Semantic unit determination method and device for large building model, equipment and medium

Abstract

The invention discloses a building large model-oriented semantic unit determination method and device, equipment and a medium. The method comprises the steps of performing data preprocessing on a basic information data set based on a preset information dictionary to obtain a target information data set meeting a preset format requirement; determining a corresponding target semantic structure in a preset semantic structure set based on a target field name of target information data in the target information data set, and determining a target semantic unit corresponding to the target information data based on the target semantic structure; and performing data arrangement on each target semantic unit based on the data prediction demand to obtain a basic semantic unit sequence corresponding to the basic information data set. Through the technical scheme of the invention, the semantic unit oriented to the information data of the large building model can be determined, so that the semantic unit fully expresses professional knowledge and complex relationships in the building field, a foundation is laid for application of the large model in the building industry, and the accuracy of a prediction result of the large building model is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a semantic unit determination method, device, equipment and medium for a large building model. Background Art

[0002] With the development of the digital age, large model technologies represented by the Generative Pre-trained Transformer (GPT) model have gradually demonstrated powerful semantic understanding and knowledge transfer capabilities in the construction industry. Similarly, they have also shown great potential in areas such as energy consumption prediction, intelligent control, and equipment and facility operation and maintenance.

[0003] However, the training of large models depends on the accuracy of the tokenization method. Tokenization is the process of dividing text data into discrete and meaningful semantic units (tokens), which is the basis for large models to understand knowledge. Therefore, the training of large models is very important.

[0004] The tokenization methods in the existing technology are mainly aimed at natural language generation scenarios, which are difficult to fully express the professional knowledge in the field of architecture, and cannot effectively handle the temporal dynamics and spatial relationships in the field of architecture. Therefore, how to perform word segmentation processing on the information data for large architectural models, obtain semantic units for large architectural models that can fully express the professional knowledge and complex relationships in the field of architecture, lay the foundation for the application of large models in the construction industry, and improve the accuracy of the prediction results of large architectural models, is a problem that needs to be solved urgently. Summary of the invention

[0005] The present invention provides a method, device, equipment and medium for determining semantic units for large building models, which can solve the problem that the semantic units obtained by the tokenization method in the prior art cannot fully express the professional knowledge and complex relationships in the field of architecture.

[0006] According to one aspect of the present invention, a method for determining semantic units for a large building model is provided, comprising:

[0007] Obtaining a basic information data set corresponding to the current prediction scenario and a current training requirement; wherein the current training requirement includes a data prediction requirement;

[0008] Performing data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; wherein the target information data set includes each target information data;

[0009] Determining a corresponding target semantic structure in a preset semantic structure set based on a target field name of the target information data, and determining a target semantic unit corresponding to the target information data based on the target semantic structure;

[0010] The data of each target semantic unit is arranged based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

[0011] According to another aspect of the present invention, there is provided a semantic unit determination device for a large building model, comprising:

[0012] A data acquisition module is used to acquire a basic information data set corresponding to the current prediction scenario and a current training requirement; wherein the current training requirement includes a data prediction requirement;

[0013] A preprocessing module, used for performing data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; wherein the target information data set includes each target information data;

[0014] A semantic unit determination module, configured to determine a corresponding target semantic structure in a preset semantic structure set based on a target field name of the target information data, and determine a target semantic unit corresponding to the target information data based on the target semantic structure;

[0015] The data arrangement module is used to arrange the data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] at least one processor; and

[0018] a memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the semantic unit determination method for a large building model described in any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for determining semantic units for a large building model as described in any embodiment of the present invention when executed.

[0021] According to another aspect of the present invention, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the method for determining semantic units for a large building model according to any embodiment of the present invention is implemented.

[0022] The technical solution of the embodiment of the present invention obtains a target information data set that meets the preset format requirements by preprocessing the basic information data set based on a preset information dictionary; then, the corresponding target semantic structure is determined in the preset semantic structure set based on the target field name of the target information data in the target information data set, and the target semantic unit corresponding to the target information data is determined based on the target semantic structure; finally, the data of each target semantic unit is arranged based on the data prediction requirements to obtain a basic semantic unit sequence corresponding to the basic information data set. Since the semantic units are divided by using the pre-constructed information dictionary and semantic structure set of the construction industry, the problem that the semantic units obtained by the tokenization method in the prior art cannot fully express the professional knowledge and complex relationships in the construction field is solved, and the information data for the large model of the building can be segmented to obtain the semantic units for the large model of the building that can fully express the professional knowledge and complex relationships in the construction field, laying the foundation for the application of large models in the construction industry and improving the accuracy of the prediction results of large models of the building.

[0023] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0025] Figure 1 is a flow chart of a semantic unit determination method for a large building model provided according to Embodiment 1 of the present invention;

[0026] Figure 2 is a flow chart of a semantic unit determination method for a large building model provided according to a second embodiment of the present invention;

[0027] Figure 3 2 is a schematic diagram of a structure of a semantic unit determination device for a large building model provided according to a third embodiment of the present invention;

[0028] Figure 4It is a structural schematic diagram of an electronic device for implementing the semantic unit determination method for a large building model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "base", "target", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] Embodiment 1

[0032] Figure 1 A flowchart of a method for determining semantic units for large building models provided in the first embodiment of the present invention. This embodiment can be applied to the word segmentation of information data for large building models to obtain semantic units for large building models that can fully express professional knowledge and complex relationships in the field of architecture. The method can be executed by a semantic unit determination device for large building models. The semantic unit determination device for large building models can be implemented in the form of hardware and / or software. The semantic unit determination device for large building models can be configured in an electronic device. Figure 1 As shown, the method includes:

[0033] S110, obtaining a basic information data set corresponding to a current prediction scenario and a current training requirement; wherein the current training requirement data predicts the requirement.

[0034] Among them, the current prediction scene may refer to the scene that needs to be predicted by the large model data at the current moment. Exemplarily, the current prediction scene may be the name of the building, or the location of the building, etc. Information data may refer to information related to the building field. Exemplarily, the information data may be floors, air-conditioning equipment, power values, or connectivity relationships, etc. Basic information data may refer to text data containing initial building information. For example, unprocessed descriptive text data containing building information. Exemplarily, the basic information data may be that the windows are square, there are 10 floors, and the height of each floor is 3 meters, etc. The basic information data set may refer to a set of basic information data corresponding to the same current prediction scene.

[0035] Among them, the data prediction requirement may refer to the data input order requirement in the prediction process. For example, the data prediction requirement may be to first input the number of floors, then the temperature of each floor, and finally the air conditioning equipment activation status of each floor. The current training requirement may refer to the various condition requirements corresponding to the current prediction scenario. For example, the current training requirement may include semantic enhancement requirements and data prediction requirements.

[0036] S120, preprocessing the basic information data set based on a preset information dictionary to obtain a target information data set that meets preset format requirements; wherein the target information data set includes each target information data.

[0037] The preset information dictionary may refer to a pre-built standard definition dictionary corresponding to the construction field. Generally, the preset information dictionary may contain standard definitions of all common construction information in the construction field. The preset format requirement may refer to a preset format judgment requirement. Exemplarily, the preset format requirement may be that there is no duplicate data, no missing values, and no abnormal values. The target information data may refer to information data that meets the preset format requirements. The target information data set may refer to a set composed of various target information data under the same current prediction scenario.

[0038] In an optional embodiment, before the basic information data set is preprocessed based on the preset information dictionary to obtain a target information data set that meets the preset format requirements, it may also include: acquiring each historical information data in the historical information database; determining the preset definition standard corresponding to the historical information data based on the basic field name of the historical information data; determining the target definition information corresponding to the historical information data based on the preset definition standard; and combining and processing the target definition information corresponding to each historical information data in the historical information database to generate a preset information dictionary.

[0039] Among them, historical information data may refer to various information data that appeared in a historical time period. For example, it may be information data that appeared in a historical year. The historical information database may refer to a database that stores historical information data. The field name may refer to a specific field representation corresponding to the information data. For example, if the information data is that the power of the cooler is 500 watts (w), the field name may be cooler or power. The basic field name may refer to the field name corresponding to the historical information data.

[0040] Among them, the preset definition standard may refer to the definition requirements of the information data set in advance. Exemplarily, the definition standard classification may be performed according to the field category of the field name. The field category may refer to the classification category corresponding to the field name. Exemplarily, the field category may be an object category including building space class objects and equipment and facility class objects, an attribute category including static attributes and dynamic attributes, and a relationship category including the relationship types between entity objects. Specifically, if the field category is an object category, the preset definition standard may be: assigning a unique code to each entity object; if the field category is an attribute category, the preset definition standard may be: defining the data type, unit and unique code for each attribute; if the field category is a relationship category, the preset definition standard may be: assigning a unique code to each relationship type. The definition information may refer to the standard definition at the semantic level generated according to the preset definition standard. The target definition information may refer to the definition information corresponding to the historical information data.

[0041] Specifically, before preprocessing the basic information data set based on the preset information dictionary to obtain the target information data set that meets the preset format requirements, the independent historical information data of each data can be determined in the historical information database. Thereafter, the preset definition standard corresponding to the historical information data can be determined using the field category of the basic field name in the historical information data, and the target definition information corresponding to the historical information data can be determined based on the preset definition standard. Exemplarily, if the basic field name of the historical information data is a building space class object, such as a floor, space, or shaft, or an equipment and facility class object, such as an air-conditioning unit or an alarm valve group, then "assigning a unique code to each entity object" is used as the preset definition standard, and a unique code is assigned to the entity object in the historical information data according to the preset definition standard as the target definition information; if the basic field name of the historical information data is a static attribute, such as rated power or brand, or a dynamic attribute, such as instantaneous air supply volume or operating status, then "defining data types, units, and unique codes for each attribute" is used as the preset definition standard, and data types such as integer, floating point, or Boolean are defined for the static attributes or dynamic attributes in the historical information data according to the preset definition standard, and units are defined, unique codes are assigned, and timestamps are attached to dynamic attributes, etc., as the target definition information; if the basic field name of the historical information data is the relationship type between entity objects, then "assigning a unique code to each relationship type" is used as the preset definition standard, and a unique code is assigned to the relationship type between entity objects in the historical information data according to the preset definition standard, such as the connectivity relationship between spaces or the service relationship between equipment and spaces, as the target definition information. Finally, the target definition information corresponding to each historical information data in the historical information database is combined and processed to generate a preset information dictionary. Therefore, the preset information dictionary can be pre-built through the field categories of each basic field name, providing a standardized data structure framework, and making standard definitions at the semantic level for the building information involved in the building, ensuring the uniqueness and traceability of each basic field name in the data processing and exchange process, improving the interoperability of the data, avoiding ambiguity when using data in various scenarios, and providing an effective basis for subsequent operations.

[0042] It is worth noting that in the embodiment of the present invention, the same historical information data may correspond to multiple target definition information. For example, if the historical information data is "each floor is 3 meters high", the target definition information may be the definition information corresponding to the entity object "floor" or the definition information corresponding to the static attribute "3 meters high". In addition, the unique codes corresponding to the various definition information are independent of each other to ensure the uniqueness of the data, and the embodiment of the present invention does not make additional elaborations on this.

[0043] S130: Determine a corresponding target semantic structure in a preset semantic structure set based on the target field name of the target information data, and determine a target semantic unit corresponding to the target information data based on the target semantic structure.

[0044] Among them, the target field name may refer to the field name corresponding to the target information data. The preset semantic structure may refer to the preset information data word segmentation structure. The preset semantic structure set may refer to the set composed of various preset semantic structures. The target semantic structure may refer to the semantic structure corresponding to the target field name. The semantic unit may refer to the word segmentation result generated according to the semantic structure. Generally, the semantic unit can be the word segmentation result of the field name filled in according to the semantic structure. The target semantic unit may refer to the semantic unit corresponding to the target information data.

[0045] In an optional embodiment, before determining the corresponding target semantic structure in the preset semantic structure set based on the target field name of the target information data, it may also include: obtaining the basic field name corresponding to the full amount of historical information data; determining the full amount of basic semantic fields and basic semantic order corresponding to the basic field name based on the field category of the basic field name; combining and processing each basic semantic field based on the basic semantic order to generate the preset semantic structure corresponding to the basic field name; combining and processing the preset semantic structures corresponding to the full amount of basic field names to generate a preset semantic structure set.

[0046] Among them, the basic semantic field can refer to the field representation of the semantic level corresponding to the basic field name. Generally, the basic semantic field corresponding to the basic field name can be determined according to the field category of the basic field name. Exemplarily, if the field category of the basic field name is an object category, the basic semantic field can be an object type code and an object instance code, wherein the object type code can represent the unique code of the entity object, and the object instance code can represent the specific internal code of the entity object under the same object type. For example: if the basic field name is a condenser, the object type code can be an air conditioning unit code, and the object instance code can be a condenser 1 code; if the field category of the basic field name is an attribute category, the basic semantic field can be an object instance code, an attribute code, and an attribute value, wherein the attribute code can represent the unique code of the attribute type, and the attribute value can represent the specific value of the attribute type. For example: if the basic field is a brand, the object instance code can be a condenser 1 code, the attribute code can be a brand code, and the attribute value can be a specific brand; if the field category of the basic field name is a relationship category, the basic semantic field can be an object instance code 1, a relationship code, and an object instance code 2, wherein the relationship code can represent the unique code of the relationship type. For example, if the basic field is a connectivity relationship, the object instance code 1 may be a condenser 1 code, the relationship code may be a connection relationship code, and the object instance code 2 may be a pump 1 code.

[0047] Among them, the basic semantic order may refer to the field arrangement order of each basic semantic field under the same field category. Usually, the basic semantic order of each basic semantic field corresponding to the basic field name can be determined according to the field category of the basic field name. Exemplarily, if the field category of the basic field name is an object category, the basic semantic order may be: [object type code]-[object instance code]; if the field category of the basic field name is an attribute category, the basic semantic order may be: [object instance code]-[attribute code]-[attribute value]. If the field category of the basic field name is a relationship category, the basic semantic order may be: [object instance code 1]-[relationship code]-[object instance code 2].

[0048] Specifically, before determining the corresponding target semantic structure in the preset semantic structure set based on the target field name of the target information data, all the basic field names corresponding to the historical information data can be determined first, and then all the basic semantic fields and basic semantic order corresponding to each basic field name can be determined according to the field category of the basic field name, and the corresponding basic semantic fields are filled in according to the basic semantic order to obtain the preset semantic structure corresponding to the basic field name, and finally, the preset semantic structures corresponding to all the basic field names are combined to generate a preset semantic structure set. Therefore, in view of the specific needs of the construction field, a professional tokenization method is proposed, which can pre-construct a preset semantic structure set through the field category of each basic field name, which can more accurately capture the professional knowledge and complex relationships in the construction field and provide an effective basis for subsequent operations.

[0049] It is worth noting that in the embodiment of the present invention, the preset information dictionary and the preset semantic structure set only need to be constructed once, and subsequent operations can directly use the constructed preset information dictionary and preset semantic structure set to determine the semantic unit, and the embodiment of the present invention does not make additional elaborations on this.

[0050] S140 , arranging data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

[0051] The data arrangement may refer to the operation of sorting the target semantic units corresponding to the same basic information data set. The basic semantic unit sequence may refer to the sequence obtained by arranging the target semantic units corresponding to the same basic information data set.

[0052] The technical solution of the embodiment of the present invention obtains a target information data set that meets the preset format requirements by preprocessing the basic information data set based on a preset information dictionary; then, the corresponding target semantic structure is determined in the preset semantic structure set based on the target field name of the target information data in the target information data set, and the target semantic unit corresponding to the target information data is determined based on the target semantic structure; finally, the data of each target semantic unit is arranged based on the data prediction requirements to obtain a basic semantic unit sequence corresponding to the basic information data set. Since the semantic units are divided by using the pre-constructed information dictionary and semantic structure set of the construction industry, the problem that the semantic units obtained by the tokenization method in the prior art cannot fully express the professional knowledge and complex relationships in the construction field is solved, and the information data for the large model of the building can be segmented to obtain the semantic units for the large model of the building that can fully express the professional knowledge and complex relationships in the construction field, laying the foundation for the application of large models in the construction industry and improving the accuracy of the prediction results of large models of the building.

[0053] Embodiment 2

[0054] Figure 2 A flowchart of a semantic unit determination method for a large building model is provided in the second embodiment of the present invention. This embodiment is refined based on the above embodiment. In this embodiment, the operation of preprocessing the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements is specifically refined. Specifically, it may include: format conversion of each basic information data in the basic information data set based on the preset information dictionary to obtain a candidate information data set; format judgment of each candidate information data in the candidate information data set based on the preset format requirements to obtain a target information data set that meets the preset format requirements. Figure 2 As shown, the method includes:

[0055] S210: Obtain a basic information data set corresponding to the current prediction scenario and current training requirements.

[0056] Wherein, the current training demand includes data prediction demand.

[0057] In an optional embodiment, the current training requirements may also include: semantic enhancement requirements. The semantic enhancement requirements may refer to the setting requirements of semantically enhanced tags set by the user according to the current actual scenario. Exemplarily, the semantic enhancement requirements may include unit setting requirements, metadata setting requirements, and timestamp setting requirements. By introducing a semantically enhanced tagging mechanism, metadata such as units, data types, numerical ranges, and timestamps can be embedded in the token, thereby improving the interpretability and application value of building data.

[0058] Specifically, the user can input the basic information data set and current training requirements corresponding to the current prediction scenario according to actual application requirements. For example, the input can be made through the keyboard and mouse, or through the drop-down option. Then, the basic information data set and current training requirements corresponding to the current prediction scenario can be obtained to provide a data basis for subsequent operations.

[0059] It is worth noting that in the embodiment of the present invention, the basic information data set may come from different sources, such as Building Information Modeling (BIM) software, sensor data or manual input, etc., and the embodiment of the present invention does not specifically limit this.

[0060] S220: Convert the format of each basic information data in the basic information data set based on a preset information dictionary to obtain a candidate information data set.

[0061] The information data to be selected may refer to information data that has only completed the format conversion operation. Exemplarily, it may be information data that has been encoded according to a preset information dictionary. Specifically, if the basic information data is a condensing machine, and the object type code of the condensing machine specified in the preset information dictionary is an air conditioning unit code, the air conditioning unit code may be used as the information data to be selected corresponding to the basic information data. The information data set to be selected may refer to a set of information data to be selected corresponding to the same basic information data set.

[0062] Specifically, after obtaining the basic information data set corresponding to the current prediction scenario, the preset information dictionary can be used to convert the format of each basic information data in the basic information data set to obtain the selected information data corresponding to each basic information data. For example, if the basic information data is the brand of the condensing machine is brand A, the preset information dictionary pre-defines the object type code of the condensing machine as "O:ACATCC", the object instance code as "Chiller1", and the attribute code as "A:brand", then the basic information data can be formatted according to the preset information dictionary to obtain the selected information data containing the fields "O:ACATCC", "Chiller1", "A:brand" and "A brand". Then, the selected information data corresponding to each basic information data are combined to obtain the selected information data set. Therefore, by processing the basic information data set according to the definition of the preset information dictionary, it can be ensured that the selected information data after format conversion meets the predefined format standard, providing an effective basis for subsequent operations.

[0063] S230: Perform format judgment on each of the candidate information data sets in the candidate information data set based on preset format requirements to obtain a target information data set that meets the preset format requirements.

[0064] The target information data set includes various target information data.

[0065] The format judgment may refer to the operation of screening the data content of the selected information data. Exemplarily, the format judgment may include judging whether there are repeated selected information data, judging whether there are missing values ​​in the selected information data, judging whether there are abnormal values ​​in the selected information data, etc.

[0066] Specifically, after converting the format of each basic information data in the basic information data set to obtain the selected information data set, data cleaning technology can be used to judge the format of each candidate information data in the selected information data set, process missing values, abnormal values ​​and duplicate data, and obtain a target information data set that meets the preset format requirements to improve data quality.

[0067] S240: Determine a corresponding target semantic structure in a preset semantic structure set based on the target field name of the target information data.

[0068] Specifically, after determining the target information data set, the target field name of the target information data can be used to determine a semantic structure consistent with the field category of the target field name in the preset semantic structure set as the target semantic structure corresponding to the target information data. For example, if the target field name is condensing machine, it can be determined that the field category corresponding to the target field name is an object category, and then the semantic structure "[object type code]-[object instance code]" in the preset semantic structure set can be used as the target semantic structure.

[0069] S250: Determine the structural category of the target semantic structure. If the structural category is an attribute category, determine a semantic enhancement sequence corresponding to the target information data based on the semantic enhancement requirement.

[0070] Among them, the structure category may refer to the structure type corresponding to the semantic structure. Exemplarily, the structure category may be an object category, an attribute category, and a relationship category. The semantic enhancement sequence may refer to a semantic field related to the semantic enhancement requirement. Generally, the semantic enhancement sequence needs to be matched according to the semantic enhancement requirement. Specifically, if the semantic enhancement requirement is a unit setting requirement, the semantic enhancement sequence may be: [unit value] and [unit symbol]. For example,

[500] and [W]. If the semantic enhancement requirement is a metadata setting requirement, the semantic enhancement sequence may be: [data type] and [data range]. For example, [META:dataType=float] and [META:range=0-100]. If the semantic enhancement requirement is a timestamp setting requirement, the semantic enhancement sequence may be: [time range]. For example, [T:2024-10-07T14:00:00].

[0071] Specifically, after determining the target semantic structure corresponding to the target information data, the structural category of the target semantic structure can be judged. If the target semantic structure is an attribute category, the semantic enhancement sequence corresponding to the target information data can be determined based on the semantic enhancement requirements.

[0072] S260: Determine a target semantic unit corresponding to the target information data based on the semantic enhancement sequence and the target semantic structure.

[0073] Specifically, after determining the semantic enhancement sequence, the semantic enhancement sequence can be filled into the target semantic structure according to a preset mark filling method to generate a target semantic unit corresponding to the target information data. Among them, the preset mark filling method can be to fill each semantic enhancement sequence into the last position of the target semantic structure, and each semantic enhancement sequence is connected by an overline. Exemplarily, if the semantic enhancement requirement is a unit setting requirement, the initial semantic unit corresponding to the target information data generated based on the target semantic structure is: [O:Chiller1]-[A:ratedPower], and the semantic enhancement sequences are

[500] and [W], then the target semantic unit can be: [O:Chiller1]-[A:ratedPower]-

[500] _[W]. If the semantic enhancement requirement is a metadata setting requirement, the initial semantic unit corresponding to the target information data generated based on the target semantic structure is: [O:Sensor1]-[A:temperature]-

[25] , the semantic enhancement sequence is [META:dataType=float] and [META:range=0-100], then the target semantic unit can be: [O:Sensor1]-[A:temperature]-

[25] _[META:dataType=float]_[META:range=0-100]. If the semantic enhancement requirement is a timestamp setting requirement, the initial semantic unit corresponding to the target information data generated based on the target semantic structure is: [O:AHU2]-[A:runStatus]-[1], the semantic enhancement sequence is [T:2024-10-07T14:00:00], then the target semantic unit can be: [O:AHU2]-[A:runStatus]-[1]_[T:2024-10-07T14:00:00].

[0074] Generally, if the semantic enhancement requirement includes multiple requirements, the target semantic units can be sorted in the order of the unit setting requirement, the metadata setting requirement, and the timestamp setting requirement. For example, if the semantic enhancement requirements are the unit setting requirement and the metadata setting requirement, the initial semantic unit corresponding to the target information data generated based on the target semantic structure is: [O:Sensor1]-[A:temperature]-

[25] , and the semantic enhancement sequence is [C], [META:dataType=float], and [META:range=0-100], then the target semantic unit can be: [O:Sensor1]-[A:temperature]-

[25] _[C]_[META:dataType=float]_[META:range=0-100].

[0075] It is worth noting that if the target semantic structure is an object category or a relationship category, there is no need to determine the semantic enhancement sequence corresponding to the target information data based on the semantic enhancement requirements, and determine the target semantic unit corresponding to the target information data based on the semantic enhancement sequence and the target semantic structure. The target semantic structure can be directly used to determine the target semantic unit corresponding to the target information data. Specifically, the target information data is filled in the position according to the format of the target semantic structure. Exemplarily, if the target semantic structure is [object type code]-[object instance code], the structure category is object category, the target information data is the air conditioning unit code ACATCC, and the condenser 1 code Chiller1, then ACATCC and Chiller1 can be filled into the target semantic structure to obtain the target semantic unit: [O:ACATCC]-[Chiller1]. Similarly, if the target semantic structure is [object instance code 1]-[relationship code]-[object instance code 2], the structural category is the relationship category, and the target information data is object instance code 1 Chiller1, relationship code Connect, and object instance code 2 Pump1, then Chiller1, Connect, and Pump1 can be filled into the target semantic structure to obtain the target semantic unit: [O:Chiller1]-[R:Connect]-[O:Pump1].

[0076] S270 , arranging data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

[0077] Specifically, after generating each target semantic unit, each target semantic unit can be arranged according to the data prediction requirements to obtain a basic semantic unit sequence corresponding to the basic information data set. For example, if the data prediction requirements are to first input the condensing machine brand, then the condensing machine power, and finally the running status of the condensing machine. Then, the generated target semantic units can be rearranged in order according to the data prediction requirements to obtain a basic semantic unit sequence corresponding to the basic information data set: {[O:Chiller1]-[A:brand]-[York], [O:Chiller1]-[A:ratedPower]-

[500] _[W], [O:AHU2]-[A:runStatus]-[1]_[T:2024-10-07T14:00:00]}.

[0078] S280: Perform result verification on the basic semantic unit sequence based on the data prediction requirement to obtain a target semantic unit sequence that meets a preset verification standard.

[0079] Among them, the result verification may refer to the operation of verifying the integrity and consistency of the basic semantic unit sequence. Exemplarily, the result verification may be to verify whether the basic semantic unit sequence contains all the semantic units corresponding to the data prediction requirements, or to verify whether the basic semantic unit sequence satisfies the data input sequence corresponding to the data prediction requirements. The preset verification standard may refer to the pre-set standard conditions for evaluating the verification results. Exemplarily, the preset verification standard may be that the basic semantic unit sequence contains all the semantic units corresponding to the data prediction requirements, and the basic semantic unit sequence satisfies the data input sequence corresponding to the data prediction requirements.

[0080] Specifically, after generating the basic semantic unit sequence, the basic semantic unit sequence can be verified using the data prediction requirements to obtain a target semantic unit sequence that contains all the semantic units corresponding to the data prediction requirements and the data input sequence that satisfies the data prediction requirements. In this way, it can be ensured that the token sequence can accurately reflect the information of the input data, laying the foundation for the application of large models in the construction industry and improving the accuracy of the prediction results of large building models.

[0081] In an optional embodiment, after the result verification of the basic semantic unit sequence based on the data prediction demand is performed to obtain a target semantic unit sequence that meets the preset verification standard, it may also include: sending the target semantic unit sequence to the target building large model corresponding to the current prediction scene, so that the target building large model generates a prediction result corresponding to the current prediction scene. Among them, the target building large model may refer to a pre-selected building large model for data prediction. The prediction result may refer to the model result generated by the target building large model based on the target semantic unit sequence. Specifically, after generating the target semantic unit sequence corresponding to the current prediction scene, the target semantic unit sequence can be input into the pre-selected target building large model, and the prediction result corresponding to the current prediction scene is generated by the target building large model.

[0082] The technical solution of the embodiment of the present invention is to obtain a candidate information data set by performing format conversion on each basic information data in a basic information data set based on a preset information dictionary, and to perform format judgment on each candidate information data in the candidate information data set based on the preset format requirements to obtain a target information data set that meets the preset format requirements, and then, based on the target field name of the target information data, determine the corresponding target semantic structure in the preset semantic structure set, determine the structure category of the target semantic structure, and if the structure category is an attribute category, determine the semantic enhancement sequence corresponding to the target information data based on the semantic enhancement requirement, determine the target semantic unit corresponding to the target information data based on the semantic enhancement sequence and the target semantic structure, and finally, arrange the data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set, and perform result verification on the basic semantic unit sequence based on the data prediction requirement to obtain a target semantic unit sequence that meets the preset verification standard. Since the semantic units are divided by using the pre-built information dictionary and semantic structure set of the construction industry, the problem that the semantic units obtained by the tokenization method in the prior art cannot fully express the professional knowledge and complex relationships in the construction field is solved. The information data oriented to large building models can be segmented to obtain semantic units oriented to large building models that can fully express the professional knowledge and complex relationships in the construction field, laying the foundation for the application of large models in the construction industry and improving the accuracy of the prediction results of large building models.

[0083] Embodiment 3

[0084] Figure 3 This is a schematic diagram of a structure of a semantic unit determination device for a large building model provided by the third embodiment of the present invention. Figure 3 As shown, the device includes: a data acquisition module 310, a pre-processing module 320, a semantic unit determination module 330 and a data arrangement module 340;

[0085] The data acquisition module 310 is used to acquire the basic information data set corresponding to the current prediction scenario and the current training requirements; wherein the current training requirements include data prediction requirements;

[0086] A preprocessing module 320 is used to perform data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; wherein the target information data set includes each target information data;

[0087] A semantic unit determination module 330, configured to determine a corresponding target semantic structure in a preset semantic structure set based on a target field name of the target information data, and determine a target semantic unit corresponding to the target information data based on the target semantic structure;

[0088] The data arrangement module 340 is used to arrange the data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

[0089] The technical solution of the embodiment of the present invention obtains a target information data set that meets the preset format requirements by preprocessing the basic information data set based on a preset information dictionary; then, the corresponding target semantic structure is determined in the preset semantic structure set based on the target field name of the target information data in the target information data set, and the target semantic unit corresponding to the target information data is determined based on the target semantic structure; finally, the data of each target semantic unit is arranged based on the data prediction requirements to obtain a basic semantic unit sequence corresponding to the basic information data set. Since the semantic units are divided by using the pre-constructed information dictionary and semantic structure set of the construction industry, the problem that the semantic units obtained by the tokenization method in the prior art cannot fully express the professional knowledge and complex relationships in the construction field is solved, and the information data for the large model of the building can be segmented to obtain the semantic units for the large model of the building that can fully express the professional knowledge and complex relationships in the construction field, laying the foundation for the application of large models in the construction industry and improving the accuracy of the prediction results of large models of the building.

[0090] Optionally, the preprocessing module 320 may be specifically used for:

[0091] Performing format conversion on each basic information data in the basic information data set based on a preset information dictionary to obtain a candidate information data set;

[0092] Based on the preset format requirement, format judgment is performed on each of the candidate information data sets in the candidate information data set to obtain a target information data set that meets the preset format requirement.

[0093] Optionally, the semantic unit determination device for a large building model may further include: an information dictionary construction module, which is used to obtain various historical information data in a historical information database before performing data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; determine a preset definition standard corresponding to the historical information data based on a basic field name of the historical information data; determine target definition information corresponding to the historical information data based on the preset definition standard; and combine and process the target definition information corresponding to various historical information data in the historical information database to generate a preset information dictionary.

[0094] Optionally, the semantic unit determination device for a large building model may also include: a semantic structure construction module, which is used to obtain the basic field name corresponding to the full amount of historical information data before determining the corresponding target semantic structure in the preset semantic structure set based on the target field name of the target information data; determine the full amount of basic semantic fields and basic semantic order corresponding to the basic field name based on the field category of the basic field name; combine and process each basic semantic field based on the basic semantic order to generate a preset semantic structure corresponding to the basic field name; combine and process the preset semantic structures corresponding to the full amount of basic field names to generate a preset semantic structure set.

[0095] Optionally, the current training requirement also includes: semantic enhancement requirement;

[0096] The semantic unit determination module 330 can be specifically used for:

[0097] Determining a structural category of the target semantic structure, and if the structural category is an attribute category, determining a semantic enhancement sequence corresponding to the target information data based on the semantic enhancement requirement;

[0098] A target semantic unit corresponding to the target information data is determined based on the semantic enhancement sequence and the target semantic structure.

[0099] Optionally, the semantic unit determination device for a large building model may further include: a data verification module, which is used to, after arranging data for each target semantic unit based on the data prediction requirement and obtaining a basic semantic unit sequence corresponding to the basic information data set, perform result verification on the basic semantic unit sequence based on the data prediction requirement to obtain a target semantic unit sequence that meets a preset verification standard.

[0100] Optionally, the semantic unit determination device for a large building model may further include: a model prediction module, which is used to verify the results of the basic semantic unit sequence based on the data prediction requirements and obtain a target semantic unit sequence that meets a preset verification standard. Thereafter, the target semantic unit sequence is sent to a target building large model corresponding to a current prediction scene, so that the target building large model generates a prediction result corresponding to the current prediction scene.

[0101] The apparatus for determining semantic units for large building models provided in the embodiment of the present invention can execute the method for determining semantic units for large building models provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0102] Embodiment 4

[0103] Figure 4 A schematic diagram of an electronic device 410 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0104] like Figure 4 As shown, the electronic device 410 includes at least one processor 420, and a memory connected to the at least one processor 420 in communication, such as a read-only memory (ROM) 430, a random access memory (RAM) 440, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 420 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 430 or the computer program loaded from the storage unit 490 to the random access memory (RAM) 440. In the RAM 440, various programs and data required for the operation of the electronic device 410 can also be stored. The processor 420, the ROM 430, and the RAM 440 are connected to each other via a bus 450. An input / output (I / O) interface 460 is also connected to the bus 450.

[0105] Multiple components in the electronic device 410 are connected to the I / O interface 460, including: an input unit 470, such as a keyboard, a mouse, etc.; an output unit 480, such as various types of displays, speakers, etc.; a storage unit 490, such as a disk, an optical disk, etc.; and a communication unit 4100, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 4100 allows the electronic device 410 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0106] The processor 420 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 420 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 420 executes the various methods and processes described above, such as the semantic unit determination method for a large building model.

[0107] The method includes:

[0108] Obtaining a basic information data set corresponding to the current prediction scenario and a current training requirement; wherein the current training requirement includes a data prediction requirement;

[0109] Performing data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; wherein the target information data set includes each target information data;

[0110] Determining a corresponding target semantic structure in a preset semantic structure set based on a target field name of the target information data, and determining a target semantic unit corresponding to the target information data based on the target semantic structure;

[0111] The data of each target semantic unit is arranged based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

[0112] In some embodiments, the semantic unit determination method for building large models can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 490. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 410 via the ROM 430 and / or the communication unit 4100. When the computer program is loaded into the RAM 440 and executed by the processor 420, one or more steps of the semantic unit determination method for building large models described above can be performed. Alternatively, in other embodiments, the processor 420 can be configured to execute the semantic unit determination method for building large models in any other appropriate manner (for example, by means of firmware).

[0113] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0114] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0115] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0116] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the 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, voice input, or tactile input).

[0117] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0118] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0119] The embodiment of the present application also discloses a computer program product, which includes a computer program. When the computer program is executed by a processor, the method for determining a semantic unit for a large building model provided in any embodiment of the present application is implemented. The program product and the method for determining a semantic unit for a large building model disclosed in each embodiment of the present application belong to the same inventive concept, so they are not described here in detail.

[0120] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0121] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for determining semantic units for a large building model, characterized in that: The method comprises: Obtaining a basic information data set corresponding to the current prediction scenario and a current training requirement; wherein the current training requirement includes a data prediction requirement; Performing data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; wherein the target information data set includes each target information data; Determining a corresponding target semantic structure in a preset semantic structure set based on a target field name of the target information data, and determining a target semantic unit corresponding to the target information data based on the target semantic structure; The data of each target semantic unit is arranged based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

2. The method according to claim 1, characterized in that The data preprocessing of the basic information data set based on the preset information dictionary to obtain a target information data set that meets the preset format requirements includes: Performing format conversion on each basic information data in the basic information data set based on a preset information dictionary to obtain a candidate information data set; Based on the preset format requirement, format judgment is performed on each of the candidate information data sets in the candidate information data set to obtain a target information data set that meets the preset format requirement.

3. The method according to claim 1, characterized in that Before the basic information data set is preprocessed based on the preset information dictionary to obtain a target information data set that meets the preset format requirements, the method further includes: Obtaining each historical information data in the historical information database; Determining a preset definition standard corresponding to the historical information data based on a basic field name of the historical information data; Determine the target definition information corresponding to the historical information data based on the preset definition standard; The target definition information corresponding to each historical information data in the historical information database is combined and processed to generate a preset information dictionary.

4. The method according to claim 1, characterized in that: Before determining the corresponding target semantic structure in the preset semantic structure set based on the target field name of the target information data, the method further includes: Get the basic field names corresponding to the full amount of historical information data; Determine the full amount of basic semantic fields and basic semantic order corresponding to the basic field name based on the field category of the basic field name; Combining and processing each basic semantic field based on the basic semantic sequence, generating a preset semantic structure corresponding to the basic field name; The preset semantic structures corresponding to all basic field names are combined and processed to generate a preset semantic structure set.

5. The method according to claim 1, characterized in that: The current training requirements also include: semantic enhancement requirements; The determining the target semantic unit corresponding to the target information data based on the target semantic structure includes: Determining a structural category of the target semantic structure, and if the structural category is an attribute category, determining a semantic enhancement sequence corresponding to the target information data based on the semantic enhancement requirement; A target semantic unit corresponding to the target information data is determined based on the semantic enhancement sequence and the target semantic structure.

6. The method according to claim 1, characterized in that After arranging data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set, the method further includes: The basic semantic unit sequence is verified based on the data prediction requirement to obtain a target semantic unit sequence that meets a preset verification standard.

7. The method according to claim 6, characterized in that After the basic semantic unit sequence is verified based on the data prediction requirement to obtain a target semantic unit sequence that meets a preset verification standard, the method further includes: The target semantic unit sequence is sent to the target building large model corresponding to the current prediction scene, so that the target building large model generates a prediction result corresponding to the current prediction scene.

8. A semantic unit determination device for a large building model, characterized in that: include: A data acquisition module is used to acquire a basic information data set corresponding to the current prediction scenario and a current training requirement; wherein the current training requirement includes a data prediction requirement; A preprocessing module, used for performing data preprocessing on the basic information data set based on a preset information dictionary to obtain a target information data set that meets the preset format requirements; wherein the target information data set includes each target information data; A semantic unit determination module, configured to determine a corresponding target semantic structure in a preset semantic structure set based on a target field name of the target information data, and determine a target semantic unit corresponding to the target information data based on the target semantic structure; The data arrangement module is used to arrange the data of each target semantic unit based on the data prediction requirement to obtain a basic semantic unit sequence corresponding to the basic information data set.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the semantic unit determination method for a large building model according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the semantic unit determination method for a large building model according to any one of claims 1 to 7 when executed.

Images