CN111597170B

CN111597170B - Method for building spatial semantic database from BIM model without damage

Info

Publication number: CN111597170B
Application number: CN202010707765.4A
Authority: CN
Inventors: 刘俊伟; 彭贵堂; 唐莉萍
Original assignee: Terra It Technology Beijing Co ltd
Current assignee: Terry Digital Technology Beijing Co ltd
Priority date: 2020-07-21
Filing date: 2020-07-21
Publication date: 2021-06-08
Anticipated expiration: 2040-07-21
Also published as: CN111597170A

Abstract

The invention discloses a method for building a spatial semantic database from a BIM (building information modeling) model without damage, belonging to the technical field of spatial geographic information data conversion methods. The method comprises the steps of establishing a new table, defining a table structure field, defining the relationship between tables, and performing table record arrangement on a BIM object according to a defined database structure; finally, building a database to obtain a lossless BIM space semantic database; the invention can directly and completely reserve the attribute structure and the geometric structure data of the BIM, manage the BIM model in a space semantic database mode, and express the semantic relation between entities through the organizational structure of the database, thereby realizing the lossless introduction of the BIM model into the space information field and providing data support for the visual application field and the intelligent analysis field.

Description

Method for building spatial semantic database from BIM model without damage

Technical Field

The invention relates to a method for building a spatial semantic database from a BIM model without damage, belongs to the technical field of a method for assisting management of smart cities or digital twin cities by using computer technology, and particularly belongs to the technical field of a spatial geographic information data conversion method.

Background

With the development of scientific technology, objects and businesses in the physical world of human beings are being reorganized and reproduced in the computer world through a series of technical means: from early design drawings and two-dimensional electronic maps to present three-dimensional design models and three-dimensional digital maps, various technologies aim to realize mapping, simulation, monitoring and guidance of a physical world through an increasingly real and accurate three-dimensional digital world.

In the process of restoring the physical world to the digital world, two typical technologies become mainstream: one is a Building Information Modeling (BIM) technology oriented to the field of constructional engineering, and because of high model fineness and rich semantic Information including rich construction space and semantic Information such as geometry, physics and rules, the BIM technology is mainly applied to digital management of the whole life cycle of engineering or constructional engineering and is widely applied to various fields of constructional engineering. The GIS (Geographic Information System) technology is oriented to the Geographic space field, and can support the storage management analysis and application of large-scale, multi-scale and massive two-dimensional and three-dimensional Geographic space data from micro to macro based on Geographic space Information.

Since early BIM and GIS were directed to different fields of application, they adopted distinct data standards and techniques, with significant differences in both geometric representation and semantic description. However, with the development of smart cities, the demand for the integration of the smart cities is more and more strong: on one hand, the GIS technology is selected as a main technology for constructing a smart city due to the congenital advantages, and the BIM data is an important data source for city visualization, refinement and precise management; on one hand, users in the field of BIM increasingly need to know the environment around the model and the association relationship between multiple BIM models, which in turn needs support of GIS. BIM enters the GIS field, repeated modeling can be reduced, new application is excited, and basic data guarantee is provided for various applications such as city infrastructure, city planning management, public security emergency, indoor and outdoor integrated navigation and the like.

The currently available method for BIM to enter the GIS domain is generally data conversion, mainly developed around two common data standards of IFC and CityGML, including data level and application level (as shown in fig. 1).

The data level integration method mainly converts the existing data format, namely IFC to CityGML, and extracts geometric information, converts a coordinate system and geometric expression and maps semantic information through related algorithms; the work focuses on comparing the difference of data models of the BIM standard and the CityGML standard in an expression mechanism and researching the conversion from a refined model to a surface model; a method for extracting a multi-level of detail (LOD) GIS model from a BIM model.

The integration method of the application level is that only at the application program level, required information is extracted from BIM, converted into a specific data format and then read by GIS analysis software; based on visual programming Engine (FME), a semantic information filtering and geometric information extracting method is designed to realize conversion from a BIM model to a City GML model under LOD2 and LOD3 detail levels.

The existing data conversion method for BIM entering the GIS field has the following problems:

1. data loss exists in the fusion of the data level; firstly, converting any BIM software into IFC, and losing a part of data; on the other hand, the conversion method inevitably generates errors and information loss problems due to the difference of geometric expression and object semantics between the two data formats and the fact that the IFC model contains semantic information far exceeding the CityGML model.

2. The fusion of the application level only faces a specific application to extract conversion data, has very large limitation, can only solve a specific problem, and has weak expansibility.

For example, in the published chinese patent document "a method for merging BIM and 3DGIS for planning management (application number 201910206369.0)", a method for merging BIM and 3DGIS for planning management is proposed. The method mentions that BIM is utilized to construct a city planning information database. Before constructing the library, the BIM information is extracted and converted, and the constructed semantic database content is 'mapping relation between semantics and BIM classification information' and does not contain geometric structure information.

For example, in another published chinese patent document, "BIM data weight reduction method and apparatus (application No. 201710058771. X)", a method for reducing BIM data weight is proposed. The method comprises the steps of determining all components corresponding to any component type in BIM data according to any component type in the BIM data, traversing component attributes corresponding to all the components, setting the same component identification for the components with the same component attribute, and storing the component attribute corresponding to any component identification as a constructed model to a component model database according to any component identification. In the technical scheme, the database is not managed by utilizing a semantic association concept, and models with the same attribute are classified, so that simplification is realized, the original BIM tree structure is lost, and the relation between components cannot be seen. That is, it does not fully follow the organizational structure of BIM, completely retains all the geometric information and attribute information of all objects of BIM, and performs additional classification on the BIM objects.

For example, the published article "method and application of expanding BIM to spatial database" (author: the department of civil engineering of central university of cinnabar-Ping) proposes a BIM expansion method and application case combining spatial database, aiming at utilizing the field expansion, file processing and fast query capability of large amount of data with simple spatial relationship which are good in the original database technology, so that more software application programs can reuse BIM data and functions. However, the method only refers to the concept of geographic space, and does not relate to the concept of semantic association for managing the database.

Such as the tools proposed by the ESRI company to import the contents of one or more BIM file workspaces into a geographic database element dataset, BIMs may be created as geographic elements. The tool also only considers geospatial concepts and does not consider semantic associations.

Chinese patent document CN109918751A also discloses a building three-dimensional semantic modeling method based on CityGML extension, which classifies buildings and components thereof and extracts basic semantic entity objects; determining the type and attribute of the entity object; building a three-dimensional semantic model of a building based on the CityGML; based on semantic mapping and geometric transformation, high-quality conversion from a building IFC model to a CityGML expansion model is realized, and the CityGML expansion model with correct geometry, semantics and topology is quickly obtained, so that high-precision CityGML expansion model data of the building under the GIS environment are quickly generated; shape details of a three-dimensional object are expressed by describing the faces, edges, vertices, and relationships between them of the object. However, in this solution, three objects IFCSpace, ifcBuildingStore, ifccording in BIM are converted into a ctygml object Room. This merging operation involves, on the one hand, a reduction of the BIM object class and, on the other hand, an increase of the BIM records or, that is, the data is modified during the data conversion.

In summary, none of the prior art methods is able to achieve lossless data conversion and storage.

However, with the deep development of the spatial information field, a technology capable of realizing lossless data conversion and storage is urgently needed in more and more application scenarios.

Disclosure of Invention

In view of the technical problems in the prior art, the invention provides a method for building a spatial semantic database by a BIM model in a lossless manner, so that the geometric information, the attribute information and the texture information of the BIM are directly organized to express the semantic relationship among entities through the database, the attribute structure and the geometric structure data of the BIM are completely reserved, the BIM model is managed in a spatial semantic database manner, and the BIM model is introduced into the spatial information field in a lossless manner from the visual application field to the intelligent analysis field.

Specifically, in order to achieve the purpose, the technical scheme and the method adopted by the invention are as follows:

a method for building a spatial semantic database from a BIM model without loss comprises the following steps:

initially setting, selecting a relational database;

the method comprises the following steps: the number of the database tables is aligned with the types of the BIM objects, and how many tables are established for how many types of objects exist in the BIM;

establishing a new table: according to the organization specification IFC2x Edition 3 of BIM data proposed by the International organization for specialization of BIM building SMART and adopted by the International organization for standardization such as ISO and the like, a table is established for all object types in the BIM, and the table name is consistent with the type name in the BIM; the coverage of all object types in the BIM is realized;

step two: the database table structure is aligned with the attribute of the BIM object;

defining the table structure field: and constructing fields of a database table according to the attribute description fields of each type of BIM object: the field name and the type are consistent with the BIM object; setting the attribute of the BIM object as the field name of the table;

step three: the relation between the database table relation and the BIM object is aligned;

defining the relationship between tables: the unique ID of the BIM object is set as a primary key of each table. The parent ID of the BIM object is set as the foreign key of the table. The BIM tree structure is described by a relational database through the matching of a main key and an external key;

step four: the database table records are aligned with the records of the BIM objects, the BIM objects are put into a warehouse, and all the BIM objects are stored into corresponding tables according to classification to form the records of the tables;

performing table record arrangement on the BIM object according to the database structure defined above;

step five: executing a library building script, and building a library to obtain a lossless BIM space semantic database;

and (3) warehousing each class of objects and each attribute in the BIM, clearly expressing the relationship of each class of objects through the relationship among the tables, completely reserving all BIM object attributes, completely reserving all BIM object relationships, and converting the geometric expression of the BIM into the spatial expression to obtain the structure of the spatial semantic database which is the same as the BIM organizational structure.

The table records in the fourth step are arranged to respectively execute the geometric conversion operation and the semantic preservation operation.

The semantic retention operation specifically includes that according to a defined table structure, the semantic attributes of the BIM data are directly used as values of corresponding fields in the table structure to be input, and complete retention of the BIM semantic data is achieved.

The geometric transformation operation is to perform coordinate transformation on the geometric information of the BIM data, and the coordinate transformation includes transformation of a geometric expression mode and coordinate transformation to obtain geometric information of a geographic space.

The conversion of the geometric expression mode refers to the model expression of the BIM object, and the model expression usually converts a triangulation network into a geographic expression, mainly comprising geographic geometric elements of points, lines and surfaces; the geometric object of the BIM model adopts a model geometric expression mode, a triangular network is used for infinitely approximating the geometric shape, and the geometric expression mode of the BIM model needs to be converted into the expression of geographic element points, lines and surfaces.

The coordinate transformation is to transform the local coordinate system adopted by the BIM into geospatial coordinates.

The conversion steps of the geometric expression mode are as follows:

firstly, selecting adjacent triangles with parallel normal vectors to carry out triangular net merging; merging the triangulation networks by removing the shared edges of two adjacent triangles; after merging, judging whether the vertexes are in the same plane: if the vertexes are in the same plane, continuing to merge until no triangle is merged; this results in a polygon with continuous and closed vertices, which can be described by sequential combinations of point sequences, such as P = { P1, P2, P3 … … pn }, n = natural numbers of 1,2,3,4, … …; the step does not modify the structure and the number of the geometric objects of the BIM, but only changes the geometric expression, so that the geometric objects are expressed by a polygon coordinate string at present through the original expression of the triangle network.

The coordinate transformation comprises the following steps:

the local coordinate system adopted by BIM, in which the coordinates of the polygon point are also the local coordinates, is generally represented by (X, Y, Z), and is converted into a geospatial coordinate system, which is generally represented by (B, L, H), and the conversion formula between the two is as follows:

wherein: b is the latitude of the earth, L is the longitude of the earth, H is the height of the earth, N is a constant, and e is a natural constant.

The technical scheme and the method have the beneficial effects that:

the problem of BIM data loss is solved: the invention can completely reserve and store the BIM data in a warehouse, and solves the problems of BIM geometric data loss, semantic mapping error and the like caused by the traditional methods such as the traditional semantic mapping, geometric structure simplification and the like.

The problem of BIM data space query retrieval is solved: BIM data generally adopts local coordinates and cannot be subjected to spatial retrieval; the BIM data is constructed into the spatial semantic database, so that spatial topological relation expression can be realized among BIM models, and spatial retrieval and spatial topological analysis can be realized on the BIM data, such as query of all wall objects which are 30 meters away from the current visual angle.

Solving the joint retrieval among a plurality of BIM data: the original BIM data is a single file and cannot be jointly searched among a plurality of BIMs. By the method, a plurality of BIMs can be put into a warehouse to realize combined query and retrieval, such as searching and positioning fire hydrants in all BIMs in a city.

Solving semantic retrieval and expansion: the BIM expressed by the existing GIS method on the market is based on expression or visualization, and is a visualization parameter of a certain attribute, for example, a window is expressed by hooking some attributes through a central point coordinate, and the query is not an object of the window but a geometric center + hooked attributes of the window. None of these methods express real-world information. The semantic spatial database is characterized in that semantic information exists, knowledge in different fields can be integrated by semantic objects through a semantic description method, the objects not only contain attributes and capabilities of the objects, but also have spatial attributes and temporal information, and other information can be conveniently expanded and fused. The semantic objects are also associated with each other, so that the flattened data structure is changed into a three-dimensional data structure, and the value between the data is generated. Meanwhile, the semantic data structured description is an important way for the BIM data to enter intelligent analysis.

The problem of indoor and outdoor coordinate reference unification is solved: the outdoor environment expressed by the GIS and the indoor environment expressed by the BIM adopt different coordinate references. The method can convert the coordinate reference of the BIM into geographic space reference which can be identified by the GIS, realizes the unification of indoor and outdoor coordinate reference, and provides a new solution/solution idea/data support for indoor and outdoor integrated navigation.

The problem of difficulty in editing BIM data is solved: BIM data editing generally needs to be performed by means of professional BIM software, the problem that BIM is difficult to edit in the GIS field can be solved through the method, and batch addition, deletion, modification and searching of BIM models can be achieved. Such as: the doors of the two rooms are 2.1 meters wide originally, one of the two rooms needs to be changed into 1.8 meters, only batch modification in a form is needed, and then refreshing is carried out, so that the model can be modified accordingly. Meanwhile, because of semantic description, if newly added information exists in the later period, for example, a smoke sensor is newly installed in a building, the information can be expanded in a database according to rules without editing BIM data again.

Drawings

FIG. 1 is a schematic perspective view of a conventional data conversion method;

FIG. 2 is a schematic perspective view of data conversion according to the method of the present invention;

FIG. 3 is a schematic diagram of the overall structure of data conversion according to the method of the present invention;

FIG. 4 is a schematic diagram of a data conversion structure according to the present invention;

FIG. 5a is a block diagram of the geometry transformation method of FIG. 4;

FIG. 5b is a block diagram of the geometry expression transformation method of FIG. 5 a;

FIG. 5c is a schematic diagram showing a comparison between before and after conversion of the geometric representation of FIG. 5 b;

FIG. 6 is a schematic representation of the expression pattern in BIM according to an embodiment of the present invention;

FIG. 7 is a block diagram of a geometric representation transformation method according to an embodiment of the present invention;

fig. 8 is a schematic diagram of a data structure of a spatial semantic database constructed according to an embodiment of the present invention.

Detailed Description

The following detailed description of the method of the present invention with reference to the drawings and examples is provided to facilitate a thorough understanding of the method of the present invention.

As shown in fig. 1, fig. 2, fig. 3, fig. 4, fig. 5a, fig. 5b, fig. 5c, fig. 6, fig. 7 and fig. 8.

Referring to fig. 2 and fig. 3, a lossless method for constructing a spatial semantic database by a BIM model includes the following steps:

initial setting, selecting a relational database (for example, selecting MySQL and PostgreSQL);

the method comprises the following steps: the number of the database tables is aligned with the types of the BIM objects, and how many tables are established for how many types of objects exist in the BIM; and reading the native BIM data and the geometric object and attribute information of the BIM through the analysis parameters provided by the BIM manufacturer.

step three: the database table relation is aligned with the relation between the BIM objects, and the ID of the BIM objects is set as the primary key of the table; setting the father node ID of the BIM object as a foreign key of the table;

performing table record arrangement on the BIM object according to the database structure defined above; the table records are sorted to perform a geometric transformation operation and a semantic preservation operation, respectively.

and (3) warehousing each class of objects and each attribute in the BIM, clearly expressing the relationship of each class of objects through the relationship among the tables, completely reserving all BIM objects, completely reserving all BIM object attributes, completely reserving all BIM object relationships, converting the geometric expression of the BIM into a spatial expression, and obtaining a structure of a spatial semantic database which is the same as the BIM organizational structure, as shown in FIG. 4.

As shown in fig. 5a and fig. 5b, the geometric transformation operation is to perform coordinate transformation on the geometric information of the BIM data, and includes transformation of geometric expression and coordinate transformation to obtain geometric information of the geographic space.

The conversion of the geometric expression mode refers to the model expression of the BIM object, usually the triangulation network is converted into the geographic expression, mainly the geographic geometric elements of points, lines and surfaces; the geometric object of the BIM model adopts a model geometric expression mode, and a triangulation network is used to approach the geometric shape infinitely, and the geometric expression mode of the BIM model needs to be converted into expressions of geographic element points, lines and surfaces (as shown in fig. 5b and fig. 5 c).

The coordinate transformation refers to transforming the independent local coordinate system adopted by the BIM into geospatial coordinates.

The conversion steps of the geometric expression mode are as follows:

firstly, selecting adjacent triangles with parallel normal vectors to carry out triangular net merging; merging the triangulation networks by removing the shared edges of two adjacent triangles; after merging, judging whether the vertexes are in the same plane: if the vertexes are in the same plane, continuing to merge until no triangle is merged; this results in a polygon with continuous vertices, which can be described by sequential combinations of point sequences, such as P = { P1, P2, P3 … … pn }, n = natural numbers of 1,2,3,4, … …; the step does not modify the structure and the number of the geometric objects of the BIM, but only changes the geometric expression, so that the geometric objects are expressed by a polygon coordinate string at present through the original expression of the triangle network. (as shown in fig. 5 c).

The coordinate transformation comprises the following steps:

the local coordinate system adopted by BIM, in which the coordinates of the point of the polygon are also the local coordinates, is generally represented by (X, Y, Z), and is converted into a geographic coordinate system, which is generally represented by (B, L, H), and the conversion formula between the two is as follows:

The geodetic coordinate system (geodetic coordinate system) is a coordinate system established by taking a reference ellipsoid as a datum plane in geodetic surveying. The location of the ground point is represented by the geodetic longitude, the geodetic latitude, and the geodetic altitude. The determination of the geodetic coordinate system comprises selecting an ellipsoid, locating the ellipsoid and determining geodetic data. The geodetic latitude, geodetic longitude and geodetic height are each represented by the capital english letter B, L, H. The GPS geodetic coordinate system (B, L, H) is in relation with a ground rectangular coordinate system (X, Y, Z), the geodetic longitude L is a dihedral angle formed by a geodetic initial meridian plane and a meridian plane where the point is located, the east is positive from the initial meridian plane, the east is called east longitude (0-180), the west is negative, and the west is called west longitude (0-180); the geodetic latitude B is an included angle between a normal line passing through the point as an ellipsoid and an equatorial plane, and is counted from the equatorial plane, north latitude (0-90) is positive, south latitude (0-90) is negative; the geodetic height H is the distance from the ground point to the ellipsoid along the normal to the ellipsoid.

The method of the present invention proposes a method for storing and managing BIM through a spatial semantic database.

The method directly inputs the geometric information, the attribute information and the texture information of the BIM into the relational geographic database, so that information loss is avoided. Firstly, establishing a table for each type of BIM object through a fixed Schema, wherein the field design and the content of the table are from completely reserving semantic attribute information in the BIM; secondly, giving a uniform geographic space coordinate to the BIM object, establishing an incidence relation between tables according to a tree structure of the BIM, and realizing the expression of the semantic relation between BIM object entities through the organization of a database, thereby realizing the complete and lossless expression of the BIM by using a spatial semantic database.

Through the spatial semantic database, BIM can be accessed into GIS application without damage: by accessing the spatial semantic database, the GIS user can realize spatial query retrieval of single or multiple BIM models in the database on one hand, and can also perform semantic level search and mining on the other hand to perform semantic association search on attributes of single or multiple BIM objects in the database. Therefore, the expansion of the BIM data from the visual application field to the intelligent analysis field is realized.

As shown in fig. 6, fig. 7 and fig. 8 show a specific embodiment applying the method of the present invention, and a typical BIM file is taken as an example to perform detailed description of the implementation steps.

A typical building object, whose expression in BIM is shown in fig. 6, is divided into: project-construction site-building-floor, the wall, thick plate and roof are arranged under the floor. Each type of object in turn has a respective attribute element. For example, buildings, there are construction specific elements, general elements and space geometry elements.

The first step is as follows: alignment of database table number with type of BIM object

Establishing a new table: establishing a table for all object types in the BIM, wherein the table name is consistent with the type name in the BIM; for the BIM object, the following steps are required to be created: project, Site, Building floor, Wall, Roof, Slab and the like, and all object types are tabulated to cover all object types in the BIM.

The second step is that: attribute alignment of database table structures to BIM objects

Defining the table structure field: and constructing fields of a database table according to the attribute description fields of each type of BIM object: the field name and type are consistent with the BIM. For the BIM object, each object has general attributes such as unique Identification (ID), class (class), name (name) and the like; also have attributes specific to its own category, such as Building having a Building year (year _ of _ construction), number of floors (number _ of _ storeys); and also comprises that the position geometric attributes (geometry) establish corresponding fields.

The third step: alignment of relationships between database table relationships and BIM objects

Defining the relationship between tables: the unique ID of the BIM object is set as a primary key of each table. The parent ID of the BIM object is set as the foreign key of the table. The BIM tree structure is described by a relational database through the matching of a main key and an external key; for the BIM object, giving a unique Identifier (ID) of each type of object as a primary key of the type of object table; a newly added field for storing its parent node ID (unique identification); for example, in this embodiment, building store (floor) is added with a parent node building id (building id).

The fourth step: record alignment of database table records with BIM objects

As shown in fig. 7, table record sorting: carrying out BIM data warehousing on the BIM object according to the database structure defined above, wherein the BIM data warehousing comprises geometric conversion and semantic preservation; aiming at the BIM model, firstly carrying out triangulation network merging, constructing geographic element expression and coordinate transformation on a specific BIM object such as a wall surface (wall), realizing the transformation of BIM geometric attributes (geometry), and sequentially putting geometric attribute values into a geometric attribute table to be constructed; and placing the geographical element expression into a corresponding object table.

Then other semantic attribute data of the BIM are directly put in storage, and response fields in the corresponding table are recorded. Complete preservation of BIM semantic data is realized.

Finally, each type of object and each attribute in the BIM are put into a warehouse, the relationship of each type of object is clearly expressed through the relationship between tables, so that the nondestructive retention of BIM data is realized, a spatial semantic database the same as the BIM organizational structure is obtained, and the expression mode after the database is built is shown in FIG. 8.

If there are multiple BIM models, repeat the above steps.

On the management level, after the BIM model is entered into the spatial database, all buildings in the park can be uniformly managed, and spatial query and attribute query are performed, for example: and querying objects with the window area larger than x square meters in the park. All objects that meet the conditions can be queried. Compared with the traditional BIM management mode, the method can only inquire the interior of a single BIM, and has great improvement.

At the editing level, all data can be modified rapidly in batches, including geometric data and attribute data. For example: the doors of the two rooms are 2.1 meters wide originally, one of the two rooms needs to be changed into 1.8 meters, only batch modification and refreshing are needed in a form, and then the model can be modified. If new data are added, the database file can be directly edited to realize the new objects.

In the indoor and outdoor navigation aspect, the BIM data and the geographic information data are unified into the same coordinate reference frame by warehousing the BIM data, so that the indoor and outdoor integrated navigation can be realized, and the navigation in a park is easily realized.

On the semantic mining level, a spatial semantic database is constructed by the BIM data, spatial attributes are superposed on the BIM data, fusion data based on a space-time model is achieved, and further semantic mining can be performed on the BIM data. For example, fire hydrants produced by the same manufacturer and having a maintenance period of more than half a year and a location within a radius of 10 kilometers from a fire station can be queried, and the requirement that the space area of the fire hydrant is less than 50 square meters can be met. Like this, where there is a need for a comprehensive search of spatial topology (radius), attributes (maintenance period, manufacturer), and the geometric objects themselves (floor space), only a spatial semantic database can support it.

In the aspect of urban intelligent analysis. Because the semantic model is structured description, the geometric attributes (such as house inclination angle, wall area and window size) of the model are in the description of the existing model, so that the model can be read by various analysis and simulation software, and the BIM model can really enter the analysis field instead of staying in the visual expression field.

Claims

1. A method for lossless construction of a spatial semantic database from a BIM model, comprising the steps of:

initially setting, selecting a relational database;

step one, the number of database tables is aligned with the type of BIM objects, and how many tables are established according to how many types of objects exist in the BIM;

establishing a new table: establishing a table according to the organization specification of the BIM data, wherein the table name is consistent with the category name in the BIM; the coverage of all object types in the BIM is realized;

defining the relationship between tables: setting the unique ID of the BIM object as a primary key of each table, setting the ID of a father node of the BIM object as an external key of the table, and realizing the description of the BIM tree structure by a relational database through the cooperation of the primary key and the external key;

performing table record arrangement on the BIM object according to the database structure;

each class of object and each attribute in the BIM are put into a warehouse, the relationship of each class of object is clearly expressed through the relationship between tables, all BIM object attributes are completely reserved, all BIM object relationships are completely reserved, the geometric expression of the BIM is converted into spatial expression, and the structure of a spatial semantic database which is the same as the BIM organization architecture is obtained;

the table records in the step four are sorted into a geometric conversion operation and a semantic preservation operation which are respectively executed;

the geometric transformation operation is to perform coordinate transformation on the geometric information of the BIM data, and the coordinate transformation comprises the transformation of a geometric expression mode and the coordinate transformation to obtain the geometric information of a geographic space;

the conversion of the geometric expression mode refers to the model expression of the BIM object, and the model expression is the conversion of a triangulation network into a geographic expression to express geographic geometric elements of points, lines and surfaces; the geometric object of the BIM model adopts a model geometric expression mode, a triangular network is used for infinitely approximating the geometric shape, the geometric expression mode of the BIM model is converted, and the geometric expression mode is converted into the expression of geographic element points, lines and surfaces;

the conversion steps of the geometric expression mode are as follows:

firstly, selecting adjacent triangles with parallel normal vectors to carry out triangular net merging; merging the triangulation networks by removing the shared edges of two adjacent triangles; after merging, judging whether the vertexes are in the same plane: if the vertexes are in the same plane, continuing to merge until no triangle is merged; this results in a polygon with continuous, closed vertices, which is described by sequential combinations of point sequences: p ═ P1, P2, P3 … … pn, n ═ 1,2,3,4, … … natural numbers; the step does not modify the structure and the quantity of the geometric objects of the BIM, only changes the geometric expression, so that the geometric objects are expressed by a polygon coordinate string at present through the original expression of the triangulation.

2. The method according to claim 1, wherein the semantic preserving operation is specifically to directly input the semantic attributes of the BIM data as the values of the corresponding fields in the table structure according to the defined table structure, so as to realize complete preservation of the BIM semantic data.

3. The method of claim 1, wherein the coordinate transformation is performed by transforming a local coordinate system adopted by the BIM into geospatial coordinates.

4. The method of claim 3, wherein the step of coordinate transformation comprises:

the local coordinate system adopted by the BIM is represented by (X, Y, Z), the local coordinate system is converted into a geographic space coordinate system represented by (B, L, H), and the conversion formula between the local coordinate system and the geographic space coordinate system is as follows:

5. The method of claim 1, wherein the method can realize BIM lossless access to GIS application through spatial semantic database: GIS users can realize the spatial query retrieval of single or multiple BIM models in the database by accessing a spatial semantic database, and can also perform semantic level search and mining to perform semantic association search on the attributes of single or multiple BIM objects in the database; therefore, the expansion of the BIM data from the visual application field to the intelligent analysis field is realized.