Building knowledge graph generation method, data processing method, middleware and system

Abstract

The embodiment of the application provides a building knowledge graph generation method, a data processing method, middleware and a system. According to the correlation between the running parameters and the values thereof in the device data from the first cloud platform and the environmental parameters and the values thereof in the regional environmental data of each region in the building from the second cloud platform, an unclear association relationship between entities of different cloud platforms, that is, a first association relationship, can be found, so that a knowledge graph is generated based on the first association relationship, the generated knowledge graph integrates data of different cloud platforms, a more comprehensive and accurate knowledge graph of the building can be obtained, and therefore comprehensive and accurate data can be quickly provided based on the knowledge graph.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for generating a building knowledge graph, a data processing method, middleware, and a system. Background Technology

[0002] With the development of technology and the abundance of information, the application of the Internet of Things (IoT) is becoming increasingly widespread. In IoT systems, applications can access cloud platforms via the network to obtain the data they need.

[0003] To improve the connection efficiency between applications and cloud platforms, technologies that utilize middleware as an intermediary for data transmission between them have emerged in recent years. For example, based on a query request from the application, the middleware retrieves a list of devices and their attribute statuses from the cloud platform and returns them to the application, enabling unified management of the status and messages of IoT devices.

[0004] In addition, in recent years, technologies that utilize knowledge graphs to assist in equipment management have emerged. For example, knowledge can be extracted from equipment-related data to build knowledge graphs for equipment monitoring and management.

[0005] It should be noted that the above description of the technical background is only for the purpose of providing a clear and complete explanation of the technical solutions of the present invention and facilitating understanding by those skilled in the art. It should not be assumed that the above technical solutions are known to those skilled in the art simply because they have been described in the background section of this invention. Summary of the Invention

[0006] Within a building, there are typically various types of equipment from different manufacturers. The data of these different devices is stored on different cloud platforms. In the existing methods mentioned above, even if knowledge graphs are used, the knowledge graphs cannot effectively integrate the data from multiple cloud platforms. This results in the application needing to retrieve data from multiple cloud platforms separately, making it impossible to quickly and accurately obtain comprehensive data.

[0007] Even though the existing methods mentioned above utilize middleware, the data is scattered across multiple cloud platforms, requiring the middleware to forward requests and merge data multiple times. This results in low data acquisition efficiency, slow processing speed, inflexible response, and increased access load on the cloud platforms.

[0008] In addition, existing middleware has low processing capabilities. It does not filter and process data, but directly returns it to the application, resulting in data redundancy and increasing the processing burden on the application.

[0009] In addition, existing middleware lacks in-depth data understanding and analysis, making it difficult to meet the customized needs of applications and provide intelligent decision-making and personalized services.

[0010] To address one or more of the aforementioned problems, embodiments of the present invention provide a method for generating a building knowledge graph, a data processing method, middleware, and a system.

[0011] According to a first aspect of the present invention, a method for generating a building knowledge graph is provided. The method includes: acquiring device data from a first cloud platform and regional environmental data of various areas within a building from a second cloud platform; the device data includes device category information, device identification information, device operating parameters and their values, and the regional environmental data includes regional information, regionally associated environmental parameters and their values; extracting a first set of entities based on the device data from the first cloud platform; the first set of entities includes device category entities, device identification entities, and device operating parameter entities; extracting a second set of entities based on the regional environmental data of various areas within the building from the second cloud platform; the second set of entities includes regional entities and environmental parameter entities; determining a first association relationship between the first set of entities and the second set of entities based on the correlation between the device operating parameters and their values ​​in the device data and the environmental parameters and their values ​​in the regional environmental data; and generating a knowledge graph of the building based at least on the first association relationship between the first set of entities and the second set of entities.

[0012] According to a second aspect of the present invention, a data processing method is provided in an Internet of Things (IoT) system, the IoT system including at least one application terminal, at least two cloud platforms, and middleware, the method being applied to the middleware, the method comprising: the middleware receiving request data from at least one application terminal; the middleware obtaining feedback data in response to the request data using a pre-generated knowledge graph based on the request data, wherein the knowledge graph is generated by the method described in the first aspect of the present invention; and the middleware sending the feedback data to the application terminal.

[0013] According to a third aspect of the present invention, a middleware is provided, the middleware comprising: a knowledge graph generation module and / or a data processing module, which is used to execute the methods described in the first and / or second aspects of the present invention.

[0014] According to a fourth aspect of the present invention, an Internet of Things (IoT) system is provided, the IoT system including at least the middleware described in the third aspect of the present invention.

[0015] The beneficial effects of the embodiments of the present invention are as follows:

[0016] Based on the correlation between the operating parameters and their values ​​in the device data from the first cloud platform and the environmental parameters and their values ​​in the regional environmental data of each area within the building from the second cloud platform, the ambiguous relationships between entities on different cloud platforms can be discovered, namely the first relationship. Based on this first relationship, a knowledge graph is generated, which integrates the data from different cloud platforms, resulting in a more comprehensive and accurate knowledge graph of the building. Based on this knowledge graph, comprehensive and accurate data can be provided quickly.

[0017] Furthermore, based on the natural language processing model, data from different platforms are preprocessed and entity recognition is performed. The explicit relationships between entities from different cloud platforms, namely the second relationship, are extracted. Based on the identified entities, the extracted second relationship, and the aforementioned first relationship, a knowledge graph is generated. Therefore, it is possible to further integrate data from different cloud platforms and improve the accuracy and comprehensiveness of the knowledge graph.

[0018] Furthermore, the equipment operating parameters in the equipment data include at least one of the equipment's set air quality parameters and equipment energy consumption parameters; the environmental parameters in the regional environmental data include at least one of the regional air quality parameters, regional energy consumption parameters, and regional building parameters; and the first association between the first group of entities and the second group of entities includes the association between the equipment category entity and equipment identification entity in the first group of entities and the regional entity in the second group of entities. For example, air quality parameters include at least one of temperature and humidity, energy consumption parameters include power consumption, and regional building parameters include regional heat load. Thus, by utilizing the correlation between air quality parameters such as equipment set temperature or humidity and regional air quality parameters such as temperature or humidity, or by utilizing the correlation between energy consumption parameters such as equipment power consumption and regional energy consumption parameters such as regional power consumption or regional building parameters such as regional heat load, the previously unclear association between the equipment in the first cloud platform and the region in the second cloud platform can be accurately determined, providing an important foundation for generating a comprehensive and accurate knowledge graph.

[0019] Furthermore, based on the similarity between the first relationship between the set air quality parameter values ​​and time and the second relationship between the set air quality parameter values ​​and time, the association between the device entity corresponding to the set air quality parameter and the regional entity corresponding to the regional air quality parameter can be determined. In this way, it is possible to determine whether the device is located in the region based on the similarity between the changes in the set temperature or humidity of the device and the changes in the corresponding parameters of the region, thereby reliably establishing the association between the two.

[0020] Alternatively, the correlation between the device entity corresponding to the device energy consumption parameter and the regional entity corresponding to the regional energy consumption parameter can be determined based on the similarity between the third relationship of the device energy consumption parameter value and time and the fourth relationship of the regional energy consumption parameter value and time. In this way, it is possible to determine whether the device is located in the region based on the similarity between the power consumption change of the device and the power consumption change of the region, thereby reliably establishing the correlation between the two.

[0021] Alternatively, the correlation between the equipment entity corresponding to the equipment energy consumption parameter and the regional entity corresponding to the regional building parameter can be determined based on the similarity between the third relationship of the equipment energy consumption parameter value and time and the fifth relationship of the regional building parameter value and time. In this way, it is possible to determine whether the equipment is located in the region based on the similarity between the change in the equipment's power consumption and the change in the region's heat load, thereby reliably establishing the correlation between the two.

[0022] Furthermore, the knowledge graph mentioned above is generated by middleware in the IoT system. In this way, the middleware can easily use the knowledge graph to quickly provide the application with comprehensive and accurate data that integrates multiple cloud platforms.

[0023] The beneficial effects of the embodiments of the present invention also include:

[0024] By leveraging the knowledge graph generated using the methods described above, middleware can perform data processing and relationship integration based on data from multiple cloud platforms, thereby directly returning the processed data to the application. This improves the response speed to the application and reduces the data processing burden on the application and the access load on the cloud platform. Furthermore, based on the generated knowledge graph, the middleware can accurately understand the intent of the application user and provide more relevant and appropriate results, thus achieving efficient connection between the application and the cloud platform.

[0025] Furthermore, based on the aforementioned knowledge graph and artificial intelligence model, the middleware generates target information corresponding to the requested data, such as missing entity information, prediction information, and suggestion information in the knowledge graph; thus, it can also provide intelligent decision-making and personalized services to meet the customized needs of the application.

[0026] Furthermore, the request data includes the user's query request, and the feedback data includes the query results for the query request. In this way, the middleware, based on the knowledge graph, can accurately utilize the user's query intent and provide more relevant and accurate query results.

[0027] Furthermore, the middleware has a unified transmission interface module for access interfaces of all applications. This improves data transmission efficiency, and external applications only need to set up an access mechanism to connect to one middleware, without having to connect to multiple cloud platforms, thus reducing the development burden of external applications.

[0028] Furthermore, the middleware can display the knowledge graph; alternatively, the middleware can display the knowledge graph through the application. This allows users to understand and use the knowledge graph more intuitively, further enhancing the user experience.

[0029] The feature information described and illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, combined with feature information in other embodiments, or substituted for feature information in other embodiments.

[0030] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, whole, step, or component, but does not exclude the presence or addition of one or more other features, wholes, steps, or components. Attached Figure Description

[0031] Many aspects of the invention can be better understood by referring to the following accompanying drawings. The components in the drawings are not drawn to scale, but are only intended to illustrate the principles of the invention. Corresponding portions in the drawings may be enlarged or reduced for ease of illustration and description of certain parts of the invention. Elements and features described in one drawing or embodiment of the invention may be combined with elements and features shown in one or more other drawings or embodiments. Furthermore, similar reference numerals in the drawings denote corresponding components in several drawings and can be used to indicate corresponding components used in more than one embodiment.

[0032] In the attached diagram:

[0033] Figure 1 This is a flowchart of a method for generating a building knowledge graph according to an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram illustrating an example of establishing the association between entities on a first cloud platform and entities on a second cloud platform according to an embodiment of the present invention.

[0035] Figure 3 This is a schematic diagram comparing the changes in equipment operating parameters and regional air quality parameters in the implementation of this invention;

[0036] Figure 4 This is a flowchart illustrating one implementation method of generating a knowledge graph according to the present invention;

[0037] Figure 5 This is a flowchart illustrating one implementation of calculating entity similarity according to an embodiment of the present invention;

[0038] Figure 6 This is a schematic diagram of a building knowledge graph, i.e., a target knowledge graph, generated based on data from multiple cloud platforms according to an embodiment of the present invention.

[0039] Figure 7 This is a schematic diagram of a knowledge graph generated by middleware based on data from multiple cloud platforms, according to an embodiment of the present invention.

[0040] Figure 8 This is a flowchart of a data processing method in an Internet of Things system according to an embodiment of the present invention;

[0041] Figure 9 This is an architecture diagram of an Internet of Things (IoT) system according to an embodiment of the present invention;

[0042] Figure 10 This is a schematic diagram illustrating an example of how middleware in this invention uses knowledge graphs to provide feedback data to the application.

[0043] Figure 11 This is a schematic diagram of the middleware in an embodiment of the present invention;

[0044] Figure 12 This is a schematic diagram of the hardware configuration of the middleware in an embodiment of this application;

[0045] Figure 13 This is a schematic diagram of an Internet of Things (IoT) system according to an embodiment of the present invention. Detailed Implementation

[0046] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings.

[0047] Example 1

[0048] This invention provides a method for generating a building knowledge graph.

[0049] Figure 1 This is a flowchart of a method for generating a building knowledge graph according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0050] Step 101: Obtain device data from the first cloud platform and regional environmental data of each area within the building from the second cloud platform;

[0051] Equipment data includes equipment category information, equipment identification information, equipment operating parameters and their values; regional environmental data includes regional information, regional associated environmental parameters and their values.

[0052] Step 102: Extract the first set of entities based on the device data from the first cloud platform; the first set of entities includes the device category entity, the device identifier entity, and the device operating parameter entity;

[0053] Step 103: Extract the second set of entities based on the regional environmental data of each area within the building from the second cloud platform; the second set of entities includes regional entities and environmental parameter entities;

[0054] Step 104: Based on the correlation between the equipment operating parameters and their values ​​in the equipment data and the environmental parameters and their values ​​in the regional environmental data, determine the first association relationship between the first group of entities and the second group of entities;

[0055] Step 105: Generate a knowledge graph of the building based at least on the first association between the first group of entities and the second group of entities.

[0056] In this way, based on the correlation between the operating parameters and their values ​​in the device data from the first cloud platform and the environmental parameters and their values ​​in the regional environmental data of each area within the building from the second cloud platform, the ambiguous relationships between entities on different cloud platforms can be discovered, namely the first relationship. Based on this first relationship, a knowledge graph can be generated, which integrates the data from different cloud platforms, resulting in a more comprehensive and accurate knowledge graph of the building. Based on this knowledge graph, comprehensive and accurate data can be provided quickly.

[0057] In some embodiments, the knowledge graph of buildings is generated by middleware in the Internet of Things system; that is, the building knowledge graph generation method of this application embodiment is executed by middleware. In this way, the middleware can easily utilize the knowledge graph to quickly provide the application with comprehensive and accurate data integrating multiple cloud platforms. However, this embodiment of the invention does not limit the entity that generates the knowledge graph; for example, other entities can also generate the knowledge graph and provide it for use by the middleware.

[0058] In some embodiments, the building can be of various types, and may also be referred to as a building. Examples of buildings in these embodiments include office buildings, shopping malls, factory workshops, schools, apartments, and ordinary residences. These embodiments do not limit the type of building.

[0059] In some embodiments, a knowledge graph of a building is used to describe the concepts within the building and their relationships, represented by an "entity-relationship-entity" model. An entity is the basic unit in the knowledge graph, representing a real-world thing or concept related to the building. Examples include people, locations, and equipment. An instance is a concrete representation of an entity, such as "air conditioner." Entity attributes describe the characteristics or properties of the entity, such as "air conditioner operating parameters."

[0060] For example, a knowledge graph includes at least two entities, the attributes of the entities, and the relationships between the entities.

[0061] For example, an entity includes at least one of a building, space, equipment, or user. In a knowledge graph, an entity is unique.

[0062] In some embodiments, the equipment can be of various types, such as air conditioning equipment (e.g., indoor units), fresh air equipment, humidifying or dehumidifying equipment, lighting equipment, household appliances, heating equipment, large machinery, etc. Furthermore, these devices can be from different manufacturers. This embodiment of the invention does not limit the type, quantity, or manufacturer of the equipment.

[0063] In some embodiments, the first cloud platform is a cloud platform that includes device data; the device data includes device category information, device identification information, device operating parameters and their values; the device operating parameters include at least one of the device's set air quality parameters and device energy consumption parameters; the device category information indicates the category of the device, and the device identification information includes the device name or device ID.

[0064] For example, the first cloud platform is the cloud platform of an air conditioner manufacturer; the device category information is indoor unit; the device identification information is indoor unit a; the device operating parameters are the air conditioner's set temperature or set humidity, or the air conditioner's power consumption, and the numerical value of the device operating parameters is, for example, a set temperature of 25°C. The above is merely an example, and this embodiment of the invention does not limit the specific content of the first cloud platform and the device data.

[0065] In some embodiments, the second cloud platform is a cloud platform that includes building data, and the regional environmental data includes regional information, regionally associated environmental parameters and their values; the environmental parameters in the regional environmental data include at least one of regional air quality parameters, regional energy consumption parameters and regional building parameters.

[0066] For example, the second cloud platform is a cloud platform used for building management; the area information is the location information of various areas within the building, such as "meeting room"; the area air quality parameter is the area temperature or area humidity; the area energy consumption parameter is the area power consumption; and the area building parameter is the area heat load. The above is merely an example, and this embodiment of the invention does not limit the specific content of the second cloud platform and the equipment data.

[0067] In this way, by utilizing the correlation between air quality parameters such as temperature or humidity set by the equipment and air quality parameters such as temperature or humidity of the region, or by utilizing the correlation between energy consumption parameters such as power consumption of the equipment and energy consumption parameters such as power consumption of the region or regional building parameters such as heat load of the region, the previously unclear relationship between the equipment in the first cloud platform and the region in the second cloud platform can be accurately determined, providing an important foundation for generating a comprehensive and accurate knowledge graph.

[0068] In step 101, device data from the first cloud platform and regional environmental data of each area within the building from the second cloud platform are obtained. For example, device data is obtained from the server of the first cloud platform and regional environmental data is obtained from the server of the second cloud platform.

[0069] In step 102, a first set of entities is extracted based on the device data from the first cloud platform. The first set of entities includes device category entities, device identifier entities, and device operating parameter entities. For example, the data obtained from the server of the first cloud platform is identified and parsed to obtain the device category (e.g., air conditioner indoor unit), device identifier (e.g., indoor unit ID), and device operating parameters (e.g., indoor unit set temperature, set humidity, power consumption, etc.), which are used as the extracted entities.

[0070] In step 103, a second set of entities is extracted based on the regional environmental data of each area within the building from the second cloud platform. The second set of entities includes regional entities and environmental parameter entities. For example, the data obtained from the server of the second cloud platform is identified and parsed to obtain each area of ​​the building and the environmental parameters of each area (such as temperature, humidity, power consumption, etc.), which are used as the extracted entities.

[0071] In step 104, based on the correlation between the equipment operating parameters and their values ​​in the equipment data and the environmental parameters and their values ​​in the regional environmental data, a first association relationship between the first group of entities and the second group of entities is determined. In some embodiments, the first association relationship between the first group of entities and the second group of entities includes the association relationship between the equipment category entity and the equipment identification entity in the first group of entities and the regional entity in the second group of entities, that is, the association relationship between the equipment entity and the regional entity.

[0072] For example, based on the similarity between the changes in the values ​​of equipment operating parameters in the equipment data over time and the changes (or trends) in the values ​​of environmental parameters in the regional environmental data over the same time period, the association between equipment entities and regional entities can be determined. That is, it can be determined whether the equipment corresponding to the equipment operating parameters is located in the region corresponding to the regional environmental parameters.

[0073] Thus, it is possible to determine the relationships that are not recorded on either the first cloud platform or the second cloud platform, thereby establishing the previously ambiguous relationships between the data on the first cloud platform and the second cloud platform.

[0074] The following example illustrates this. Figure 2 This is a schematic diagram illustrating an example of establishing an association between entities on a first cloud platform and entities on a second cloud platform according to an embodiment of the present invention.

[0075] like Figure 2As shown, the data in the first cloud platform may include device data, such as the set temperature and humidity of the air conditioner, and the location information of the air conditioning system, i.e., air conditioning system 1 includes indoor unit a, indoor unit b, and indoor unit c, and air conditioning system 2 includes indoor unit e and indoor unit f. In addition, it may also include the installation location information of each indoor unit. However, this installation location information may be unclear or incomplete. Furthermore, the zoning of the building floor plan may change. Therefore, the relationship between indoor units and areas cannot be determined solely based on the data from the first cloud platform. That is, it is impossible to determine which indoor units are installed in the same area, or which indoor units are installed in a certain area.

[0076] like Figure 2 As shown, the data in the second cloud platform can include regional environmental data, such as temperature, humidity, and location information (e.g., conference room) for each region.

[0077] For example, based on the judgment in step 104 above, if it is determined that the temperature changes of indoor units a, b, and e over time are similar to the temperature changes of the conference room at the same time, then it is determined that indoor units a, b, and e are associated with the conference room. That is, indoor units a, b, and e are all located in the same area, "conference room," thus establishing the first association between the equipment entity and the area entity. Figure 2 The first association relationship is represented as 1). Additionally, as... Figure 2 As shown, the first association can also be represented in other forms, such as the first association 2, which indicates that for the area "meeting room", the associated equipment is indoor unit a, indoor unit b and indoor unit e. In addition, it can also indicate the air conditioning system to which each indoor unit belongs.

[0078] The following section provides a detailed explanation of how, in step 104, the first association relationship between the first group of entities and the second group of entities is determined based on the correlation between the equipment operating parameters and their values ​​in the equipment data and the environmental parameters and their values ​​in the regional environmental data.

[0079] In some embodiments, the association between the device category entity and the device identifier entity corresponding to the set air quality parameter and the regional entity corresponding to the regional air quality parameter is determined based on the similarity between a first relationship between the value and time of the set air quality parameter and a second relationship between the value and time of the regional air quality parameter.

[0080] In some embodiments, regional air quality parameters can be obtained by using IAQ (Indoor Air Quality) sensors located within the region.

[0081] For example, the relationship between the indoor unit of the air conditioner and the area can be determined by the similarity between the value of the air conditioner's set temperature and the value of the area temperature within the same period of time.

[0082] Specifically, obtain the device operating parameters and their values ​​from the first cloud platform, such as the air conditioner set temperature. Use the function y = F(t) to represent the relationship between the set temperature value of each air conditioner indoor unit and time, i.e., the first relationship.

[0083] Additionally, regional air quality parameters from the second cloud platform are obtained, such as temperature values ​​detected by IAQ sensors in each region. Similarly, the relationship between temperature values ​​and time in each region is represented by y' = F(t), i.e., the second relationship.

[0084] Then, the similarity between the first relation y and the second relation y' is compared, for example, by calculating their mean squared error (MSE) according to the following formula (1):

[0085]

[0086] Among them, y t y′ represents the time function representing the set temperature of the indoor unit of the air conditioner. t The time function representing the temperature of a region, where time t ranges from 0 to n.

[0087] The smaller the MSE value, the greater the similarity between y and y', that is, the higher the correlation. For example, when the MSE is less than the preset threshold, it can be determined that the indoor unit of the air conditioner corresponding to function y is set in the area corresponding to function y'.

[0088] Figure 3 This is a schematic diagram comparing the changes in equipment operating parameters and regional air quality parameters in the implementation of this invention.

[0089] like Figure 3 As shown, in each indoor unit of the air conditioner, the set temperature and time relationship functions of indoor units a and b of air conditioning system 1 and indoor unit e of air conditioning system 2 are y1, y2 and y3 respectively, the set temperature and time relationship functions of indoor units e and f of air conditioning system 2 are y4 and y5 respectively, and the temperature and time relationship function measured by the IAQ sensor in the conference room is y'.

[0090] Based on the calculation results of MSE, combined with Figure 3As shown, the set temperature versus time functions y1, y2, and y4 of indoor units a and b of air conditioning system 1 and indoor unit e of air conditioning system 2 are similar to the temperature versus time function y' of the conference room. Therefore, it can be determined that indoor units a and b of air conditioning system 1 and indoor unit e of air conditioning system 2 are associated with the area "conference room," that is, indoor units a and b of air conditioning system 1 and indoor unit e of air conditioning system 2 are located in the same area "conference room." However, the set temperature versus time functions y3 and y4 of indoor units c of air conditioning system 1 and indoor unit f of air conditioning system 2 are significantly different from the temperature versus time function y' of the conference room. Therefore, it can be determined that these indoor units are not located in the area "conference room."

[0091] In this way, it is possible to determine whether a device is located in a region based on the similarity between changes in the device's set temperature or humidity and changes in corresponding parameters in the region, thereby reliably establishing a correlation between the two.

[0092] In some embodiments, the association between the device category entity corresponding to the device energy consumption parameter and the device identifier entity and the regional entity corresponding to the regional energy consumption parameter is determined based on the similarity between the third relationship of the value and time of the device energy consumption parameter and the fourth relationship of the value and time of the regional energy consumption parameter.

[0093] In some embodiments, device energy consumption parameters, such as device power consumption, can be determined by counting the electricity meters installed on the device, or calculated based on parameters such as device current; area energy consumption parameters, such as area power consumption, can be determined by counting the electricity meters installed in the area.

[0094] For example, the correlation between the air conditioner indoor unit and the area can be determined based on the similarity between the power consumption of the indoor unit and the power consumption of the area within the same period of time.

[0095] Specifically, obtain the energy consumption parameters of the devices on the first cloud platform, such as the power consumption of the indoor air conditioner unit. Use the function z = F(t) to represent the relationship between the power consumption of each indoor air conditioner unit and time, i.e., the third relationship.

[0096] Additionally, regional energy consumption parameters of the second cloud platform are obtained, such as the power consumption of each region. Similarly, the relationship between regional power consumption and time is represented by z' = F(t), i.e., the fourth relationship.

[0097] Then, the similarity between the third relation z and the fourth relation z' is compared, for example, by calculating their mean squared error (MSE) according to the following formula (2):

[0098]

[0099] Among them, z tThe time function representing the power consumption of the indoor unit of the air conditioner, z′ t The time function representing the regional power consumption, where time t ranges from 0 to n.

[0100] The smaller the MSE value, the greater the similarity between z and z', that is, the higher the correlation. For example, when the MSE is less than the preset threshold, it can be determined that the indoor unit of the air conditioner corresponding to function z is set in the area corresponding to function z'.

[0101] In this way, it is possible to determine whether a device is located in a region based on the similarity between the device's power consumption changes and the region's power consumption changes, thereby reliably establishing the correlation between the two.

[0102] In some embodiments, the association between the equipment category entity corresponding to the equipment energy consumption parameter and the equipment identification entity and the regional entity corresponding to the regional building parameter is determined based on the similarity between the third relationship of the value and time of the equipment energy consumption parameter and the fifth relationship of the value and time of the regional building parameter.

[0103] In some embodiments, equipment energy consumption parameters, such as equipment power consumption, can be determined by counting the electricity meters installed on the equipment, or calculated based on parameters such as equipment current; and regional building parameters, such as regional heat load, can be calculated using formulas.

[0104] For example, the correlation between the air conditioner indoor unit and the area can be determined based on the similarity between the power consumption of the air conditioner indoor unit and the changes in the area's heat load over the same period of time.

[0105] Specifically, obtain the energy consumption parameters of the devices on the first cloud platform, such as the power consumption of the indoor air conditioner unit. Use the function z = F(t) to represent the relationship between the power consumption of each indoor air conditioner unit and time, i.e., the third relationship.

[0106] Additionally, regional building parameters, such as the heat load of each region, are obtained from the second cloud platform. Similarly, the relationship between regional heat load and time is represented by Q = F(t), i.e., the fifth relationship.

[0107] For example, the heat load Q of each region can be calculated using the following formula (3):

[0108] Q = Q solar +Q skin +Q buil +Q air +Q outair +I g (3)

[0109] Among them, Q solar Q represents the heat load from solar radiation. skin Q represents the cross-flow heat load. builQ represents the heat storage capacity of facilities (such as furniture) within the area. air Q represents the heat storage capacity of air. outair Indicates heat exchange load, I g This indicates the internal load. Furthermore, the specific calculation methods for these parameters can be found in relevant technical documents, and will not be elaborated upon here.

[0110] Then, the similarity between the third relation z and the fifth relation Q is compared, for example, by calculating their mean squared error (MSE) according to the following formula (4):

[0111]

[0112] Among them, z t Q represents the time function of the power consumption of the indoor unit of the air conditioner. t The time function representing the regional heat load, where time t ranges from 0 to n.

[0113] The smaller the MSE value, the greater the similarity between z and z', that is, the higher the correlation. For example, when the MSE is less than the preset threshold, it can be determined that the indoor unit of the air conditioner corresponding to function z is set in the area corresponding to function z'.

[0114] In this way, it is possible to determine whether a device is located in a region based on the similarity between the device's power consumption changes and the region's heat load changes, thereby reliably establishing the correlation between the two.

[0115] The above explains how to determine the first association relationship between the first group of entities and the second group of entities in step 104 based on the correlation between the equipment operating parameters and their values ​​in the equipment data and the environmental parameters and their values ​​in the regional environmental data. After determining the first association relationship between the first group of entities and the second group of entities, in step 105, at least based on the first association relationship between the first group of entities and the second group of entities, a knowledge graph of the building is generated.

[0116] In some embodiments, the method further includes:

[0117] Based on a Natural Language Processing (NLP) model, data from a first cloud platform and data from a second cloud platform are preprocessed; entity recognition is performed on the preprocessed data; and a second association relationship representing the relationship between entities is extracted, such as the explicit association relationship between entities in data from different cloud platforms.

[0118] Therefore, in step 105, a knowledge graph is generated based on the identified entities, the extracted second relationships, and the first relationships determined in step 104. This allows for further and more comprehensive integration of data from different cloud platforms, improving the accuracy and comprehensiveness of the knowledge graph.

[0119] Figure 4 This is a flowchart illustrating one implementation method of generating a knowledge graph according to the present invention.

[0120] like Figure 4 As shown, the process of generating a knowledge graph can be roughly divided into several stages: data preprocessing, entity relationship identification, graph database storage, and optimization. Specifically, firstly, the raw input data of the cloud platform undergoes data cleaning, preprocessing, and structuring. Then, using generative AI natural language processing techniques (such as BERT, GPT, and other pre-trained language models), the integrated building data is further preprocessed before algorithm model training. Entity recognition technology is used to identify key entities in the building data, such as equipment names, models, and characteristics. Relationship extraction technology is used to clarify the relationships between entities, such as connections and functions of equipment, i.e., extracting association relationships. Based on the extracted entities and association relationships (such as first and second association relationships), a knowledge graph of the buildings is constructed and stored and queried using a graph database.

[0121] The following is an exemplary description of the implementation method for entity relationship recognition (also known as entity linking or entity mapping). For example, firstly, entity preprocessing is performed, including cleaning entity names and standardizing formats (e.g., removing keywords and unifying terminology); then, feature extraction is performed, including extracting features of "objects" and "buildings," such as whether attributes are consistent (area, number of floors, etc.); next, mapping rules are defined, including formulating rules such as synonyms and word root similarity, for example, pre-defining object = building = building; then, similarity calculation is performed. If a word embedding model is used, the similarity between "building" and "building" is calculated. If it exceeds a preset threshold, mapping is performed; finally, verification and adjustment are performed, including manual verification of the initial mapping results and adjustment of rules to improve mapping accuracy.

[0122] For example, various calculation methods can be used when performing similarity calculations, such as Euclidean distance, cosine similarity, and Jaccard similarity. Taking cosine similarity as an example, the specific steps are as follows:

[0123] First, use word embedding models (such as Word2Vec, FastText, BERT, etc.) to map the entity "building" to a vector space; using the Word2Vec / FastText model, input "building", the model will return a vector of fixed dimensions; for example, the vector dimension can be 100 dimensions or 200 dimensions.

[0124] Next, feature representation is performed, defining the feature space, including assigning a feature dimension to each attribute of the "building"; vectors are constructed, including encoding relevant information into numerical values; for example, the feature vector of a building might be an array containing area, number of floors, and environmental facilities, such as 1000, 10, 11000, 10, 1 (square meters, number of floors, nearby schools); the "object" vector is defined as A, and the building vector as B;

[0125] Finally, similarity is calculated. Cosine similarity measures the angle between two vectors; if they point in similar directions, the cosine value is close to 1. The calculation formula is as follows:

[0126]

[0127] Where A and B are the feature vectors of two entities, * is the dot product, and ||A|| and ||B|| are the Euclidean norms of the vectors.

[0128] Figure 5 This is a flowchart illustrating one implementation of calculating entity similarity according to an embodiment of the present invention. For example... Figure 5 As shown, "objects" and "buildings" are input into the model, and their respective feature vectors A and B are output. Then, the cosine similarity between feature vectors A and B is calculated.

[0129] Through steps 101-105 above, a knowledge graph of the building is generated.

[0130] Figure 6 This is a schematic diagram of a building knowledge graph, i.e., a target knowledge graph, generated based on data from multiple cloud platforms according to an embodiment of the present invention. Figure 6 As shown, data from the first and second cloud platforms are fully integrated to obtain a comprehensive and accurate knowledge graph.

[0131] As described above, in some embodiments, the knowledge graph is generated and stored by middleware.

[0132] Figure 7 This is a schematic diagram illustrating the generation of a knowledge graph from data across multiple cloud platforms by middleware, according to an embodiment of the present invention. For example... Figure 7 As shown, the middleware's knowledge graph generation module receives different data from different modules or databases in the first and second cloud platforms, integrates this data, and generates a building (structure) knowledge graph. Additionally, application clients 1 to N access the middleware, and the middleware uses the generated knowledge graph to generate feedback data for the application clients, where N is an integer greater than 1. In this way, the middleware can easily utilize this knowledge graph to quickly provide the application clients with comprehensive and accurate data integrated from multiple cloud platforms. For a detailed explanation of the process, please refer to the specific description in the embodiments described later.

[0133] In the above embodiments, multiple cloud platforms, including a first cloud platform and a second cloud platform, were used as examples for illustration. However, the embodiments of the present invention may also include more cloud platforms that store other data, that is, they can integrate data from more cloud platforms. The methods and processes for data integration are similar to those described above and will not be repeated here.

[0134] As can be seen from the above embodiments, based on the correlation between the operating parameters and their values ​​in the device data from the first cloud platform and the environmental parameters and their values ​​in the regional environmental data of each area within the building from the second cloud platform, an ambiguous relationship between entities on different cloud platforms can be discovered, namely the first relationship. Based on this first relationship, a knowledge graph is generated, which integrates the data from different cloud platforms, resulting in a more comprehensive and accurate knowledge graph of the building. Based on this knowledge graph, comprehensive and accurate data can be provided quickly.

[0135] Example 2

[0136] This invention also provides a data processing method in an Internet of Things (IoT) system. The IoT system includes at least one application terminal, at least two cloud platforms, and middleware, and the data processing method is applied to the middleware.

[0137] Figure 8 This is a flowchart of a data processing method in an Internet of Things (IoT) system according to an embodiment of the present invention. Figure 8 As shown, the method includes:

[0138] Step 801: The middleware receives request data from at least one application client;

[0139] Step 802: The middleware uses a pre-generated knowledge graph to obtain feedback data in response to the request data based on the request data;

[0140] Step 803: The middleware sends the feedback data to the application.

[0141] In some embodiments, the pre-generated knowledge graph is generated using the method described in the embodiments of the present invention. The specific generation process can be referred to the description in Embodiment 1, which will not be repeated here.

[0142] In some embodiments, the middleware generates the knowledge graph through a knowledge graph generation module. Alternatively, the knowledge graph generation module may be based on a generative artificial intelligence (AI) model.

[0143] In some embodiments, the middleware uses an artificial intelligence model and a knowledge graph to generate target information corresponding to the requested data.

[0144] For example, the artificial intelligence model is a generative AI model, such as a model based on generative adversarial networks (GANs) or variational autoencoders (VAEs).

[0145] For example, the target information includes at least one of the following: missing entity information in the knowledge graph, predictive information, and suggestive information. Predictive information may include, for example, predicted device behavior, while suggestive information may include, for example, filter replacement suggestions or device usage suggestions.

[0146] In this way, middleware can also provide intelligent decision-making and personalized services to meet the customized needs of applications.

[0147] In some embodiments, the request data includes the user's query request, and correspondingly, the feedback data includes the query results in response to the query request. In this way, the middleware, based on a knowledge graph, can accurately utilize the user's query intent and provide more relevant and accurate query results.

[0148] In some embodiments, the middleware is provided with a unified transmission interface module for access interfaces of all applications. This improves data transmission efficiency, and external applications only need to set up an access mechanism to connect to one middleware, without having to connect to multiple cloud platforms, thus reducing the development burden of external applications.

[0149] In some embodiments, the middleware displays the knowledge graph; alternatively, the middleware displays the knowledge graph through the application. This allows users to understand and use the knowledge graph more intuitively, further enhancing the user experience.

[0150] Figure 9 This is an architecture diagram of an Internet of Things (IoT) system according to an embodiment of the present invention. Figure 9 As shown, the Internet of Things (IoT) system includes N application terminals, middleware, and M cloud platforms, where M and N are integers greater than 1.

[0151] like Figure 9 As shown, the device connection and parsing module of the middleware is used to connect with various cloud platforms. Through a unified interface and data protocol, it integrates various types of data, including building data, equipment data, and electricity consumption data. This module parses the data protocols and interfaces of various cloud platforms, performs data cleaning, preprocessing, structuring and other processing on the acquired data, and finally stores it in the middleware's database (DB).

[0152] The middleware's knowledge graph generation module generates a knowledge graph using data from the database. First, the data processing and analysis module uses generative AI's natural language processing techniques (such as BERT, GPT, and other pre-trained language models) to preprocess the integrated data in the database before training the algorithm model. Then, entity recognition technology is used to identify key entities in the building data, such as equipment names, models, and characteristics. Relationship extraction technology clarifies the relationships between entities, such as connections and functions between devices, i.e., extracting associations. Based on the extracted entities and relationships, a knowledge graph of the building is constructed and stored and retrieved using a graph database. The specific generation process can be found in Example 1 and will not be repeated here. Additionally, the middleware can display the generated knowledge graph through a visualization module.

[0153] After generating the knowledge graph, the middleware receives request data from the application through a unified transmission interface. The data processing module (e.g., a generative AI model) uses the generated knowledge graph to execute the aforementioned data processing methods and returns the processing results (e.g., query results) to the corresponding application through the unified transmission interface.

[0154] In some embodiments, the middleware’s unified transport interface can provide feedback data to the application through two forms: “code interface calls” and “question and answer (text)”.

[0155] Figure 10 This is a schematic diagram illustrating an example of how middleware in this invention uses a knowledge graph to provide feedback data to the application. For example... Figure 10 As shown, the middleware generates a knowledge graph based on building data from multiple cloud platforms, including the first and second cloud platforms. Feedback data can be provided to the application in two forms: "code interface calls" and "question-and-answer (text)". For "code interface calls", the middleware searches for relevant air conditioning equipment data in the knowledge graph and then returns it in a format suitable for application processing. For "question-and-answer (text)", the middleware searches for relevant air conditioning equipment data in the knowledge graph and returns intelligent recommendation results based on the application's search content.

[0156] As can be seen from the above embodiments, by using the knowledge graph generated by the method described in the embodiments of the present invention, data processing and relationship integration based on data from multiple cloud platforms can be performed in the middleware, thereby directly returning the processed data to the application, improving the response speed to the application, and reducing the data processing burden on the application and the access load on the cloud platform; furthermore, based on the generated knowledge graph, the middleware accurately understands the intent of the application user and provides more appropriate and relevant results, thereby achieving efficient connection between the application and the cloud platform.

[0157] Example 3

[0158] This invention also provides a middleware.

[0159] In some embodiments, middleware may also be referred to as middleware device, middleware apparatus, middleware, or middleware apparatus.

[0160] Figure 11 This is a schematic diagram of the middleware in an embodiment of the present invention, as shown below. Figure 11 As shown, middleware 1100 includes:

[0161] Knowledge graph generation module 1101, which is used to execute the building knowledge graph generation method described in Embodiment 1; and

[0162] The data processing module 1102 is used to execute the data processing method described in Embodiment 2.

[0163] For example, the knowledge graph generation module 1101 acquires device data from a first cloud platform and regional environmental data of various areas within a building from a second cloud platform. The device data includes device category information, device identification information, device operating parameters and their values; the regional environmental data includes regional information, regionally associated environmental parameters and their values. Based on the device data from the first cloud platform, a first set of entities is extracted. The first set of entities includes device category entities, device identification entities, and device operating parameter entities. Based on the regional environmental data of various areas within a building from the second cloud platform, a second set of entities is extracted. The second set of entities includes regional entities and environmental parameter entities. Based on the correlation between the device operating parameters and their values ​​in the device data and the environmental parameters and their values ​​in the regional environmental data, a first association relationship between the first set of entities and the second set of entities is determined. At least based on the first association relationship between the first set of entities and the second set of entities, a knowledge graph of the building is generated.

[0164] For example, the data processing module 1102 receives request data from at least one application; based on the request data, it uses a pre-generated knowledge graph to obtain feedback data in response to the request data; and sends the feedback data to the application.

[0165] In some embodiments, the functions of the above-mentioned units can be implemented with reference to the relevant steps in Embodiments 1 and 2, and will not be repeated here.

[0166] Figure 12 This is a schematic diagram of the hardware configuration of the middleware in an embodiment of this application. Figure 12As shown, the middleware 1200 may include a processor 1210 and a memory 1220; the memory 1220 is coupled to the processor 1210. The memory 1220 may store various data, including the knowledge graph 1230; it also stores an information processing program 1240, and executes the program 1240 under the control of the processor 1210 to perform the building knowledge graph generation method described in Embodiment 1 and the data processing method described in Embodiment 2.

[0167] The processor 1240 can be configured to: acquire device data from a first cloud platform and regional environmental data of various areas within a building from a second cloud platform; the device data includes device category information, device identification information, device operating parameters and their values, and the regional environmental data includes regional information, regional associated environmental parameters and their values; extract a first set of entities based on the device data from the first cloud platform; the first set of entities includes device category entities, device identification entities, and device operating parameter entities; extract a second set of entities based on the regional environmental data of various areas within the building from the second cloud platform; the second set of entities includes regional entities and environmental parameter entities; determine a first association relationship between the first set of entities and the second set of entities based on the correlation between the device operating parameters and their values ​​in the device data and the environmental parameters and their values ​​in the regional environmental data; and generate a knowledge graph of the building based at least on the first association relationship between the first set of entities and the second set of entities.

[0168] Additionally, the processor 1240 can be configured to: receive request data from at least one application; obtain feedback data in response to the request data using a pre-generated knowledge graph based on the request data; and send the feedback data to the application.

[0169] In addition, such as Figure 12 As shown, the middleware 1200 may also include a communication module 1250, etc. The communication module 1250 may include, for example, a unified transmission interface module for access by all applications and a device connection and parsing module for connecting to multiple cloud platforms; for details, please refer to the relevant descriptions in Embodiment 1 and Embodiment 2.

[0170] In addition, middleware 1200 is not necessarily required. Figure 12 All components shown; in addition, intermediate component 1200 may also include Figure 12 For components not shown, please refer to relevant technologies.

[0171] In addition, for details regarding the structure or function of the middleware, please refer to the specific content about the middleware in Embodiments 1 and 2, which will not be repeated here.

[0172] As can be seen from the above embodiments, the middleware can discover the ambiguous relationship between entities on different cloud platforms, namely the first relationship, based on the correlation between the operating parameters and their values ​​in the device data from the first cloud platform and the environmental parameters and their values ​​in the regional environmental data of each area within the building from the second cloud platform. Based on this first relationship, a knowledge graph is generated, which integrates the data from different cloud platforms, resulting in a more comprehensive and accurate knowledge graph of the building. Based on this knowledge graph, comprehensive and accurate data can be provided quickly.

[0173] Furthermore, by leveraging the knowledge graph generated using the method described in this embodiment of the invention, data processing and relationship integration based on data from multiple cloud platforms can be performed in the middleware, thereby directly returning the processed data to the application, improving the response speed to the application, and reducing the data processing burden on the application and the access load on the cloud platform; moreover, based on the generated knowledge graph, the middleware accurately understands the intent of the application user and provides more appropriate and relevant results, thereby achieving efficient linking between the application and the cloud platform.

[0174] Example 4

[0175] Embodiment 4 of the present invention provides an Internet of Things (IoT) system.

[0176] Figure 13 This is a schematic diagram of an Internet of Things (IoT) system according to an embodiment of the present invention, such as... Figure 13 As shown, the Internet of Things system 1300 includes N application terminals 1301-1 to 1301-N, middleware 1302, and M cloud platforms 1303-1 to 1303-N, where M and N are integers greater than 1.

[0177] Middleware 1302 integrates data from M cloud platforms 1303-1 to 1303-N to generate and store a knowledge graph of buildings; one or more of the N application terminals 1301-1 to 1301-N send request data through the unified transmission interface module of the middleware; based on the request data, middleware 1302 uses the knowledge graph to obtain feedback data in response to the request data; and middleware 1302 sends the feedback data to the corresponding application terminal through the unified transmission interface module.

[0178] In addition, for details regarding the structure or function of the Internet of Things (IoT) system, please refer to the specific content of the IoT system in Embodiments 1 and 2, which will not be repeated here.

[0179] As can be seen from the above embodiments, in the Internet of Things system, the middleware can discover the ambiguous relationship between entities on different cloud platforms, namely the first relationship, based on the correlation between the operating parameters and their values ​​in the device data from the first cloud platform and the environmental parameters and their values ​​in the regional environmental data of each area in the building from the second cloud platform. Based on this first relationship, a knowledge graph is generated, which integrates the data from different cloud platforms and can obtain a more comprehensive and accurate knowledge graph of the building. Based on this knowledge graph, comprehensive and accurate data can be provided quickly.

[0180] Furthermore, by leveraging the knowledge graph generated using the method described in this embodiment of the invention, data processing and relationship integration based on data from multiple cloud platforms can be performed in the middleware, thereby directly returning the processed data to the application, improving the response speed to the application, and reducing the data processing burden on the application and the access load on the cloud platform; moreover, based on the generated knowledge graph, the middleware accurately understands the intent of the application user and provides more appropriate and relevant results, thereby achieving efficient linking between the application and the cloud platform.

[0181] The apparatus and methods described above in the embodiments of the present invention can be implemented in hardware or in combination with software. The present invention relates to a computer-readable program that, when executed by a logic component, enables the logic component to implement the aforementioned apparatus or constituent parts, or to implement the various methods or steps described above.

[0182] The embodiments of the present invention also relate to storage media for storing the above programs, such as hard disks, magnetic disks, optical disks, DVDs, flash memory, etc.

[0183] This invention also provides a computer-readable program, wherein when the program is executed in middleware, the program causes the middleware to perform the above-described method for generating a building knowledge graph according to this invention.

[0184] This invention also provides a computer-readable program, wherein when the program is executed in middleware, the program causes the middleware to perform the data processing method described in this invention.

[0185] This invention also provides a computer-readable program product comprising a computer program, wherein when the program is executed in middleware, the program causes the middleware to perform the above-described method for generating a building knowledge graph according to this invention.

[0186] This invention also provides a computer-readable program product comprising a computer program, wherein when the program is executed in middleware, the program causes the middleware to perform the data processing method described in this invention.

[0187] This invention also provides a storage medium for a computer-readable program, wherein when the program is executed in middleware, the program causes the middleware to perform the above-described method for generating a building knowledge graph according to this invention.

[0188] This invention also provides a storage medium for a computer-readable program, wherein when the program is executed in middleware, the program causes the middleware to perform the data processing method described in this invention.

[0189] It should be noted that the limitations of each step involved in the embodiments of the present invention are not considered as limiting the order of steps without affecting the implementation of the specific solution. The steps listed first can be executed first, or they can be executed later, or they can even be executed simultaneously. As long as the embodiments of the present invention can be implemented, they should be considered to fall within the protection scope of the present invention.

[0190] The present invention has been described above with reference to specific embodiments. However, those skilled in the art should understand that these descriptions are exemplary and not intended to limit the scope of protection of the present invention. Those skilled in the art can make various modifications and variations to the present invention based on its spirit and principles, and these modifications and variations are also within the scope of the present invention.

Claims

1. A method for generating a building knowledge graph, characterized in that, The method includes: Acquire device data from a first cloud platform and regional environmental data for each area within a building from a second cloud platform; the device data includes device category information, device identification information, device operating parameters and their values, and the regional environmental data includes regional information, regionally associated environmental parameters and their values; Based on the device data from the first cloud platform, a first group of entities is extracted; the first group of entities includes a device category entity, a device identifier entity, and a device operating parameter entity. Based on the regional environmental data of each area within the building from the second cloud platform, a second set of entities is extracted; the second set of entities includes regional entities and environmental parameter entities; Based on the correlation between the equipment operating parameters and their values ​​in the equipment data and the environmental parameters and their values ​​in the regional environmental data, a first association relationship is determined between the first group of entities and the second group of entities; and A knowledge graph of the building is generated based at least on the first association relationship between the first group of entities and the second group of entities.

2. The method according to claim 1, characterized in that, The method further includes: Based on the Natural Language Processing (NLP) model, data from the first cloud platform and data from the second cloud platform are preprocessed; Entity recognition is performed on the preprocessed data; Extract the second association relation that represents the relationship between entities. At least based on the first association relationship between the first group of entities and the second group of entities, a knowledge graph of the building is generated, including: The knowledge graph is generated based on the identified entities, the extracted second relationship, and the first relationship.

3. The method according to claim 1, characterized in that, The equipment operating parameters in the equipment data include at least one of the equipment's set air quality parameters and equipment energy consumption parameters; The environmental parameters in the regional environmental data include at least one of regional air quality parameters, regional energy consumption parameters, and regional building parameters. The first association between the first group of entities and the second group of entities includes the association between the device category entity and the device identifier entity in the first group of entities and the region entity in the second group of entities.

4. The method according to claim 3, characterized in that, The air quality parameters include at least one of temperature and humidity. The energy consumption parameters include power consumption. The regional building parameters include the regional heat load.

5. The method according to claim 3 or 4, characterized in that, Based on the correlation between the equipment operating parameters and their values ​​in the equipment data and the environmental parameters and their values ​​in the regional environmental data, a first association relationship is determined between the first group of entities and the second group of entities, including: Based on the similarity between the first relationship between the value and time of the set air quality parameter and the second relationship between the value and time of the regional air quality parameter, determine the association between the equipment category entity and equipment identifier entity corresponding to the set air quality parameter and the regional entity corresponding to the regional air quality parameter; or, Based on the similarity between the numerical value and time of the device energy consumption parameter and the numerical value and time of the regional energy consumption parameter, determine the association between the device category entity and device identifier entity corresponding to the device energy consumption parameter and the regional entity corresponding to the regional energy consumption parameter; or, Based on the similarity between the third relationship of the device energy consumption parameter values ​​and time and the fifth relationship of the regional building parameters values ​​and time, the association relationship between the device category entity and the device identifier entity corresponding to the device energy consumption parameter and the regional entity corresponding to the regional building parameter is determined.

6. The method according to claim 1, characterized in that, The method is executed by middleware in the Internet of Things system.

7. A data processing method in an Internet of Things (IoT) system, the IoT system comprising at least one application terminal, at least two cloud platforms, and middleware, the method being applied to the middleware. Its features are, The method includes: The middleware receives request data from at least one application client; The middleware uses a pre-generated knowledge graph to obtain feedback data in response to the request data, wherein the knowledge graph is generated by the method of any one of claims 1-6; The middleware sends the feedback data to the application.

8. The method according to claim 7, characterized in that, The method further includes: The middleware uses an artificial intelligence model and the knowledge graph to generate target information corresponding to the request data.

9. The method according to claim 8, characterized in that, The target information includes at least one of the following: missing entity information, prediction information, and suggestion information in the knowledge graph.

10. The method according to claim 7, characterized in that, The request data includes the user's query request. The feedback data includes the query results in response to the query request.

11. The method according to any one of claims 7-10, characterized in that, The middleware includes a unified transmission interface module, which serves as the access interface for all application terminals.

12. The method according to any one of claims 7-10, characterized in that, The middleware displays the knowledge graph; or... The middleware is used to display the knowledge graph through applications.

13. A middleware, characterized in that, The middleware includes: A knowledge graph generation module, which is used to perform the method described in any one of claims 1-6.

14. The middleware according to claim 13, characterized in that, The middleware also includes: A data processing module for performing the method described in any one of claims 7-12.

15. An Internet of Things (IoT) system, characterized in that, The Internet of Things system includes at least the middleware described in claim 13 or 14.

Images