CN117455416B

CN117455416B - Data analysis method and system applied to building project management

Info

Publication number: CN117455416B
Application number: CN202311757322.6A
Authority: CN
Inventors: 陈金湖; 陈继丹; 陈魁; 丁峰; 狄德龙; 李成山; 李明; 陈正华
Original assignee: Guangzhou Nanhua Engineering Management Co ltd
Current assignee: Guangzhou Nanhua Engineering Management Co ltd
Priority date: 2023-12-20
Filing date: 2023-12-20
Publication date: 2024-03-22
Anticipated expiration: 2043-12-20
Also published as: CN117455416A

Abstract

According to the data analysis method and system applied to the building project management, the BIM model can be automatically generated through the drawing data, and the existing drawing data does not need to be re-built, so that additional work is omitted, the working efficiency is improved, and the working cost is reduced. The BIM model construction network is debugged through drawing data templates of multiple types of modes and corresponding BIM models, so that the BIM model construction network can acquire the capacity of constructing the corresponding BIM model based on drawing data of each mode, the generalization capacity of the BIM model construction network is improved, and the corresponding BIM model can be constructed based on drawing data of at least one mode through the BIM model construction network, so that the generalization capacity of constructing the BIM model is improved.

Description

Data analysis method and system applied to building project management

Technical Field

The application relates to the field of building project management and artificial intelligence, in particular to a data analysis method and system applied to building project management.

Background

Along with the advancement of the digitization process in the construction industry, BIM is an integral part of construction project management. BIM has the characteristics of fine geometric structure, abundant semantic information and abundant parameter characteristics, and can more intuitively and more accurately express the geometric, physical, additional attribute, manufacturing process and other information of railway projects. However, at present, the application of BIM has not completely replaced CAD, and most of the mainstream ways of design enterprises are still mainly two-dimensional planes, and reverse design from CAD drawing data to BIM becomes urgent. In a general manner, CAD information is transferred to the BIM for re-modeling by manpower, which greatly wastes work output and affects project progress.

Disclosure of Invention

The application provides a data analysis method and system applied to building project management.

According to an aspect of the present application, there is provided a data analysis method applied to construction project management, the method comprising:

acquiring a plurality of first network training templates, wherein the first network training templates comprise drawing data templates and BIM model templates corresponding to the drawing data templates, and the plurality of first network training templates comprise drawing data templates of multiple types of modes;

debugging a BIM model construction network according to a plurality of first network training templates, wherein the BIM model construction network is used for constructing a corresponding BIM model based on drawing data of at least one mode in the multi-class modes;

building a network through the BIM model, and building a BIM model corresponding to the first drawing data based on the acquired first drawing data;

wherein, according to a plurality of the first network training templates, debugging the BIM model to construct a network comprises:

determining a template characterization vector corresponding to the BIM model template according to an alternative generation component library and the BIM model template, wherein the alternative generation component library comprises a plurality of alternative generation components, and the template characterization vector characterizes the probability that each alternative generation component belongs to the BIM model template;

Processing the drawing data template through the BIM model construction network to obtain an inference characterization vector, wherein the inference characterization vector characterizes the probability that each alternative generating component belongs to an inference BIM model corresponding to the drawing data template;

and debugging the BIM model to construct a network according to the template characterization vector and the reasoning characterization vector.

In an optional embodiment, the BIM model building network includes a token vector extraction operator, a feature embedding mapping operator, and a mapping analysis operator, and the processing the drawing data template by the BIM model building network to obtain an inference token vector includes:

extracting the characterization vector from the drawing data template through the characterization vector extraction operator to obtain an inference drawing characterization vector corresponding to each drawing partition in the drawing data template;

integrating, embedding and mapping the obtained multiple inference drawing characterization vectors through the feature embedding and mapping operator to obtain an inference integration characterization vector;

and analyzing according to the reasoning integration characterization vector by the mapping analysis operator to obtain the reasoning characterization vector.

In an optional embodiment, the inference characterization vector includes inference local information corresponding to a plurality of distribution spaces in the inference BIM model, the inference local information includes an inference vector element corresponding to each of the alternative generating members, and the inference vector element characterizes a probability that the alternative generating member belongs to a generating member of the corresponding distribution space in the inference BIM model;

Analyzing the mapping analysis operator according to the reasoning integration characterization vector to obtain the reasoning characterization vector, wherein the method comprises the following steps:

analyzing the reasoning integration characterization vector through the mapping analysis operator to obtain first reasoning local information, and determining an alternative generation component corresponding to the largest reasoning vector element in the first reasoning local information as a first reasoning generation component;

and repeatedly analyzing the inference integration characterization vector based on the obtained inference generating components through the mapping analysis operator to obtain second inference local information, and determining an alternative generating component corresponding to the maximum inference vector element in the second inference local information as a second inference generating component until the obtained inference generating component is a first mark generating component, wherein the first mark generating component represents the edge of the BIM model.

In an optional implementation manner, the template characterization vector includes template local information corresponding to a plurality of distribution spaces in the BIM model template, the template local information characterizes probabilities that each of the alternative generating components belongs to generating components corresponding to the distribution spaces in the BIM model template, the inference characterization vector includes inference local information corresponding to a plurality of distribution spaces in the inference BIM model, and the inference local information characterizes probabilities that each of the alternative generating components belongs to generating components corresponding to the distribution spaces in the inference BIM model;

The debugging the BIM model to construct a network according to the template characterization vector and the reasoning characterization vector comprises the following steps:

determining a sub-error corresponding to each distribution space according to the error between the reasoning local information corresponding to each distribution space and the template local information;

determining an integrated error according to the plurality of sub-errors, wherein the integrated error is positively correlated with the plurality of sub-errors;

and debugging the BIM model according to the integrated error to construct a network.

In an optional implementation manner, the building a network through the BIM model, based on the obtained first drawing data, builds a BIM model corresponding to the first drawing data, including:

identifying the first drawing data through the BIM model construction network to obtain a characterization vector, determining a plurality of target generating components belonging to the BIM model from a plurality of candidate generating components in a candidate generating component library based on the characterization vector, wherein the characterization vector characterizes the probability that each candidate generating component belongs to the BIM model corresponding to the first drawing data;

each of the target generating means is combined to obtain the BIM model.

In an optional implementation manner, the BIM model building network includes a token vector extraction operator, a feature embedding mapping operator and a mapping parsing operator, the identifying, by the BIM model building network, the first drawing data to obtain a token vector, and determining, based on the token vector, a plurality of target generating components belonging to the BIM model from a plurality of candidate generating components in a candidate generating component library, where the determining includes:

extracting the characterization vector of the first drawing data through the characterization vector extraction operator to obtain drawing characterization vectors corresponding to each drawing partition in the first drawing data;

integrating, embedding and mapping the obtained multiple drawing characterization vectors through the feature embedding and mapping operator to obtain an integrated characterization vector;

analyzing according to the integrated characterization vector through the mapping analysis operator to obtain the characterization vector, and determining a plurality of target generation components belonging to the BIM model from a plurality of candidate generation components based on the characterization vector;

the characterization vector comprises local information corresponding to a plurality of distribution spaces in the BIM model, the local information comprises vector elements corresponding to each alternative generation component, and the vector elements characterize the probability that the alternative generation components belong to the generation components of the corresponding distribution spaces in the BIM model;

Analyzing, by the mapping analysis operator, according to the integrated token vector to obtain the token vector, and determining, based on the token vector, a plurality of target generation components belonging to the BIM model from a plurality of candidate generation components, including:

analyzing the integrated characterization vector through the mapping analysis operator to obtain first local information, and determining an alternative generating component corresponding to the largest vector element in the first local information as a first target generating component;

and repeatedly analyzing the integrated characterization vector based on the obtained target generating component through the mapping analysis operator to obtain second local information, and determining an alternative generating component corresponding to the largest vector element in the second local information as a second target generating component until the obtained target generating component is a first mark generating component, wherein the first mark generating component represents the edge of the BIM model.

In an optional embodiment, the characterization vector includes local information corresponding to a plurality of distribution spaces in the BIM model, the local information includes a vector element corresponding to each candidate generation element, and the vector element characterizes a probability that the candidate generation element belongs to a generation element of the corresponding distribution space in the BIM model;

analyzing the integrated characterization vector based on different generation components through the mapping analysis operator to obtain a plurality of characterization vectors, determining an alternative generation component corresponding to the largest vector element in each local information in the characterization vector as a target generation component aiming at each characterization vector, and determining the obtained plurality of target generation components as a target generation component library corresponding to the characterization vector;

determining an integration vector element corresponding to each target generating component library, wherein the integration vector element is positively associated with vector elements corresponding to a plurality of target generating components in the target generating component library;

and determining a plurality of target generating components in the target generating component library with the largest integration vector elements in the target generating component library.

In an optional implementation manner, the parsing, by the mapping parsing operator, the integrated token vector based on different generating components respectively, to obtain a plurality of token vectors, for each token vector, determining, as a target generating component, an alternative generating component corresponding to a largest vector element in each local information in the token vector, and determining, as a target generating component library corresponding to the token vector, the obtained plurality of target generating components, including:

The number of the characterization vectors is a preset number, the mapping analysis operator analyzes the integrated characterization vectors to obtain first local information, and candidate generation components corresponding to the maximum preset number of vector elements in the first local information are respectively determined to be first target generation components;

for each first target generating component, repeatedly analyzing the integrated characterization vector based on the first target generating component through the mapping analysis operator to obtain second local information, determining an alternative generating component corresponding to the largest vector element in the second local information as a second target generating component, merging the second target generating component into a target generating component library where the first target generating component is located until the obtained target generating component is a first mark generating component, merging the first mark generating component into a target generating component library where the first target generating component is located, wherein the first mark generating component represents the edge of a BIM model, and each first target generating component is contained in a different target generating component library;

the first drawing data comprises a second mark generating component, the second mark generating component represents the edge of the drawing partition, the first drawing data is subjected to the extraction of the characterization vector by the characterization vector extraction operator to obtain drawing characterization vectors corresponding to each drawing partition in the first drawing data, and the method comprises the following steps:

Identifying second mark generating components in the first drawing data through the token vector extraction operator, determining the generating component before the first second mark generating component as one drawing partition, and determining the generating component between every two second mark generating components as one drawing partition;

and extracting the characterization vector from each drawing partition through the characterization vector extraction operator to obtain the drawing characterization vector corresponding to each drawing partition.

In an optional embodiment, before the building of the BIM model corresponding to the first drawing data based on the obtained first drawing data by the building of the network through the BIM model, the building method further includes generating components which are repeated with other generating components in the obtained second drawing data to obtain the first drawing data;

the obtaining a plurality of first network training templates includes:

acquiring a plurality of second network training templates, wherein the second network training templates comprise drawing data templates and BIM model example data corresponding to the drawing data templates, the BIM model example data comprise a plurality of families and corresponding constituent members of each family, and the plurality of second network training templates comprise drawing data templates of multiple types of modes;

For each second network training template, splitting the BIM model example data into a plurality of BIM model templates according to a plurality of families in the BIM model example data, wherein each BIM model template comprises a component corresponding to one family, and combining the drawing data template with the plurality of BIM model templates respectively to obtain a plurality of first network training templates.

According to another aspect of the present application, there is provided a data analysis system applied to construction project management, including a data upload terminal and a BIM model building apparatus communicatively connected to each other, the BIM model building apparatus including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method described above.

The application at least comprises the following beneficial effects:

according to the data analysis method and system applied to the building project management, the BIM model is automatically generated through the drawing data, and the model does not need to be built again for the existing drawing data, so that additional work is omitted, the working efficiency is improved, and the working cost is reduced. The BIM model construction network is debugged through drawing data templates of multiple types of modes and corresponding BIM models, so that the BIM model construction network can acquire the capacity of constructing the corresponding BIM model based on drawing data of each mode, the generalization capacity of the BIM model construction network is improved, and the corresponding BIM model can be constructed based on drawing data of at least one mode through the BIM model construction network, so that the generalization capacity of constructing the BIM model is improved.

It should be understood that the description of this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.

Drawings

The accompanying drawings illustrate exemplary embodiments and, together with the description, serve to explain exemplary implementations of the embodiments. The illustrated embodiments are for exemplary purposes only and do not limit the scope of the claims. Throughout the drawings, identical reference numerals designate similar, but not necessarily identical, elements.

Fig. 1 shows a schematic view of a scenario of a data analysis system applied to building project management according to an embodiment of the present application.

Fig. 2 and 3 show flowcharts of a data analysis method applied to building project management according to an embodiment of the present application, respectively.

Fig. 4 shows a functional block architecture schematic of a data analysis device according to an embodiment of the present application.

Fig. 5 shows a schematic composition diagram of a BIM model building apparatus according to an embodiment of the present application.

Detailed Description

Exemplary embodiments of the present application are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present application to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope of the present application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

In this application, unless otherwise indicated, the use of the terms "first," "second," etc. to describe various elements is not intended to limit the positional relationship, timing relationship, or importance of the elements, but is merely used to distinguish one element from another. In some examples, a first element and a second element may refer to the same instance of the element, and in some cases, they may also refer to different instances based on the description of the context.

The terminology used in the description of the various illustrated examples in this application is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context clearly indicates otherwise, the elements may be one or more if the number of the elements is not specifically limited. Furthermore, the term "and/or" as used in this application encompasses any and all possible combinations of the listed items.

Fig. 1 shows a schematic diagram of a data analysis system 100 applied to building project management provided according to an embodiment of the present application. The data analysis system 100 applied to building project management includes one or more data uploading terminals 101, a BIM model building device 120, and one or more communication networks 110 coupling the one or more data uploading terminals 101 to the BIM model building device 120.

In embodiments of the present application, BIM model building device 120 may run one or more services or software applications that enable execution of data analysis methods applied to building project management, such as applications that build a BIM model. In some embodiments, BIM model building device 120 may also provide other services or software applications that may include non-virtual environments and virtual environments. In some embodiments, these services may be provided as web-based services or cloud services, for example, provided to users of the data uploading terminal 101 under a software as a service (SaaS) model.

In the configuration shown in fig. 1, BIM model building device 120 may include one or more components that implement the functions performed by BIM model building device 120. These components may include software components, hardware components, or a combination thereof that are executable by one or more processors. A user operating data upload terminal 101 may in turn utilize one or more applications to interact with BIM model building device 120 to utilize the services provided by these components. It should be appreciated that a variety of different system configurations are possible, which may be different from the data analysis system 100 applied to building project management. Accordingly, FIG. 1 is one example of a system for implementing the various methods described herein and is not intended to be limiting. The user may upload drawing data, such as CAD drawing data, using the data upload terminal 101. The data uploading terminal 101 may provide an interface that enables a user of the data uploading terminal 101 to interact with the data uploading terminal 101. The data uploading terminal 101 may also output information to the user via the interface.

The data uploading terminal 101 may include various types of computer devices, such as portable handheld devices, general purpose computers (such as personal computers and laptop computers), workstation computers, smart screen devices, self-service terminal devices, and the like. These computer devices may run various types and versions of software applications and operating systems, such as MICROSOFT Windows, APPLE iOS, UNIX-like operating systems, linux, or Linux-like operating systems (e.g., GOOGLE Chrome OS); or include various mobile operating systems such as MICROSOFT Windows Mobile OS, iOS, windows Phone, android. Portable handheld devices may include cellular telephones, smart phones, tablet computers, personal Digital Assistants (PDAs), and the like. Network 110 may be any type of network known to those skilled in the art that may support data communications using any of a number of available protocols, including but not limited to TCP/IP, SNA, IPX, etc. For example only, the one or more networks 110 may be a Local Area Network (LAN), an ethernet-based network, a token ring, a Wide Area Network (WAN), the internet, a virtual network, a Virtual Private Network (VPN), an intranet, an extranet, a blockchain network, a Public Switched Telephone Network (PSTN), an infrared network, a wireless network (e.g., bluetooth, WIFI), and/or any combination of these and/or other networks.

BIM model building device 120 may include one or more general purpose computers, special purpose server computers (e.g., PC (personal computer) servers, UNIX servers, mid-end servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and/or combination. BIM model building device 120 may include one or more virtual machines running a virtual operating system, or other computing architecture that involves virtualization (e.g., one or more flexible pools of logical storage devices that may be virtualized to maintain virtual storage devices of servers). In various embodiments, BIM model building device 120 may run one or more services or software applications that provide the functionality described below.

The computing unit in BIM model building device 120 may run one or more operating systems including any of the operating systems described above as well as any commercially available server operating systems. BIM model building device 120 may also run any of a variety of additional server applications and/or middle tier applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, etc.

In some implementations, BIM model building device 120 may include one or more applications to analyze and incorporate data feeds and/or event updates received from a user of data uploading terminal 101. BIM model building device 120 may also include one or more applications to display data feeds and/or real-time events via one or more display devices of data uploading terminal 101.

In some implementations, BIM model building device 120 may be a server of a distributed system or a server that incorporates a blockchain. BIM model building device 120 may also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence techniques. The cloud server is a host product in a cloud computing service system, so as to solve the defects of large management difficulty and weak service expansibility in the traditional physical host and virtual private server (VPS, virtual Private Server) service.

The data analysis system 100 applied to building project management may also include one or more databases 130. In some embodiments, these databases may be used to store data and other information. For example, one or more of databases 130 may be used to store building elements, drawing data, BIM model data, and the like. Database 130 may reside in various locations. For example, the database used by BIM model building device 120 may be local to BIM model building device 120 or may be remote from BIM model building device 120 and may communicate with BIM model building device 120 via a network-based or dedicated connection. Database 130 may be of different types. In some embodiments, the database used by BIM model building device 120 may be, for example, a relational database. One or more of these databases may store, update, and retrieve the databases and data from the databases in response to the commands.

In some embodiments, one or more of databases 130 may also be used by applications to store application data. The databases used by the application may be different types of databases, such as key value stores, object stores, or conventional stores supported by the file system.

The system 100 of fig. 1 may be configured and operated in various ways to enable application of the various methods and apparatus described in accordance with the present application.

Referring to fig. 2, a flowchart of a data analysis method applied to building project management according to an embodiment of the present application is shown, where the method may be executed by a BIM model building device 120, and may specifically include the following steps:

step 110: a plurality of first network training templates is acquired.

The first network training template comprises a drawing data template and a BIM model template corresponding to the drawing data template. The drawing data template is drawing data which needs BIM model conversion, such as CAD drawing, and comprises a plurality of drawing partitions, each drawing partition corresponds to a different drawing area, and each drawing area is a complete component, such as a wall body. The BIM model template corresponding to the drawing data template is a matched BIM model obtained after information acquisition and conversion of the drawing data template. Each first network training template comprises a drawing data template and a BIM model template corresponding to the drawing data template, the plurality of first network training templates comprise drawing data templates with multiple types of modes, the modes of the drawing data templates are design elements corresponding to the drawing data templates, such as curtain walls, partitions, steps, stair layers and other design elements in the building field.

Step 120: and debugging the BIM model according to the plurality of first network training templates to construct a network.

And repeatedly debugging the BIM model building network according to drawing data templates and BIM model templates in the plurality of first network training templates to obtain a debugged BIM model building network. And in each debugging process, debugging the BIM model according to one or more first network training templates in the plurality of first network training templates to construct a network. The BIM model construction network is debugged through the drawing data templates of the multiple types of modes and the BIM models corresponding to the drawing data templates, so that the BIM model construction network can acquire the capacity of how to construct the corresponding BIM model based on the drawing data of each mode, and the debugged BIM model construction network is used for constructing the corresponding BIM model based on the drawing data of at least one mode in the multiple types of modes, so that the generalization capacity of the BIM model construction network is improved.

Step 130: and constructing a BIM model corresponding to the first drawing data based on the acquired first drawing data through a BIM model construction network.

According to the method and the device for constructing the BIM model, the first drawing data of at least one mode is obtained, and because the BIM model construction network has generalization capability, the BIM model corresponding to the first drawing data is constructed based on the first drawing data through the BIM model construction network.

Referring to fig. 3, a flowchart of another data analysis method applied to building project management according to an embodiment of the present application may specifically include the following steps:

step 210: a plurality of first network training templates is acquired.

The first network training templates comprise drawing data templates and BIM model templates corresponding to the drawing data templates, and the drawing data templates of multiple types of modes are included in the first network training templates. The drawing data templates comprise a plurality of drawing partitions, and BIM model templates corresponding to the drawing data templates are matched BIM models obtained after information acquisition and conversion of the drawing data templates. The BIM model building network debugging system comprises a plurality of first network training templates, a BIM model building network, a plurality of second network training templates, a BIM model building network and a computer, wherein the plurality of first network training templates are used for debugging the BIM model building network, and the BIM model building network is debugged repeatedly according to drawing data templates and the BIM model templates in the plurality of first network training templates to obtain a debugged BIM model building network. And at each debugging, debugging the BIM model to construct a network according to one or more first network training templates in the plurality of first network training templates. Specifically, in a debugging process, the method for performing debugging according to the first network training template may include the following steps:

Step 220: and determining a template characterization vector corresponding to the BIM model template according to the alternative generation component library and the BIM model template.

In each round of debugging, a first network training template is obtained from a plurality of first network training templates, and an alternative generation component library is obtained. And determining a template characterization vector corresponding to the BIM model template according to the BIM model template in the alternative generation component library and the first network training template. The candidate generation component library comprises a plurality of candidate generation components, and the template characterization vector characterizes the probability that each candidate generation component belongs to the BIM model template. The alternative generating component library comprises a plurality of components for generating the BIM model.

Optionally, for each candidate generating component in the candidate generating component library, determining whether the candidate generating component belongs to a generating component in the BIM model template, if the candidate generating component belongs to a generating component in the BIM model template, determining a first vector element (a feature value, for example, 1) as a vector element corresponding to the candidate generating component, and if the candidate generating component does not belong to a generating component in the BIM model template, determining a second vector element (for example, 0) as a vector element corresponding to the candidate generating component, and combining the vector elements corresponding to each candidate generating component as the template characterization vector. As other embodiments, the template characterization vector includes template local information corresponding to a plurality of distribution spaces in the BIM model template (i.e., locations distributed in the BIM model), the template local information characterizing a probability that each of the candidate generating members belongs to a generating member of the corresponding distribution space in the BIM model template. For example, for each distribution space in the BIM model template, determining whether each candidate generation component belongs to a generation component in the distribution space, if the candidate generation component belongs to a generation component in the distribution space, determining the first vector element as a vector element corresponding to the candidate generation component, if the candidate generation component does not belong to a generation component in the distribution space, determining the second vector element as a vector element corresponding to the candidate generation component, determining the vector element corresponding to each candidate generation component as template local information corresponding to the distribution space, and obtaining template local information corresponding to each distribution space. And determining the template local information corresponding to the plurality of distribution spaces as a template characterization vector corresponding to the BIM model template.

Step 230: and constructing a characterization vector extraction operator in the network through the BIM model, extracting the characterization vector of the drawing data template, and obtaining an inference drawing characterization vector corresponding to each drawing partition in the drawing data template.

The BIM model building network comprises a characterization vector extraction operator, wherein the characterization vector extraction operator is used for extracting a characterization vector of the drawing data, and the characterization vector is used for characterizing characteristic information of the drawing data. The method comprises the steps of obtaining a drawing data template in a first network training template, wherein the drawing data template comprises a plurality of drawing partitions, extracting a characterization vector from the drawing data template through a characterization vector extraction operator to obtain an reasoning drawing characterization vector corresponding to each drawing partition in the drawing data template, and obtaining a plurality of reasoning drawing characterization vectors, wherein the reasoning drawing characterization vectors corresponding to the drawing partitions characterize the semantics of the drawing partitions. Optionally, the drawing data template includes a second mark generating component, where the second mark generating component represents an edge of the drawing partition, such as a corner of a wall, and a junction between two walls, and the outside of the second mark generating component is not connected with a new component (wall). Identifying second mark generating components in the drawing data template through a token vector extraction operator, determining the generating component before the first second mark generating component as a drawing partition, and determining the generating component between every two second mark generating components as a drawing partition. And extracting the characterization vector from each drawing partition through a characterization vector extraction operator to obtain an inference drawing characterization vector corresponding to each drawing partition. Identifying second mark generating components in the drawing data template by the token vector extraction operator, determining the generating component before the first second mark generating component as a drawing partition, determining the generating component between the first second mark generating component and the second mark generating component as a drawing partition, and determining the generating component between the second mark generating component and the third mark generating component as a drawing partition. The drawing data template may further include a third mark generating means representing a start means of the drawing data template, the generating means between the third mark generating means and a first second mark generating means subsequent to the third mark generating means being determined as one drawing division.

Optionally, the token vector extraction operator includes a token vector extraction operator and a token vector embedding mapping operator, and token vector extraction is performed on each drawing partition in the drawing data template through the token vector extraction operator to obtain one hot code corresponding to each drawing partition, and then embedding mapping is performed on the one hot code corresponding to each drawing partition (i.e. the process of refining the coding features), so as to obtain an inference drawing token vector corresponding to each drawing partition. The token vector embedding mapping operator can be a token vector embedding mapping operator which is debugged in advance, and the token vector is extracted from the one hot code to obtain a drawing token vector.

Step 240: and constructing a feature embedding mapping operator in the network through the BIM model, and carrying out integration embedding mapping on the obtained multiple inference drawing characterization vectors to obtain an inference integration characterization vector.

The BIM model building network further comprises a feature embedding mapping operator, wherein the feature embedding mapping operator is used for integrating and embedding mapping on the drawing characterization vectors to obtain a plurality of reasoning drawing characterization vectors, and the feature embedding mapping operator is used for integrating and embedding mapping on the reasoning drawing characterization vectors to obtain reasoning and integrating embedding mapping, so that semantic information of the reasoning and integrating embedding mapping characterization drawing data template is obtained.

Step 250: and constructing a mapping analysis operator in the network through the BIM model, and analyzing according to the reasoning integration characterization vector to obtain the reasoning characterization vector.

The BIM model building network further comprises a mapping analysis operator, wherein the mapping analysis operator is used for analyzing the reasoning integration characterization vector so as to finish decoding, in other words, the mapping analysis operator is a decoder, after the reasoning integration characterization vector corresponding to the drawing data template is obtained, the mapping analysis operator is used for analyzing according to the reasoning integration characterization vector so as to obtain the reasoning characterization vector corresponding to the drawing data template. The inference characterization vector is used for generating an inference BIM model, and the inference characterization vector represents the probability that each alternative generating component in the alternative generating component library belongs to the inference BIM model corresponding to the drawing data template. Since the inference token vector represents the probability that the alternative generating means belongs to the inference BIM model, the generating means comprised in the inference BIM model can be determined based on the inference token vector, that is, the inference BIM model corresponding to the drawing data template depends on the inference token vector.

Optionally, the inference characterization vector includes inference local information corresponding to a plurality of distribution spaces in the inference BIM model, the inference local information includes an inference vector element corresponding to each candidate generation component, and the inference vector element characterizes a probability that the candidate generation component belongs to the generation component of the corresponding distribution space in the inference BIM model. Then, analyzing according to the reasoning integration characterization vector by mapping the analysis operator to obtain the reasoning characterization vector, including: analyzing the reasoning integration characterization vector through a mapping analysis operator to obtain first reasoning local information, and determining an alternative generating component corresponding to the maximum reasoning vector element in the first reasoning local information as a first reasoning generating component; repeatedly analyzing the inference integration characterization vector based on the obtained inference generating components through a mapping analysis operator to obtain second inference local information, and determining an alternative generating component corresponding to the largest inference vector element in the second inference local information as a second inference generating component until the obtained inference generating component is a first mark generating component, wherein the first mark generating component represents the edge of the BIM model.

And resolving the reasoning integration characterization vector through a mapping resolution operator to obtain first reasoning local information, wherein the first reasoning local information is the reasoning local information corresponding to the first distribution space in the reasoning BIM model, and comprises reasoning vector elements corresponding to each alternative generation component, and the reasoning vector elements corresponding to the alternative generation components characterize the probability that the alternative generation components belong to the generation components on the first distribution space. Then, determining the largest reasoning vector element in the first reasoning local information, wherein the probability that the candidate generating component corresponding to the largest reasoning vector element belongs to the generating component on the first distribution space is largest, and determining the candidate generating component corresponding to the largest reasoning vector element as the first reasoning generating component in the reasoning BIM model, namely, the reasoning generating component of the first distribution space in the reasoning BIM model. After determining the first inference generating component, repeatedly through a mapping analysis operator, analyzing the inference integration characterization vector based on the obtained inference generating component, in other words, based on the first inference generating component, to obtain second inference local information, wherein the second inference local information is the inference local information corresponding to the second distribution space in the inference BIM model, the second inference local information comprises an inference vector element corresponding to each alternative generating component, and the inference vector element corresponding to the alternative generating component characterizes the probability that the alternative generating component belongs to the generating component on the second distribution space. And determining the largest reasoning vector element in the second reasoning local information, wherein the probability that the candidate generating component corresponding to the largest reasoning vector element belongs to the generating component on the second distribution space is largest, and determining the candidate generating component of the largest reasoning vector element as the second reasoning generating component in the reasoning BIM model, namely the second reasoning generating component on the second distribution space in the reasoning BIM model. After determining the second inference generating component, repeatedly analyzing the inference integration characterization vector based on the obtained inference generating component, namely based on the first inference generating component and the second inference generating component through a mapping analysis operator to obtain third inference local information, and correspondingly, determining an alternative generating component corresponding to the largest inference vector element in the third inference local information as a third inference generating component in the inference BIM model, namely a third inference generating component in the inference BIM model in a distribution space.

In this way, the inference integration token vector is parsed based on the obtained inference generating components each time, and the next inference generating component is obtained until the obtained inference generating components are first mark generating components, the first mark generating components represent edges of the BIM model, and if the inference generating components are first mark generating components, each generation component in the obtained BIM model is represented, the parsing of the inference integration token vector is stopped. As one implementation mode, the obtained one or more inference generating components are obtained to conduct characterization vector extraction on the one or more inference generating components, a generated component characterization vector is obtained, the generated component characterization vector characterizes semantic information of at least one inference generating component, and the integrated characterization vector and the generated component characterization vector are analyzed through a mapping analysis operator to obtain next inference local information.

Step 260: and debugging the BIM model according to the template characterization vector and the reasoning characterization vector to construct a network.

Because the template characterization vector characterizes the probability that each alternative generating component belongs to the BIM model template, the inference characterization vector characterizes the probability that each alternative generating component belongs to the inference BIM model corresponding to the drawing data template, the smaller the error between the inference characterization vector and the template characterization vector is, the smaller the error between the inference BIM model and the BIM model template obtained according to the BIM model construction network is, the stronger the accuracy of the BIM model construction network is, the BIM model construction network can be debugged according to the error between the inference characterization vector and the template characterization vector, and the debugging BIM model construction network is simpler. Because the application embodiment debugs the BIM model construction network through the drawing data templates of the multiple types of modes and the corresponding BIM models, the BIM model construction network can acquire the capacity of how to construct the corresponding BIM model based on the drawing data of each mode, and the debugged BIM model construction network is used for constructing the corresponding BIM model based on the drawing data of at least one mode in the multiple types of modes, so that the generalization capacity of the BIM model construction network is improved.

Optionally, the template characterization vector includes template local information corresponding to a plurality of distribution spaces in the BIM model template, the template local information characterizes probabilities that each candidate generating component belongs to generating components of the corresponding distribution spaces in the BIM model template, and the inference characterization vector includes inference local information corresponding to the plurality of distribution spaces in the inference BIM model, and the inference local information characterizes probabilities that each candidate generating component belongs to generating components of the corresponding distribution spaces in the inference BIM model. Debugging the BIM model to construct a network according to the template characterization vector and the reasoning characterization vector, wherein the method specifically comprises the following steps of: determining a sub-error corresponding to each distribution space according to the error between the reasoning local information corresponding to each distribution space and the template local information; determining an integrated error according to the plurality of sub-errors, wherein the integrated error is positively correlated with the plurality of sub-errors; and debugging the BIM model according to the integration error to construct a network. The larger the error between the reasoning local information and the template local information is, the larger the sub-errors are, and the integration error is positively correlated with the plurality of sub-errors, in other words, the larger the sub-errors are, the larger the integration error is, and the integration error can be the error obtained by adding the plurality of sub-errors. In other words, the larger the error between the reasoning local information and the template local information corresponding to each distribution space is, the larger the integration error is, and the lower the accuracy of building the BIM model into the network is. And then the BIM model is debugged according to the integration errors to construct a network, so that the integration errors are gradually converged, and the accuracy of embedding the mapping network can be improved. Optionally, the feature embedding mapping operator and the mapping analysis operator in the BIM model building network are convectors, the feature embedding mapping operator focuses other generating components before the current generating component through internal attention (Self-attention) and forward network (FNN) to obtain overall context information, and the mapping analysis operator comprises Self-attention, FNN and an information focusing module (Attention Mechanism) to enable the mapping analysis operator to identify feature information to be focused.

Based on the above, the embodiment of the application learns the BIM model template, loads the BIM model template and the drawing data template to the BIM model construction network for debugging, adopts BIM model templates corresponding to drawing data templates of different modes to autonomously analyze and learn comprehensive rich information of the BIM model, so that the construction of the BIM model is not constrained, and the generalization capability of the BIM model construction network is improved.

In the above operation flow, only in each round of debugging, the debugging is performed according to one first network training template, and according to different first network training templates, the above operation flows 220-260 can be repeatedly performed, so as to debug the BIM model according to a plurality of first network training templates to construct a network.

Step 270: and constructing a BIM model corresponding to the first drawing data based on the acquired first drawing data through a BIM model construction network.

The method comprises the steps that the obtained first drawing data corresponding to at least one mode is used for building a BIM model corresponding to the first drawing data through a BIM model building network because the BIM model building network has generalization capability. The BIM model is automatically generated through the drawing data, and the existing drawing data does not need to be re-built, so that additional work is omitted, the working efficiency is improved, and the working cost is reduced.

According to the method and the device for constructing the BIM model, the BIM model construction network is debugged through drawing data templates of multiple types of modes and the corresponding BIM models, so that the BIM model construction network can acquire the capacity of constructing the corresponding BIM model based on drawing data of each mode, the generalization capacity of the BIM model construction network is improved, and therefore the corresponding BIM model can be constructed based on drawing data of at least one mode through the BIM model construction network, and the generalization capacity of constructing the BIM model is improved. In addition, because the template characterization vector characterizes the probability that each alternative generating component belongs to the BIM model template, the inference characterization vector characterizes the probability that each alternative generating component belongs to the inference BIM model corresponding to the drawing data template, the smaller the error between the inference characterization vector and the template characterization vector is, the smaller the error between the inference BIM model and the BIM model template obtained according to the BIM model construction network is, the higher the precision of the BIM model construction network is, the BIM model construction network is debugged according to the error between the inference characterization vector and the template characterization vector, and the debugging of the BIM model construction network is easier.

In another embodiment, an embodiment of the present application provides a data analysis method applied to building project management, including:

Step 310: and acquiring first drawing data.

The first drawing data includes a plurality of drawing partitions. Optionally, deleting the generating component which is repeated with other generating components in the second drawing data to obtain the first drawing data. As one embodiment, each time a drawing partition is obtained, a second mark generating member is provided at an edge of the drawing partition, the second mark generating member represents the edge of the drawing partition, and a plurality of drawing partitions provided with the second mark generating member are combined to obtain first drawing data to determine each drawing partition in the first drawing data based on the second mark generating member in the first drawing data.

Step 320: and carrying out characterization vector extraction on the first drawing data by a characterization vector extraction operator in the BIM model construction network to obtain drawing characterization vectors corresponding to each drawing partition in the first drawing data.

Optionally, the first drawing data includes a second mark generating member, and the second mark generating member represents an edge of the drawing partition. Identifying second mark generating components in the first drawing data through a token vector extraction operator, determining the generating component before the first second mark generating component as a drawing partition, and determining the generating components between every two second mark generating components as a drawing partition; and extracting the characterization vector from each drawing partition through a characterization vector extraction operator to obtain a drawing characterization vector corresponding to each drawing partition.

Step 330: and constructing a feature embedding mapping operator in a network through the BIM model, and carrying out integration embedding mapping on the obtained multiple drawing characterization vectors to obtain an integration characterization vector.

Step 340: and constructing a mapping analysis operator in a network through the BIM model, analyzing according to the integrated characterization vector to obtain a characterization vector, and determining a plurality of target generation components belonging to the BIM model from a plurality of candidate generation components in a candidate generation component library based on the characterization vector.

The representation vector represents the probability that each alternative generating component belongs to the BIM model corresponding to the first drawing data. Optionally, the characterization vector includes local information corresponding to a plurality of distribution spaces in the BIM model, the local information includes a vector element corresponding to each candidate generating member, and the vector element characterizes a probability that the candidate generating member belongs to the generating member of the corresponding distribution space in the BIM model. Resolving according to the integrated token vector by a mapping resolving operator to obtain a token vector, and determining a plurality of target generating components belonging to the BIM model from a plurality of candidate generating components in a candidate generating component library based on the token vector, wherein the method may include: analyzing the integrated characterization vector through a mapping analysis operator to obtain first local information, and determining an alternative generating component corresponding to the largest vector element in the first local information as a first target generating component; and repeatedly analyzing the integral characterization vector based on the obtained target generating component through a mapping analysis operator to obtain second local information, and determining an alternative generating component corresponding to the largest vector element in the second local information as a second target generating component until the obtained target generating component is a first mark generating component, wherein the first mark generating component represents the edge of the BIM model. As other embodiments, the token vector includes local information corresponding to a plurality of distribution spaces in the BIM model, the local information includes vector elements corresponding to each candidate generation element, and the vector elements characterize probabilities that the candidate generation element belongs to the generation element of the corresponding distribution space in the BIM model.

Resolving according to the integrated token vector by a mapping resolving operator to obtain a token vector, and determining a plurality of target generating components belonging to the BIM model from a plurality of candidate generating components in a candidate generating component library based on the token vector, wherein the method may include: analyzing the integral characterization vector based on different generation components through a mapping analysis operator to obtain a plurality of characterization vectors, determining an alternative generation component corresponding to the largest vector element in each local information in the characterization vector as a target generation component aiming at each characterization vector, and determining the obtained plurality of target generation components as a target generation component library corresponding to the characterization vector; determining an integration vector element corresponding to each target generating component library, wherein the integration vector element is positively associated with vector elements corresponding to a plurality of target generating components in the target generating component library; a plurality of target generating components in a target generating component library with the largest integration vector element are determined in the plurality of target generating component libraries. The plurality of characterization vectors are respectively used for determining BIM models, the plurality of characterization vectors are different, the BIM models determined by each characterization vector are different, the plurality of characterization vectors are generated in order to determine the most accurate characterization vector in the plurality of characterization vectors, the most accurate BIM model is obtained based on the determined characterization vector, the candidate generation component corresponding to the largest vector element in the local information is the highest probability of belonging to the BIM model, the candidate generation component corresponding to the largest vector element in the local information is determined to be the target generation component, and the plurality of target generation components determined based on one characterization vector are determined to be the target generation component library corresponding to the characterization vector.

The larger the vector elements corresponding to the plurality of target generating components in the target generating component library, the larger the integrated vector elements corresponding to the target generating component library, and the smaller the vector elements corresponding to the plurality of target generating components in the target generating component library, the smaller the integrated vector elements corresponding to the target generating component library. For example, the vector elements corresponding to the plurality of target generating components are integrated to obtain the integrated vector elements corresponding to the target generating component library. Because the larger the vector element corresponding to the target generating member is, the larger the probability that the target generating member belongs to the generating member in the BIM model is, the larger the integrated vector element corresponding to the target generating member library is, the probability that the target generating member in the target generating member library belongs to the generating member in the BIM model is, the higher the accuracy of the BIM model determined based on the plurality of target generating members in the target generating member library is, and the plurality of target generating members in the target generating member library with the largest integrated vector element are determined from the plurality of target generating member libraries.

In one embodiment, the number of token vectors is a preset number, and the mapping analysis operator is used for respectively analyzing the integrated token vectors based on different generating components to obtain a plurality of token vectors, which may include: analyzing the integrated characterization vector through a mapping analysis operator to obtain first local information, and respectively determining candidate generation components corresponding to the maximum preset number of vector elements in the first local information as first target generation components; and repeatedly analyzing the integral characterization vector based on the first target generating components by using a mapping analysis operator for each first target generating component to obtain second local information, determining an alternative generating component corresponding to the largest vector element in the second local information as a second target generating component, merging the second target generating component into a target generating component library where the first target generating component is located until the obtained target generating component is a first mark generating component, merging the first mark generating component into the target generating component library where the first target generating component is located, wherein the first mark generating component represents the edge of the BIM model, and each first target generating component is contained in a different target generating component library.

And resolving the integral characterization vector through a mapping resolving operator to obtain first local information corresponding to a first distribution space in the BIM model, wherein the first local information comprises vector elements corresponding to each alternative generation component, and the vector elements corresponding to the alternative generation components characterize the probability that the alternative generation components belong to the generation components on the first distribution space. And determining the largest preset number of vector elements in the first local information, wherein the probability that the candidate generating components corresponding to the preset number of vector elements belong to generating components on the first distribution space is higher, and determining the candidate generating components corresponding to the preset number of vector elements as the first target generating components in the BIM model, namely the generating components on the first distribution space in the BIM model. For each first target generating component, determining a second local information corresponding to a second distribution space of the BIM model based on the first target generating component and the integrated characterization vector again, determining a second target generating component based on the second local information, taking the first target generating component and the second target generating component as generating components in a target generating component library, and taking the first local information and the second local information as local information in the characterization vector. Repeatedly, determining a third local information of a third distribution space of the BIM model based on the first target generating means, the second target generating means and the integrated token vector, determining a third target generating means based on the third local information, merging the third target generating means into the target generating means library again, and merging the third local information into the token vector, and proceeding in this manner until the first tag generating means is merged into the target generating means library. Repeatedly analyzing the integral characterization vector based on the first target generating component through a mapping analysis operator to obtain second local information, taking an alternative generating component corresponding to the largest vector element in the second local information as a second target generating component, merging the second target generating component into a target generating component library where the first target generating component is located until the obtained target generating component is the first mark generating component, merging the first mark generating component into the target generating component library where the first target generating component is located. For each first target generating component, according to the operation, a plurality of characterization vectors and a target generating component library corresponding to each characterization vector can be obtained.

Step 350: each of the target generating members is combined to obtain a BIM model.

After a plurality of target generating components are obtained, each target generating component is combined to obtain a BIM model corresponding to the first drawing data. Wherein a plurality of target generating members may be combined according to a construction order of the target generating members to obtain the BIM model.

On the premise of the data analysis method applied to building project management, a plurality of BIM model building networks can be debugged and deployed, and each BIM model building network is used for building different groups of constituent members in the BIM model and specifically comprises the following steps:

step 410: a plurality of second network training templates is acquired.

The second network training templates comprise drawing data templates and BIM model example data corresponding to the drawing data templates, the drawing data templates comprise a plurality of drawing partitions, and the drawing data templates of multiple types of modes are included in the second network training templates.

Step 420: for each second network training template, splitting BIM model example data into a plurality of BIM model templates according to a plurality of families in the BIM model example data, and respectively combining the drawing data templates with the plurality of BIM model templates to obtain a plurality of first network training templates.

And aiming at each second network training template, acquiring BIM model example data in the second network training templates, decomposing the BIM model example data into a plurality of BIM model templates according to a plurality of families (the families are units of component items in BIM and are carriers of parameter information, and are primitive groups containing a general attribute set and related graphic representations) in the BIM model example data, wherein each BIM model template comprises a corresponding component member of the family. In other words, for each family, the constituent members corresponding to that family are determined as one BIM model template. Then, for the drawing data template in the second network training template, a plurality of BIM model templates are corresponding to the drawing data template, and the drawing data template and the BIM model templates are respectively combined to obtain a plurality of first network training templates. That is, for each BIM model template of a plurality of BIM model templates, the BIM model template is combined with the drawing data template to obtain a first network training template. For example, if the BIM model example data of the drawing data template is decomposed into a BIM model template a and a BIM model template B according to the family, the drawing data template corresponds to the BIM model template a and the BIM model template B, the drawing data template and the BIM model template a are combined to obtain one first network training template, and the drawing data template and the BIM model template B are combined to obtain another first network training template.

Step 430: and debugging one BIM model building network based on a plurality of first network training templates corresponding to the same family respectively to obtain a plurality of BIM model building networks.

The family corresponding first network training templates are the BIM model templates in the first network training templates that include the family corresponding constituent members. Because there are cases where different BIM model example data includes a family, there are multiple first network training templates corresponding to the same family. For each family, determining a plurality of first network training templates corresponding to the family, and debugging a BIM model building network according to the plurality of first network training templates, wherein the BIM model building network is used for building a component corresponding to the family based on drawing data of at least one mode in a plurality of types of modes so as to obtain the BIM model building network corresponding to each family, and the BIM model building network has generalization capability. The manner in which the BIM model is adapted to build a network in accordance with the plurality of first network training templates has been described above and will not be described.

Step 440: and respectively constructing a network through a plurality of BIM models, and constructing a corresponding component of each family based on the first drawing data.

And (3) acquiring first drawing data of at least one mode, wherein the debugged BIM model building networks have generalization capability, so that the BIM model building networks are respectively used for building corresponding constituent members of each family based on the first drawing data. And constructing a network aiming at the BIM model corresponding to each group, constructing a network through the BIM model, and constructing the component members corresponding to the group based on the first drawing data so as to obtain the component members corresponding to each group.

Step 450: and combining the constructed plurality of constituent members with a plurality of families to obtain a BIM model corresponding to the first drawing data.

After the plurality of constituent members are constructed, the plurality of constituent members are combined with the plurality of families to obtain a BIM model corresponding to the first drawing data. Wherein each constituent member corresponds to a family, each constituent member is combined with the corresponding family to obtain a local model, and a plurality of local models are combined to obtain the BIM model.

According to the embodiment of the application, the BIM model construction network is debugged through the drawing data templates of multiple types of modes and the corresponding BIM models, so that the BIM model construction network can acquire the capacity of constructing the corresponding BIM model based on the drawing data of each mode, the generalization capacity of the BIM model construction network is improved, and the corresponding BIM model can be constructed based on the drawing data of at least one mode through the BIM model construction network, so that the generalization capacity of constructing the BIM model is improved. In addition, the family in BIM model example data is taken as the direction, the BIM model example data is decomposed into different BIM model templates, BIM model construction networks for constructing constituent members corresponding to different families are obtained through respective debugging, the granularity of BIM model construction is smaller, and the accuracy of the BIM model obtained through construction is higher.

In accordance with another aspect of the present application, there is also provided a data analysis apparatus, referring to fig. 4, an apparatus 400 includes:

the template obtaining module 410 is configured to obtain a plurality of first network training templates, where the first network training templates include drawing data templates and BIM model templates corresponding to the drawing data templates, and the plurality of first network training templates include drawing data templates of multiple types of modalities;

the network debugging module 420 is configured to debug a BIM model to construct a network according to the plurality of first network training templates, where the BIM model constructing network is configured to construct a corresponding BIM model based on drawing data of at least one mode of the multiple types of modes;

the model construction module 430 is configured to construct a network through the BIM model, and construct a BIM model corresponding to the first drawing data based on the acquired first drawing data;

wherein, the network debug module 420 is specifically configured to:

According to embodiments of the present application, there is also provided an electronic device, a readable storage medium and a computer program product. Referring to fig. 5, which is a block diagram of the electronic device 1000 of the BIM model building device of the present application, the electronic device 1000 includes a computing unit 1001 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 1002 or a computer program loaded from a storage unit 1008 into a Random Access Memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the electronic apparatus 1000 can also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other by a bus 1004. An input/output (I/O) interface 1005 is also connected to bus 1004.

Various components in the electronic device 1000 are connected to the I/O interface 1005, including: an input unit 1006, an output unit 1007, a storage unit 1008, and a communication unit 10010. The input unit 1006 may be any type of device capable of inputting information to the electronic device 1000, the input unit 1006 may receive input numeric or character information and generate key signal inputs related to user settings and/or function control of the electronic device, and may include, but is not limited to, a mouse, a keyboard, a touch screen, a trackpad, a trackball, a joystick, a microphone, and/or a remote control. The output unit 1007 may be any type of device capable of presenting information and may include, but is not limited to, a display, speakers, video/audio output terminals, vibrators, and/or printers. Storage unit 1008 may include, but is not limited to, magnetic disks, optical disks. The communication unit 10010 allows the electronic device 1000 to exchange information/data with other devices through a computer network, such as the internet, and/or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers and/or chipsets, such as bluetooth (TM) devices, 802.11 devices, wiFi devices, wiMax devices, cellular communication devices, and/or the like.

The computing unit 1001 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 1001 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 performs the various methods and processes described above, such as method 200. For example, in some embodiments, the method 200 may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 1000 via the ROM 1002 and/or the communication unit 1009. One or more of the steps of the method 200 described above may be performed when the computer program is loaded into RAM 1003 and executed by the computing unit 1001. Alternatively, in other embodiments, the computing unit 1001 may be configured to perform the method 200 in any other suitable way (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present application may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this application, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present application may be performed in parallel, sequentially or in a different order, provided that the desired results of the technical solutions disclosed herein are achieved, and are not limited herein.

Although embodiments or examples of the present application have been described with reference to the accompanying drawings, it is to be understood that the foregoing methods, systems, and apparatus are merely exemplary embodiments or examples, and that the scope of the present invention is not limited by these embodiments or examples but only by the claims following the grant and their equivalents. Various elements of the embodiments or examples may be omitted or replaced with equivalent elements thereof. Furthermore, the steps may be performed in a different order than described in the present application. Further, various elements of the embodiments or examples may be combined in various ways. It is important that as technology evolves, many of the elements described herein may be replaced by equivalent elements that appear after the application.

Claims

1. A data analysis method applied to building project management, characterized in that it is applied to a BIM model building apparatus, the method comprising:

Debugging the BIM model to construct a network according to the template characterization vector and the reasoning characterization vector;

the BIM model construction network comprises a token vector extraction operator, a feature embedding mapping operator and a mapping analysis operator, and the drawing data template is processed through the BIM model construction network to obtain an inference token vector, and the method comprises the following steps:

analyzing according to the reasoning integration characterization vector by the mapping analysis operator to obtain the reasoning characterization vector;

the inference characterization vector comprises inference local information corresponding to a plurality of distribution spaces in the inference BIM model, the inference local information comprises inference vector elements corresponding to each alternative generation component, and the inference vector elements characterize the probability that the alternative generation components belong to the generation components of the corresponding distribution spaces in the inference BIM model;

2. The method of claim 1, wherein the template characterization vector includes template local information corresponding to a plurality of distribution spaces in the BIM model template, the template local information characterizing a probability that each of the candidate generation element belongs to a generation element of a corresponding distribution space in the BIM model template, the inference characterization vector includes inference local information corresponding to a plurality of distribution spaces in the inference BIM model, the inference local information characterizing a probability that each of the candidate generation element belongs to a generation element of a corresponding distribution space in the inference BIM model;

3. The method of claim 1, wherein the constructing a network by the BIM model, based on the obtained first drawing data, constructs a BIM model corresponding to the first drawing data, includes:

each of the target generating means is combined to obtain the BIM model.

4. A method according to claim 3, wherein the BIM model building network comprises a token vector extraction operator, a feature embedding mapping operator and a mapping parsing operator, the identifying the first drawing data by the BIM model building network resulting in a token vector, determining a plurality of target generating components belonging to the BIM model among a plurality of candidate generating components in a candidate generating component library based on the token vector, comprising:

5. The method of claim 4, wherein the characterization vector includes local information corresponding to a plurality of distribution spaces in the BIM model, the local information including a vector element corresponding to each of the alternative generating members, the vector element characterizing a probability that the alternative generating member belongs to a generating member of the corresponding distribution space in the BIM model;

6. The method according to claim 5, wherein the parsing, by the mapping parsing operator, the integrated token vector based on different generating means respectively, to obtain a plurality of token vectors, determining, for each token vector, an alternative generating means corresponding to a largest vector element in each local information in the token vector as a target generating means, and determining the obtained plurality of target generating means as a target generating means library corresponding to the token vector, includes:

7. The method of claim 1, wherein the constructing a network by the BIM model, based on the obtained first drawing data, further includes deleting, from the obtained second drawing data, a generating member that is repeated with other generating members in the second drawing data, and obtaining the first drawing data, before constructing a BIM model corresponding to the first drawing data;

the obtaining a plurality of first network training templates includes:

8. A data analysis system for building project management, comprising a data uploading terminal and a BIM model building device communicatively connected to each other, the BIM model building device comprising:

at least one processor;

and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.