Internet of Things equipment management method and device, equipment and medium

Abstract

The invention discloses an Internet of Things equipment management method and device, equipment and a medium. The Internet of Things equipment management method comprises the following steps: acquiring equipment information of equipment to be accessed; analyzing the equipment information by utilizing a priori concept of an Internet of Things knowledge base on the basis of a large language model, generating a field label of the equipment information, mapping the field label to a corresponding concept node in a pre-constructed Internet of Things knowledge graph, and performing semantic binding to realize equipment access; wherein the knowledge graph of the internet of things is a semantic network structure constructed by equipment information, abstract concepts of equipment, events and relations of the events. According to the method, the knowledge graph and the large language model technology are combined, the dynamic semantic access management of the Internet of Things equipment is realized, the access management efficiency of different equipment can be improved, the expandability is relatively high, and the maintenance cost is relatively low.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and Internet of Things (IoT) technology, and in particular to an IoT device management method, apparatus, device, and medium. Background Technology

[0002] As the Internet of Things (IoT) continues to expand, the types of connected devices are becoming increasingly diverse, with a large number of manufacturers, heterogeneous protocols, and inconsistent semantics. This leads to increasing difficulties in data interaction and device management between different fields and platforms within IoT systems. Current technologies typically use device gateways or middleware to implement device access through manually predefined data models or manual configuration rules. While this meets basic device access requirements, it relies heavily on manual implementation. Given the wide variety of IoT device types and the heterogeneity of protocols, device access management is inefficient, has high maintenance costs, and poor scalability. Summary of the Invention

[0003] This invention provides a method, apparatus, device, and medium for managing Internet of Things (IoT) devices, in order to solve the technical problems of low efficiency, high maintenance costs, and poor scalability caused by the reliance on manual intervention for the access management of different types of devices in existing IoT systems.

[0004] To address the aforementioned technical problems, firstly, an IoT device management method is provided, comprising: Obtain device information for the device to be connected; Based on a large language model, the device information is parsed using prior concepts from the IoT knowledge base to generate field tags for the device information. These field tags are then mapped to corresponding concept nodes in a pre-built IoT knowledge graph for semantic binding, enabling device access. The IoT knowledge graph is a semantic network structure built from device information, abstract concepts of devices, events, and their relationships.

[0005] Secondly, an Internet of Things (IoT) device management apparatus is provided, including a unit for performing the above-described IoT device management method.

[0006] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described Internet of Things (IoT) device management method.

[0007] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described Internet of Things (IoT) device management method.

[0008] In the aforementioned IoT device management methods, devices, equipment, and media, the device information of the device to be accessed is parsed using prior concepts from the IoT knowledge base based on a large language model. Field tags for the device information are generated, and these field tags are mapped to corresponding concept nodes in a pre-built IoT knowledge graph for semantic binding, thus enabling device access. The IoT knowledge graph is a semantic network structure built from device information, abstract concepts of devices, events, and their relationships. This invention combines the semantic understanding capabilities of a large language model with the unified modeling technology of a knowledge graph, employing a unified knowledge representation framework. This allows for semantic association of information such as devices and events from multiple vendors, protocols, and networks, effectively breaking down data silos and enabling "plug-and-play" semantic access for different devices. It improves the efficiency of access management for different devices, has high scalability, and low maintenance costs. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating the IoT device management method in the first embodiment of the present invention.

[0010] Figure 2 yes Figure 1 The diagram shows the specific process flow of S140 in the IoT device management method.

[0011] Figure 3 yes Figure 2 The diagram shows the specific process flow of S141 in the IoT device management method.

[0012] Figure 4 This is a flowchart illustrating the IoT device management method in the second embodiment of the present invention.

[0013] Figure 5 yes Figure 4 The diagram shows the specific process flow of S260 in the IoT device management method.

[0014] Figure 6 This is a schematic diagram of the structure of the Internet of Things (IoT) device management device provided in an embodiment of the present invention.

[0015] Figure 7 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0018] It should also be understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0019] Please see Figure 1 , Figure 1 This is a schematic flowchart illustrating an IoT device management method according to a first embodiment of the present invention. In the embodiment shown in the figures, the IoT device management method includes the following steps S110-140: S110. Based on the large language model, the device information of the connected devices in the Internet of Things system is parsed to obtain all device-related entities, abstract concepts, running events and their semantic relationships.

[0020] In this step, semantic understanding technology using a large language model is used to automatically parse the meaning of fields in the device information, obtaining all device-related entities, abstract concepts, operational events, and their semantic relationships. Device information includes core entity information such as device ID, device type (e.g., IoT device or gateway), device product type (e.g., temperature sensor or humidity sensor), protocols used by the device, and third-party usage information, as well as device operational information. Third-party usage information includes users, tenants, and third-party applications. Abstract concepts include device type classification (e.g., abstract classification of IoT device types, such as temperature and humidity sensors, door locks, and cameras), device product type classification (e.g., abstract classification of temperature and humidity sensor types, such as environmental monitoring devices), device usage protocol classification, and event types. Operational events include device online / offline status, alarms, command issuance, and remote upgrades.

[0021] In this invention, an extensible JSON Schema description structure can be used, combined with mapping rules, to semantically unify and abstractly represent heterogeneous data of device information from different protocols, manufacturers, and business domains, thereby achieving standardized data management across scenarios and device types. In this embodiment, a concept table can also be pre-established to store abstract concept names, categories, descriptions, and mapping rules. The mapping rules can be determined based on a rule engine. The concept table can be stored in a local relational database, and integrity and consistency checks are performed when new or modified abstract concepts are added to ensure semantic uniformity within the system. Furthermore, a mapping relationship between abstract concepts and core entities can be established based on entity-concept mapping rules, achieving entity-concept association management. The entity-concept mapping rules are association rules determined based on the semantic relationship between device-related entities and abstract concepts, supporting multiple entity fields pointing to the same concept, facilitating subsequent cross-entity and cross-domain semantic retrieval. Understandably, event data can also be uniformly processed into a JSON Schema structure to ensure unified parsing and management of event data across devices and protocols.

[0022] Furthermore, in this invention, the relational database can also establish relational tables to store various predefined relation types used to describe the semantic relationships between devices, events, and abstract concepts. Relationship types can include affiliation relationships (devices belong to a product category or organizational unit), product composition relationships (such as the composition relationship between components or modules and complete products), protocol usage relationships, access gateway relationships (the access relationship between edge devices and gateway nodes), event generation relationships (the types of events generated by the device, such as device online / offline, alarms, remote upgrades, and control command execution), and access relationships (the device being accessed by a specific user, application, or service). Understandably, the relational tables can also store the mapping relationships between core entities and abstract concepts and events related to device information.

[0023] S120. Using all device-related entities, abstract concepts, and operational events as nodes, establish edges for all nodes based on the semantic relationships of all device-related entities, abstract concepts, and operational events to obtain an IoT knowledge graph.

[0024] In this invention, the IoT knowledge graph is a semantic network structure built from device information, abstract concepts of devices, events, and their relationships. During creation, high-level concept nodes can be created, and these nodes, along with entity nodes and event nodes, form a complete knowledge graph network through semantic relationships. In this embodiment, events such as device online / offline status, alarms, command issuance, and remote upgrades are uniformly incorporated into the graph network, providing a structured semantic foundation for subsequent cross-domain semantic retrieval and dynamic device access.

[0025] Specifically, the IoT knowledge graph is constructed using core entity information, abstract concepts, and operational events such as device ID, device type, device product type, device protocol, and third-party usage information of the device as nodes. All nodes of core entities and event nodes constitute device knowledge nodes. The semantic relationships between device-related entities, abstract concepts, and operational events are used to construct the IoT knowledge graph. Each edge of the IoT knowledge graph can carry attributes (event time, source system, mapping rules, etc.).

[0026] Preferably, information such as node ID, node type, relationship type, edge attributes, and timestamps in the IoT knowledge graph can also be stored in a database relation table to ensure the integrity and consistency of relation data.

[0027] S130. Obtain device information of the device to be connected.

[0028] In this step, when a new device connects to the IoT system, the device information of the device to be connected is obtained. The device information can be presented in JSON Schema format or in the form of a data packet.

[0029] S140. Based on the large language model, the prior concepts of the IoT knowledge base are used to parse the device information, generate field tags for the device information, and map the field tags to the corresponding concept nodes in the pre-built IoT knowledge graph for semantic binding to realize device access.

[0030] In this step, the IoT knowledge base includes the aforementioned relational database, and the prior concepts can be information related to abstract concepts pre-stored in the relational database.

[0031] Specifically, such as Figure 2 As shown, this step includes S141-S142: S141. Based on the large language model, the prior concepts of the Internet of Things knowledge base are used to perform entity recognition and matching on the device information, and field labels containing conceptual entities are generated.

[0032] Specifically, such as Figure 3 As shown, this step includes S1411-S1412: S1411. Based on the large language model, perform semantic understanding and entity recognition on the device information to obtain entity fields.

[0033] S1412. Perform multi-dimensional semantic matching between the entity fields and the prior concepts of the Internet of Things knowledge base to generate field labels containing entities.

[0034] For steps S1411-S1412 above, based on the semantic understanding and entity recognition capabilities of the Large Language Model (LLM), semantic analysis is performed on the device information to achieve field-level named entity extraction and automatic classification. The entity fields are matched with the prior concepts of the IoT knowledge base to obtain field labels containing abstract concept entities.

[0035] S142. Create a device knowledge node based on the entity field, search for the concept node corresponding to the concept entity labeled by the field in the pre-built IoT knowledge graph, and establish a semantic binding between the device knowledge node and the concept node to realize device access.

[0036] In this step, the conceptual entities of the field labels are mapped to the corresponding conceptual nodes in the IoT knowledge graph. Core entity nodes and event nodes are created based on the entity fields obtained from the device information and semantically bound to the conceptual nodes to enable the access of new devices.

[0037] As can be seen from the above scheme, the IoT device management method in this embodiment utilizes the semantic understanding capabilities of large language models and the unified modeling technology of knowledge graphs to achieve adaptive construction of device semantic information. It can semantically associate information such as devices and events from multiple manufacturers, protocols, and networks, solving the problem of semantic inconsistency across manufacturers, protocols, and systems. New devices can be connected without manual configuration, achieving plug-and-play functionality. It can achieve rapid integration of large-scale IoT devices, reduce manual modeling and access costs, lower the barrier to device access and long-term operation and maintenance costs, improve the scalability of IoT systems, and can be widely applied in smart cities, industrial IoT, energy, transportation and other scenarios to promote semantic interconnection of devices.

[0038] It should be noted that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0039] Reference Figure 4 , Figure 4 This is a flowchart illustrating the IoT device management method provided in the second embodiment of the present invention. As shown in the figure, the method includes the following steps S210-S270: S210. Based on the large language model, the device information of the connected devices in the Internet of Things system is parsed to obtain all device-related entities, abstract concepts, running events and their semantic relationships.

[0040] This step is similar to step S110, and will not be described again here.

[0041] S220. Using all device-related entities, abstract concepts, and operational events as nodes, establish edges for all nodes based on the semantic relationships of all device-related entities, abstract concepts, and operational events to obtain an IoT knowledge graph.

[0042] This step is similar to step S120, and will not be described again here.

[0043] S230, Obtain device information of the device to be connected.

[0044] This step is similar to step S130, and will not be described again here.

[0045] S240. Based on the large language model, the prior concepts of the IoT knowledge base are used to parse the device information, generate field tags for the device information, and map the field tags to the corresponding concept nodes in the pre-built IoT knowledge graph for semantic binding, thereby realizing device access.

[0046] This step is similar to step S140, and will not be described again here.

[0047] S250, Receive retrieval information.

[0048] In this step, the retrieval information includes device information, functions, and / or semantic tags; wherein, the functions are the functions that the device can perform, such as linkage triggering, remote control, data sharing, health collaboration, etc., and the semantic tags are tags related to business or application scenarios, such as "health monitoring", "security management", "environmental control", which facilitate semantic-level query.

[0049] S260. Based on the large language model, perform semantic retrieval in the pre-constructed IoT knowledge graph according to the retrieval information to obtain the nodes and semantic relationship information that match the retrieval information.

[0050] In this invention, when searching in the Internet of Things knowledge graph, the scope of influence or collaboration chain can be queried along the relationship chain between devices based on the search information. For example, a smoke sensor triggers an alarm on a smart door lock or a voice controller, thus achieving accurate cross-domain retrieval.

[0051] Specifically, such as Figure 5 As shown, this step includes S261-S262: S261. Based on the large language model, perform semantic embedding processing on the retrieved information to generate query semantic information, and based on the large language model, perform semantic embedding processing on each node and its relationship in the Internet of Things knowledge graph to generate graph embedding information.

[0052] In this invention, the retrieval information can be input in natural language and embedded based on a large language model to obtain query semantic information, thereby realizing the semantic mapping from natural language to knowledge graph. For example, if the retrieval information is "Which devices can collaborate with smartwatches in health monitoring scenarios?", the large language model will parse it into query semantic information: the target device includes "smartwatch 001", the relationship type includes "collaboration", and the semantic tag includes "health".

[0053] S262. Based on the query semantic information, a semantic similarity matching method is used to perform semantic retrieval in the graph embedding information of the Internet of Things knowledge graph to obtain the node with the highest semantic similarity to the query semantic information and semantic relationship information.

[0054] For steps S261-S262 above, embedding processing is performed based on a large language model to obtain vectorized query semantic information. Furthermore, each node and its relationship in the IoT knowledge graph can be embedded to obtain vectorized graph embedding information. The most similar nodes and semantic relationship information are found in the graph embedding information through semantic similarity matching, thereby obtaining the most relevant devices and relationship chains. This enables cross-domain semantic retrieval of IoT devices, supports real-time discovery, collaboration, and linkage, meets the security and efficient management needs in complex network environments, and also supports multi-dimensional path queries, comprehensively improving intelligent operation and maintenance and decision support capabilities.

[0055] S270. When device information changes or an event occurs, the nodes and relationships of the IoT knowledge graph are updated using hot-loading technology via Kafka message queues.

[0056] In this invention, event triggers can be used to monitor the dynamic operating status of connected IoT devices. When an event occurs or device information changes (such as adding a new protocol or changing the protocol), the new or changed information or event can be captured through the Kafka message queue to achieve asynchronous transmission and decoupling. Hot loading technology is used to realize the synchronous update of incremental data in the IoT knowledge graph. During synchronous updates, batch import and incremental updates are supported to ensure that the nodes and relationships in the IoT knowledge graph remain up-to-date.

[0057] Understandably, at the initial stage of this invention, all existing core entities, abstract concepts, events and their relationships stored in the database can be fully imported into the Neo4j graph database to establish corresponding nodes and edges, thereby constructing an IoT knowledge graph. This step S270 can be executed after the IoT knowledge graph is constructed, or it can be executed after a new device is connected. That is, when there are subsequent additions or changes to devices, events or concepts, they can be asynchronously and incrementally synchronized to the IoT knowledge graph through Kafka.

[0058] As can be seen, in the above scheme, the IoT device management method of this embodiment can realize dynamic semantic access and cross-domain retrieval of IoT devices, improve the efficiency of device data access and the level of query intelligence. That is, through a unified IoT knowledge graph, it can solve the semantic inconsistency problem across manufacturers, protocols and systems, and support cross-domain, multi-dimensional queries and device status analysis. In the future, it can provide more accurate alarm, event tracing and trend prediction capabilities. Moreover, it uses big language model semantic understanding technology to realize the automatic parsing of the field meaning of device information when the device is connected. Combined with prior concepts and the constructed IoT knowledge graph, semantic mapping and binding can be performed. New device access does not require manual configuration, which can support the rapid integration of large-scale IoT devices, reduce the cost of manual modeling and docking, and also provide hot loading and incremental update mechanisms to ensure the high availability and flexible operation and maintenance of the system. It can support large-scale distributed deployment and further improve the scalability of the IoT system.

[0059] It should be understood that the above method embodiments are described as a series of actions for the sake of simplicity. However, those skilled in the art should know that the present invention is not limited by the described order of actions. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0060] Reference Figure 6 , Figure 6 This is a flowchart illustrating an IoT device management apparatus provided in an embodiment of the present invention. In the embodiment shown in the figures, the IoT device management apparatus includes a semantic parsing unit 301, a graph construction unit 302, an acquisition unit 303, an access management unit 304, a retrieval unit 305, and an incremental update unit 306. Detailed descriptions of each functional unit are as follows: The semantic parsing unit 301 is used to parse the device information of the connected devices in the Internet of Things system based on the large language model, and obtain all device-related entities, abstract concepts, running events and their semantic relationships. The graph construction unit 302 is used to establish edges of all nodes based on the semantic relationships of all device-related entities, abstract concepts and running events, respectively, to obtain an Internet of Things knowledge graph. Acquisition unit 303 is used to acquire device information of the device to be connected; Access management unit 304 is used to parse the device information based on the prior concepts of the IoT knowledge base using a large language model, generate field tags for the device information, and map the field tags to the corresponding concept nodes in the pre-built IoT knowledge graph for semantic binding to realize device access; The acquisition unit 303 is also used to receive retrieval information; The retrieval unit 305 is used to perform semantic retrieval in a pre-constructed Internet of Things knowledge graph based on the large language model and the retrieval information to obtain node and semantic relationship information that matches the retrieval information; The incremental update unit 306 is used to update the nodes and relationships of the Internet of Things knowledge graph using hot-loading technology through the Kafka message queue when device information changes or an event occurs.

[0061] In one embodiment, the access management unit 304 is specifically used for: Based on the large language model, the prior concepts of the Internet of Things knowledge base are used to perform entity recognition and matching on the device information, and field labels containing conceptual entities are generated. Based on the device information, a device knowledge node is created, and a concept node corresponding to the concept entity of the field label is searched in the pre-built IoT knowledge graph. The semantic binding between the device knowledge node and the concept node is established to realize device access.

[0062] In one embodiment, the access management unit 304 is further specifically used for: Based on a large language model, semantic understanding and entity recognition are performed on the device information to obtain entity fields; The entity fields are subjected to multidimensional semantic matching with the prior concepts of the IoT knowledge base to generate field labels containing the entities.

[0063] In one embodiment, the retrieval unit 305 is specifically used for: The retrieval information is semantically embedded based on a large language model to generate query semantic information, and the nodes and their relationships in the Internet of Things knowledge graph are semantically embedded based on the large language model to generate graph embedding information. Based on the query semantic information, a semantic similarity matching method is used to perform semantic retrieval in the graph embedding information of the Internet of Things knowledge graph to obtain the node with the highest semantic similarity to the query semantic information and semantic relationship information.

[0064] As can be seen, the present invention provides an IoT device management device that, by combining knowledge graph and large language model technology, realizes dynamic semantic access and cross-domain intelligent retrieval of IoT devices, thereby improving the efficiency of device data access and the level of query intelligence.

[0065] For specific limitations regarding the IoT device management device, please refer to the limitations on the IoT device management method mentioned above, which will not be repeated here. Each unit in the aforementioned IoT device management device can be implemented entirely or partially through software, hardware, or a combination thereof. These units can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each unit.

[0066] In one embodiment, a computer device is provided, the internal structure of which can be shown as follows: Figure 7 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external devices via a network connection. Furthermore, the computer device may also include a display screen and input devices (e.g., mouse, keyboard, etc.) for interactive purposes.

[0067] Specifically, when the processor in the computer device executes the computer program, it implements each step of the Internet of Things device management method provided in the first and second embodiments described above.

[0068] In one embodiment, the present invention may also provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the various steps of the Internet of Things device management method provided in the first and second embodiments described above.

[0069] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0070] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0071] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. An Internet of Things device management method, characterized by, The Internet of Things device management method comprises: obtaining device information of a device to be accessed; based on the prior concept of the Internet of Things knowledge base, the device information is parsed by using a large language model to generate field labels of the device information, and the field labels are mapped to corresponding concept nodes in a pre-constructed Internet of Things knowledge graph for semantic binding to realize device access; wherein the Internet of Things knowledge graph is a semantic network structure constructed based on device information, abstract concepts of devices, events and their relationships. 2.The IoT device management method of claim 1, wherein, The device information is parsed based on the prior concept of the Internet of Things knowledge base by using a large language model to generate field labels of the device information, and the field labels are mapped to corresponding concept nodes in a pre-constructed Internet of Things knowledge graph for semantic binding, specifically including: based on the prior concept of the Internet of Things knowledge base, the device information is parsed by using a large language model to generate field labels containing concept entities; According to the device information, a device knowledge node is created, and a corresponding concept node of the field label concept entity is searched in the pre-constructed Internet of Things knowledge graph, and semantic binding between the device knowledge node and the concept node is established to realize device access.

3. The IoT device management method of claim 2, wherein, The device information is parsed based on the prior concept of the Internet of Things knowledge base by using a large language model to generate field labels containing concept entities, specifically including: based on the large language model, the semantic understanding and entity recognition of the device information are obtained; The entity field is matched with the prior concept of the Internet of Things knowledge base in multiple dimensions to generate a field label containing entities. 4.The IoT device management method of claim 1, wherein, The Internet of Things device management method further comprises: based on the large language model, the device information of the accessed devices in the Internet of Things system is parsed to obtain all device-related entities, abstract concepts, running events and their semantic relationships; wherein the abstract concepts include device type classification, device product type classification, device use protocol classification and event type; all device-related entities include device ID, device type, device product type, device use protocol and third-party use information of the device; all nodes are connected based on the semantic relationship of all device-related entities, abstract concepts and running events to obtain an Internet of Things knowledge graph. 5.The IoT device management method of claim 1, wherein, The Internet of Things device management method further comprises: receiving search information; wherein the search information includes device information, functions and / or semantic labels; based on the large language model, the semantic search is performed in the pre-constructed Internet of Things knowledge graph according to the search information to obtain nodes and semantic relationship information matched with the search information.

6. The IoT device management method of claim 5, wherein, The semantic search is performed in the pre-constructed Internet of Things knowledge graph according to the search information based on the large language model to obtain nodes and semantic relationship information matched with the search information, specifically including: The big language model is used for semantic embedding processing on the search information to generate query semantic information, and the big language model is used for semantic embedding processing on each node and the relationship of the Internet of Things knowledge graph to generate graph embedding information; According to the query semantic information, a semantic similarity matching method is used for semantic retrieval in the graph embedding information of the Internet of Things knowledge graph to obtain node and semantic relationship information with the highest semantic similarity to the query semantic information.

7. The IoT device management method of claim 1, wherein, After the device access is implemented, the method further includes: When device information changes or an event occurs, the nodes and relationships of the Internet of Things knowledge graph are updated by using a hot loading technology through a Kafka message queue.

8. An Internet of Things device management apparatus characterized by comprising: The method comprises a unit for executing the Internet of Things device management method according to any one of claims 1-7.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the steps of the Internet of Things device management method according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to implement the steps of the Internet of Things device management method according to any one of claims 1-7.

Images