuracy of parameter analysis of the BIM model, and provides a theoretical basis for subsequent optimization of the BIM model.

Description

Technical Field

[0001] This invention relates to the field of BIM modeling technology, and specifically to a Bayesian-based method and apparatus for analyzing the quality and refinement of BIM models. Background Technology

[0002] With the continuous development and application of Building Information Modeling (BIM) technology, BIM models have become the primary working medium in the construction industry. The core of BIM is to create a virtual 3D model of the building project and, using digital technology, provide this model with a complete and accurate database of building information. BIM models not only contain the geometric and attribute information of the building but also its non-geometric information, such as material and equipment information. To ensure the quality and accuracy of the BIM model, each parameter needs to be analyzed and optimized. During this analysis and optimization process, it is necessary to determine the degree of influence of each parameter on the quality and accuracy of the BIM model in order to make targeted adjustments and optimizations.

[0003] In existing technologies, the analysis of parameters and the determination of their influence typically employ statistical methods and expert experience. However, these methods are inherently subjective and biased in determining the degree of parameter influence. Furthermore, the analysis of the interrelationships between multiple parameters often utilizes methods such as logistic regression, but these methods frequently overlook the fuzziness and uncertainty between parameters, leading to poor accuracy in the analytical results. Summary of the Invention

[0004] This invention aims to address the problem of poor accuracy in existing parameter analysis methods by proposing a Bayesian-based method and apparatus for analyzing the quality and precision of BIM models.

[0005] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0006] Firstly, a Bayesian-based method for analyzing the quality granularity of BIM models is provided, the method comprising:

[0007] The quality level of BIM models is classified, and the parameter factors corresponding to each quality level are determined.

[0008] Determine the prior probabilities of each parameter factor and each type of quality granularity, and determine the posterior probabilities of each parameter factor and each type of quality granularity based on the prior probabilities and a Bayesian network.

[0009] The impact of each parameter factor and various quality refinements on the quality refinement of the BIM model is analyzed based on the posterior probability analysis.

[0010] Furthermore, determining the prior probabilities of each parameter factor specifically includes:

[0011] Based on empirical or statistical methods, the influence of parameter factors on historical BIM models is determined. After quantifying the influence of parameter factors on historical BIM models, the prior probabilities of the corresponding parameter factors are obtained. The probability that the historical BIM model meets the model accuracy requirements in various quality granularity evaluations is used as the prior probability of the corresponding quality granularity category.

[0012] Furthermore, based on the prior probabilities and using a Bayesian network, the posterior probabilities of each parameter factor and various quality levels are determined, specifically including:

[0013] A Bayesian network is pre-constructed between each parameter factor and the corresponding quality granularity of the category. The prior probabilities of each parameter factor and each type of quality granularity are input into the corresponding Bayesian network to obtain the posterior probabilities of each parameter factor and each type of quality granularity.

[0014] Furthermore, the categories of BIM model quality precision include one or more of the following: geometric precision, physical precision, functional precision, temporal precision, and analytical precision.

[0015] Furthermore, the parameter factors corresponding to the geometric accuracy include one or more of geometry, size, and proportion; the parameter factors corresponding to the physical accuracy include one or more of material, weight, density, and thickness; the parameter factors corresponding to the functional accuracy include one or more of pipes, connections, and valves; the parameter factors corresponding to the time accuracy include one or more of date, time, and schedule; and the parameter factors corresponding to the analytical accuracy include one or more of structural analysis, thermal analysis, and solar radiation analysis.

[0016] Secondly, a Bayesian-based BIM model quality refinement analysis device is provided, the device comprising:

[0017] The classification unit is used to classify the quality level of the BIM model and determine the parameter factors corresponding to each quality level.

[0018] A determination unit is used to determine the prior probabilities of each parameter factor and various quality granularities, and to determine the posterior probabilities of each parameter factor and various quality granularities based on the prior probabilities and a Bayesian network.

[0019] The analysis unit is used to analyze the influence of each parameter factor and various quality refinements on the quality refinement of the BIM model based on the posterior probability.

[0020] Furthermore, the determining unit is specifically used for:

[0021] Based on empirical or statistical methods, the influence of parameter factors on historical BIM models is determined. After quantifying the influence of parameter factors on historical BIM models, the prior probabilities of the corresponding parameter factors are obtained. The probability that the historical BIM model meets the model accuracy requirements in various quality granularity evaluations is used as the prior probability of the corresponding quality granularity category.

[0022] Furthermore, the determining unit is specifically used for:

[0023] A Bayesian network is pre-constructed between each parameter factor and the corresponding quality granularity of the category. The prior probabilities of each parameter factor and each type of quality granularity are input into the corresponding Bayesian network to obtain the posterior probabilities of each parameter factor and each type of quality granularity.

[0024] Furthermore, the categories of BIM model quality precision include one or more of the following: geometric precision, physical precision, functional precision, temporal precision, and analytical precision.

[0025] Furthermore, the parameter factors corresponding to the geometric accuracy include one or more of geometry, size, and proportion; the parameter factors corresponding to the physical accuracy include one or more of material, weight, density, and thickness; the parameter factors corresponding to the functional accuracy include one or more of pipes, connections, and valves; the parameter factors corresponding to the time accuracy include one or more of date, time, and schedule; and the parameter factors corresponding to the analytical accuracy include one or more of structural analysis, thermal analysis, and solar radiation analysis.

[0026] The beneficial effects of this invention are as follows: The Bayesian-based BIM model quality refinement analysis method and apparatus of this invention divides the quality refinement of the BIM model into multiple categories. For each category of quality refinement, a Bayesian network is used to obtain the corresponding parameter factors and the posterior probability of the corresponding category of quality refinement. The posterior probability is then used as the influence weight of each parameter factor and the corresponding category of quality refinement. This allows for the analysis of the influence of each parameter factor and each category of quality refinement on the quality refinement of the BIM model based on the influence weight. This invention analyzes the influence of parameter factors on quality refinement through posterior probability analysis, avoiding the influence of subjectivity and bias on the analysis, improving the accuracy of parameter analysis of the BIM model, and providing a theoretical basis for subsequent BIM model optimization. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating the Bayesian-based BIM model quality refinement analysis method described in this embodiment of the invention.

[0028] Figure 2 This is a structural diagram of various quality levels and corresponding parameter factors of the BIM model described in the embodiments of the present invention.

[0029] Figure 3 This is a schematic diagram of the Bayesian network structure corresponding to the geometric precision described in the embodiments of the present invention;

[0030] Figure 4 This is a schematic diagram of the structure of the Bayesian-based BIM model quality refinement analysis device described in an embodiment of the present invention. Detailed Implementation

[0031] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0032] This invention aims to improve the accuracy of parameter analysis of BIM models and proposes a Bayesian-based method and apparatus for analyzing the quality granularity of BIM models. The main technical solutions include: classifying the quality granularity of BIM models and determining the parameter factors corresponding to each type of quality granularity; determining the prior probabilities of each parameter factor and each type of quality granularity; determining the posterior probabilities of each parameter factor and each type of quality granularity based on the prior probabilities and a Bayesian network; and analyzing the influence of each parameter factor and each type of quality granularity on the quality granularity of the BIM model based on the posterior probabilities.

[0033] Specifically, the analysis of BIM model quality granularity mainly involves determining the degree of influence of various parameters on the BIM model's quality granularity. To achieve this, existing technologies directly determine the influence weights of each parameter based on expert experience or statistical methods. This approach is highly subjective and biased, resulting in low accuracy. Therefore, this invention categorizes BIM model quality granularity into multiple classes. For each class of quality granularity, a Bayesian network is used to obtain the posterior probabilities of the corresponding parameter factors and quality granularity. These posterior probabilities are then used as the influence weights of each parameter factor and quality granularity. This allows for the analysis of the influence of corresponding parameter factors and various quality granularity levels on the BIM model's quality granularity based on these influence weights, thereby avoiding the influence of subjectivity and bias on parameter analysis.

[0034] Example

[0035] Please see Figure 1 The Bayesian-based BIM model quality refinement analysis method described in this embodiment of the invention includes the following steps:

[0036] Step 1: Classify the quality level of the BIM model and determine the parameter factors corresponding to each quality level.

[0037] Please see Figure 2In this embodiment, the quality precision of the BIM model is divided into the following categories: geometric precision, physical precision, functional precision, temporal precision, analytical precision, etc., and the corresponding parameter factors for each category of quality precision are determined.

[0038] Geometric accuracy is related to the basic parameter factors of the BIM model, such as geometry, size, and scale, which reflect the correctness and accuracy of the BIM.

[0039] Physical accuracy is related to the physical parameters of the BIM model, such as material, weight, density, and thickness, which reflect the BIM model's performance in reality.

[0040] Functional accuracy is related to the functional parameter factors of the BIM model, such as pipes, connections, and valves, which reflect the performance and behavior of the BIM model in actual use.

[0041] Time accuracy is related to the time parameter factors of the BIM model, such as date, time, and schedule, which reflect the state and changes of the BIM model at different points in time.

[0042] The accuracy of the analysis is related to the analysis parameters of the BIM model, such as structural analysis, thermal analysis, and solar radiation analysis, which reflect the more in-depth performance and characteristics of the BIM model.

[0043] Step 2: Determine the prior probabilities of each parameter factor and each type of quality granularity, and determine the posterior probabilities of each parameter factor and each type of quality granularity based on the prior probabilities and a Bayesian network.

[0044] In this embodiment, the specific method for determining the prior probability of each parameter factor includes the following steps:

[0045] Based on empirical or statistical methods, the influence of parameter factors on historical BIM models is determined. After quantifying the influence of parameter factors on historical BIM models, the prior probabilities of the corresponding parameter factors are obtained. The probability that the historical BIM model meets the model accuracy requirements in various quality granularity evaluations is used as the prior probability of the corresponding quality granularity category.

[0046] Specifically, prior probabilities can be determined using expert experience or statistical methods. For example, to determine the prior probability of a parameter factor like geometry, one can use empirical or statistical methods to determine the degree of influence of geometry on several previous BIM models, then quantify the degree of influence to obtain the prior probability corresponding to the geometry. Similarly, to determine the prior probability of a quality level like geometric accuracy, one can use empirical or statistical methods to determine the probability that historical BIM models meet the model accuracy requirements in geometric accuracy evaluations, and use this probability as the prior probability of geometric accuracy.

[0047] In this embodiment, a corresponding Bayesian network model needs to be pre-established for each quality level of refinement. A Bayesian network is a probabilistic graphical model. A Bayesian network is a directed acyclic graph (DAG) consisting of nodes representing variables and directed edges connecting these nodes. Nodes represent random variables, and directed edges between nodes represent the relationships between nodes (from a parent node to its child node). The strength of the relationship is expressed using conditional probability, and information without a parent node is expressed using prior probability.

[0048] In this embodiment, a Bayesian network can be built using Netica software. Specifically, for each quality granularity and its corresponding parameter factor, each parameter factor is treated as a node, and the quality granularity of the corresponding category is treated as a node. Then, nodes with obvious causal or correlational relationships are connected, and the connection starts from the cause node and points to the result node. The relationship strength is expressed through conditional probability, thus completing the construction of the Bayesian network. Figure 3 A schematic diagram of a Bayesian network corresponding to geometric precision is shown.

[0049] In this embodiment, after constructing the Bayesian network and determining the prior probabilities of each parameter factor and each type of quality granularity, for each type of quality granularity, the corresponding parameter factor and the prior probability of the quality granularity are input into the corresponding Bayesian network to obtain the posterior probability of each parameter factor and that type of quality granularity.

[0050] Step 3: Analyze the impact of each parameter factor and each type of quality refinement on the quality refinement of the corresponding category of the BIM model based on the posterior probability analysis.

[0051] After obtaining the posterior probabilities of the corresponding parameter factors and the corresponding categories of quality refinement, the posterior probabilities are used as the influence weights of the parameter factors and the corresponding categories of quality refinement. The influence of each parameter factor and each category of quality refinement on the quality refinement of the BIM model can be analyzed based on the influence weights. That is, the greater the influence weight, the greater the influence of the corresponding parameter factor or the corresponding category of quality refinement on the quality refinement of the BIM model, and the smaller the influence weight, the smaller the influence of the corresponding parameter factor or the corresponding category of quality refinement on the quality refinement of the BIM model.

[0052] In summary, the Bayesian-based BIM model quality refinement analysis method described in this embodiment categorizes the quality refinement of the BIM model into multiple classes. For each class of quality refinement, a Bayesian network is used to obtain the posterior probabilities of the corresponding parameter factors and the quality refinement of the corresponding class. These posterior probabilities are then used as the influence weights of each parameter factor and the corresponding quality refinement class. This allows for the analysis of the degree of influence of each parameter factor and each type of quality refinement on the quality refinement of the BIM model based on these influence weights. This invention analyzes the influence of parameter factors on quality refinement through posterior probability analysis, avoiding the influence of subjectivity and bias on the analysis, improving the accuracy of parameter analysis of the BIM model, and providing a theoretical basis for subsequent BIM model optimization.

[0053] Based on the above technical solution, this embodiment also proposes a Bayesian-based BIM model quality refinement analysis device. Please refer to [link to relevant documentation]. Figure 4 The device includes:

[0054] The classification unit is used to classify the quality level of the BIM model and determine the parameter factors corresponding to each quality level.

[0055] A determination unit is used to determine the prior probabilities of each parameter factor and various quality granularities, and to determine the posterior probabilities of each parameter factor and various quality granularities based on the prior probabilities and a Bayesian network.

[0056] The analysis unit is used to analyze the influence of each parameter factor and various quality refinements on the quality refinement of the BIM model based on the posterior probability.

[0057] It is understood that since the Bayesian-based BIM model quality refinement analysis device described in the embodiments of the present invention is a device for implementing the Bayesian-based BIM model quality refinement analysis method described in the embodiments, the device disclosed in the embodiments is relatively simple to describe because it corresponds to the method disclosed in the embodiments. For relevant parts, please refer to the description of the method.

[0058] The Bayesian-based BIM model quality granularity analysis method and apparatus described in this embodiment allows for the selection of different BIM model quality granularities during BIM model implementation, based on varying application areas and project requirements. By analyzing the granularity of BIM model quality, the ideal application effect of certain BIM software in specific fields can be determined, leading to the search for more suitable BIM software or the upgrading and optimization of existing BIM software. Simultaneously, a refined BIM model enables better project quality management and cost control, thereby improving overall project efficiency. Furthermore, BIM model granularity analysis facilitates better communication and collaboration with project teams, improving project management efficiency by focusing on the key aspects of BIM model granularity. This allows team members to understand the impact of different granularity BIM models on the project. This method can also be extended to factor analysis in various stages of BIM model application, demonstrating promising development prospects.

Claims

1. A Bayesian-based method for analyzing the quality refinement of a BIM model, characterized in that, The method includes: The quality level of BIM models is classified, and the parameter factors corresponding to each quality level are determined. Determine the prior probabilities of each parameter factor and each type of quality granularity, and determine the posterior probabilities of each parameter factor and each type of quality granularity based on the prior probabilities and a Bayesian network. The impact of each parameter factor and various quality levels on the quality level of the BIM model is analyzed based on the posterior probability analysis. The determination of the prior probabilities of each parameter factor and various quality granularities specifically includes: Based on empirical or statistical methods, the influence of parameter factors on historical BIM models is determined. After quantifying the influence of parameter factors on historical BIM models, the prior probabilities of the corresponding parameter factors are obtained. The probability that the historical BIM model meets the model accuracy requirements in various quality granularity evaluations is used as the prior probability of the corresponding quality granularity category.

2. The Bayesian-based analysis method of BIM model quality refinement of claim 1, wherein, Based on the prior probabilities and using a Bayesian network, the posterior probabilities of each parameter factor and various quality levels are determined, specifically including: A Bayesian network is pre-constructed between each parameter factor and the corresponding quality granularity of the category. The prior probabilities of each parameter factor and each type of quality granularity are input into the corresponding Bayesian network to obtain the posterior probabilities of each parameter factor and each type of quality granularity.

3. The Bayesian-based analysis method of BIM model quality refinement of claim 1 or 2, wherein, The categories of BIM model quality precision include one or more of the following: geometric precision, physical precision, functional precision, temporal precision, and analytical precision.

4. The Bayesian-based analysis method of BIM model quality refinement of claim 3, wherein, The parameter factors corresponding to the geometric accuracy include one or more of geometry, size, and proportion; the parameter factors corresponding to the physical accuracy include one or more of material, weight, density, and thickness; the parameter factors corresponding to the functional accuracy include one or more of pipes, connections, and valves; the parameter factors corresponding to the time accuracy include one or more of date, time, and schedule; and the parameter factors corresponding to the analytical accuracy include one or more of structural analysis, thermal analysis, and solar radiation analysis.

5. A Bayesian-based analysis apparatus for BIM model quality refinement, characterized in that, The device includes: The classification unit is used to classify the quality level of the BIM model and determine the parameter factors corresponding to each quality level. A determination unit is used to determine the prior probabilities of each parameter factor and various quality granularities, and to determine the posterior probabilities of each parameter factor and various quality granularities based on the prior probabilities and a Bayesian network. The analysis unit is used to analyze the influence of each parameter factor and various quality refinements on the quality refinement of the BIM model based on the posterior probability. The determining unit is specifically used for: Based on empirical or statistical methods, the influence of parameter factors on historical BIM models is determined. After quantifying the influence of parameter factors on historical BIM models, the prior probabilities of the corresponding parameter factors are obtained. The probability that the historical BIM model meets the model accuracy requirements in various quality granularity evaluations is used as the prior probability of the corresponding quality granularity category.

6. The Bayesian-based analysis apparatus of BIM model quality refinement of claim 5, wherein, The determining unit is specifically used for: A Bayesian network between each parameter factor and a corresponding quality fineness of the category is constructed in advance, prior probabilities of each parameter factor and each quality fineness of the category are input into the corresponding Bayesian network, and posterior probabilities of each parameter factor and each quality fineness of the category are obtained.

7. The Bayesian-based analysis apparatus of BIM model quality refinement of claim 5 or 6, wherein, The categories of the BIM model quality fineness include one or more of geometric accuracy, physical accuracy, functional accuracy, time accuracy and analysis accuracy.

8. The Bayesian-based analysis apparatus of BIM model quality refinement of claim 7, wherein, The parameter factors corresponding to the geometric accuracy include one or more of geometric shape, size and proportion, the parameter factors corresponding to the physical accuracy include one or more of material, weight, density and thickness, the parameter factors corresponding to the functional accuracy include one or more of pipe, connection and valve, the parameter factors corresponding to the time accuracy include one or more of date, time and progress, and the parameter factors corresponding to the analysis accuracy include one or more of structure analysis, thermal analysis and sunshine analysis.

Images