Information fusion method and device, computer equipment and storage medium

Abstract

The invention relates to an information fusion method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring first attribute information of each electrical device in a target building in a low-voltage power distribution system, second attribute information of each electrical device in a BMS (Battery Management System), and a device connection relationship between the electrical devices; according to the first attribute information and the second attribute information of each electrical device, determining a similarity feature between the first attribute information and the second attribute information of each electrical device; according to the first attribute information and the second attribute information of each piece of electrical equipment and the equipment topological structure diagram of the target building, determining attribute characteristics of each piece of electrical equipment; and performing information fusion on the first attribute information and the second attribute information according to the similarity feature between the first attribute information and the second attribute information of each electrical device and the attribute feature of each electrical device. By adopting the method, the information fusion and attribute mapping process of each electrical device in the target building in the low-voltage power distribution system and the BMS can be accurately completed.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to an information fusion method, apparatus, computer device, and storage medium. Background Art

[0002] With the rapid development of intelligent buildings, the construction industry is facing a transformation from the traditional high-energy consumption mode to the direction of intelligence and greenness. The core objectives of intelligent buildings include improving equipment management efficiency and optimizing system operation performance, especially achieving refined control through the collaboration of multiple subsystems. For this purpose, the intelligent building automation system must be closely integrated with the low-voltage power distribution and utilization system and the building management system. By optimizing the device identification mapping, unified management of distributed devices and adjustable loads can be achieved, maximizing the operation efficiency and stability of the system.

[0003] Currently, before information fusion, entity disambiguation needs to be performed on the data to be fused. Entity disambiguation is a key technology for mapping named entities in text to a known and unambiguous structured knowledge base and is one of the fundamental researches in the field of natural language processing. In the context of intelligent buildings, the accuracy of entity disambiguation directly affects the optimized management of building equipment identification, attribute association, and system collaboration. Currently, entity disambiguation technology plays an important role in fields such as knowledge base construction, information retrieval, machine translation, and topic tracking. However, existing methods often only focus on single entity references and ignore the impact of device topology relationships and the similarity between candidate entities on the disambiguation results. Summary of the Invention

[0004] Based on this, it is necessary to provide an information fusion method, apparatus, computer device, and storage medium for the above technical problems.

[0005] In a first aspect, this application provides an information fusion method, including:

[0006] Obtain the first attribute information of each electrical device in the low-voltage power distribution system within the target building, the second attribute information of each electrical device in the building management system (BMS), and the device connection relationship between each electrical device; wherein, the device identifiers of each electrical device in the low-voltage power distribution system and the BMS are the same;

[0007] Determine the similarity feature between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device;

[0008] Determine the attribute feature of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology structure diagram of the target building;

[0009] Fuse the first attribute information and the second attribute information of each electrical device according to the similarity characteristics between the first attribute information and the second attribute information of each electrical device and the attribute characteristics of each electrical device.

[0010] In one embodiment, the first attribute information of each electrical device includes the first device name and the first device characteristics of the electrical device, the second attribute information of each electrical device includes the second device name and the second device characteristics of the electrical device, and the similarity characteristics include syntactic similarity characteristics and semantic similarity characteristics; determining the similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device includes:

[0011] Perform word segmentation on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device;

[0012] Perform word segmentation on the first device characteristics and the second device characteristics of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device;

[0013] Determine the syntactic similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device;

[0014] Determine the semantic similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device.

[0015] In one embodiment, determining the syntactic similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device includes:

[0016] For each electrical device, take the string length of the first device word segmentation of the electrical device as the first length;

[0017] Take the string length of the second device word segmentation of the electrical device as the second length;

[0018] Determine the edit distance between the first device name and the second device name of the electrical device according to the first length and the second length;

[0019] Determine the syntactic similarity characteristics between the first attribute information and the second attribute information of the electrical device according to the edit distance.

[0020] In one embodiment, determining the semantic similarity feature between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device includes:

[0021] For each electrical device, using the bag-of-words model, representing the first feature word segmentation of the electrical device as a vector to obtain the first attribute vector of the electrical device;

[0022] Using the bag-of-words model, representing the second feature word segmentation of the electrical device as a vector to obtain the second attribute vector of the electrical device;

[0023] Determining the semantic similarity feature between the first attribute information and the second attribute information of the electrical device according to the cosine similarity between the first attribute vector and the second attribute vector of the electrical device.

[0024] In one embodiment, the first attribute information of each electrical device includes the first device name and the first device feature of the electrical device, the second attribute information of each electrical device includes the second device name and the second device feature of the electrical device, and the attribute features include relevance features and importance features;

[0025] Determining the attribute features of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology structure diagram of the target building includes:

[0026] Performing word segmentation processing on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device;

[0027] Performing word segmentation processing on the first device feature and the second device feature of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device;

[0028] According to the device topology structure diagram of the target building, determining the position information of each electrical device in the target building and the device connection relationship between each electrical device;

[0029] Determining the relevance feature of each electrical device according to the device connection relationship between each electrical device and the first feature word segmentation and the second feature word segmentation of each electrical device;

[0030] Using the PageRank algorithm, determining the importance feature of each electrical device according to the position information of each electrical device in the target building and the first feature word segmentation and the second feature word segmentation of each electrical device.

[0031] In one embodiment, determining the relevance features of each electrical device according to the device connection relationship between each electrical device, and the first feature segmentation and the second feature segmentation of each electrical device includes:

[0032] Construct an equipment association knowledge base of the target building according to the device connection relationship between each electrical device, and the first feature segmentation and the second feature segmentation of each electrical device;

[0033] Determine the relevance features of each electrical device according to the equipment association knowledge base.

[0034] In one embodiment, performing information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity feature between the first attribute information and the second attribute information of each electrical device and the attribute feature of each electrical device includes:

[0035] For any electrical device, determine the first weight of the similarity feature between the first attribute information and the second attribute information of the electrical device, and the second weight of the attribute feature of the electrical device;

[0036] Take the product of the first weight of the electrical device and the similarity feature as the first intermediate parameter;

[0037] Take the product of the second weight of the electrical device and the attribute feature as the second intermediate parameter;

[0038] Take the sum between the first intermediate parameter and the second intermediate parameter as the comprehensive feature of the electrical device;

[0039] Perform information fusion on the first attribute information and the second attribute information of each electrical device according to the comprehensive feature of the electrical device.

[0040] In a second aspect, the present application also provides an information fusion device, including:

[0041] An information acquisition module, configured to acquire the first attribute information of each electrical device in the low-voltage power distribution system in the target building, the second attribute information of each electrical device in the building management system BMS in the target building, and the device connection relationship between each electrical device; wherein, the device identifiers of each electrical device in the low-voltage power distribution system and the BMS are the same;

[0042] A first determination module, configured to determine the similarity feature between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device;

[0043] A second determination module, configured to determine the attribute characteristics of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology diagram of the target building;

[0044] An information fusion module, configured to perform information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity characteristics between the first attribute information and the second attribute information of each electrical device and the attribute characteristics of each electrical device.

[0045] In a third aspect, the present application further provides a computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0046] Obtain the first attribute information of each electrical device in the low-voltage power distribution system in the target building, the second attribute information of each electrical device in the building management system BMS in the target building, and the device connection relationship between each electrical device; wherein, the device identifiers of each electrical device in the low-voltage power distribution system and the BMS are the same;

[0047] Determine the similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device;

[0048] Determine the attribute characteristics of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology diagram of the target building;

[0049] Perform information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity characteristics between the first attribute information and the second attribute information of each electrical device and the attribute characteristics of each electrical device.

[0050] In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0051] Obtain the first attribute information of each electrical device in the low-voltage power distribution system in the target building, the second attribute information of each electrical device in the building management system BMS in the target building, and the device connection relationship between each electrical device; wherein, the device identifiers of each electrical device in the low-voltage power distribution system and the BMS are the same;

[0052] Determine the similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device;

[0053] Determine the attribute characteristics of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology diagram of the target building;

[0054] Perform information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity characteristics between the first attribute information and the second attribute information of each electrical device and the attribute characteristics of each electrical device.

[0055] In a fifth aspect, the present application also provides a computer program product, including a computer program, which when executed by a processor implements the following steps:

[0056] Obtain the first attribute information of each electrical device in the low-voltage power distribution system in the target building, the second attribute information of each electrical device in the building management system BMS in the target building, and the device connection relationship between each electrical device; wherein, the device identifiers of each electrical device in the low-voltage power distribution system and the BMS are the same;

[0057] Determine the similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device;

[0058] Determine the attribute characteristics of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology diagram of the target building;

[0059] Perform information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity characteristics between the first attribute information and the second attribute information of each electrical device and the attribute characteristics of each electrical device.

[0060] After obtaining the first attribute information of each electrical device in the low-voltage power distribution system in the target building and the second attribute information in the BMS, and the device connection relationship between each electrical device, the above information fusion method, device, computer device and storage medium determine the similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device; since the similarity characteristics can accurately map the difference degree between the first attribute information and the second attribute information, it can ensure accurate disambiguation of the first attribute information and the second attribute information during the information fusion process; in addition, according to the first attribute information and the second attribute information of each electrical device, and the device topology diagram of the target building, determine the attribute characteristics of each electrical device; since the attribute characteristics can characterize the association characteristics between each electrical device, therefore, the attribute mapping process in the subsequent information fusion process can be accurately realized, and therefore, the first attribute information and the second attribute information of each electrical device can be accurately and flexibly fused. Description of the Drawings

[0061] To more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0062] Figure 1 It is an application environment diagram of the information fusion method in an embodiment;

[0063] Figure 2 It is a flowchart of the information fusion method in an embodiment;

[0064] Figure 3 It is a flowchart of determining the similarity feature in an embodiment;

[0065] Figure 4 It is a flowchart of determining the syntactic similarity feature in an embodiment;

[0066] Figure 5 It is a flowchart of determining the semantic similarity feature in an embodiment;

[0067] Figure 6 It is a flowchart of determining the attribute features of each electrical device in an embodiment;

[0068] Figure 7 It is a flowchart of performing information fusion in an embodiment;

[0069] Figure 8A It is a comparison diagram of the experimental results of different models in an embodiment;

[0070] Figure 8B It is a comparison diagram of the contribution rates of different features in an embodiment;

[0071] Figure 9 It is a flowchart of the information fusion method in another embodiment;

[0072] Figure 10 It is a structural block diagram of the information fusion device in an embodiment;

[0073] Figure 11 It is an internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0074] In order to make the purpose, technical solutions and advantages of the present application clearer, the following further details the present application in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0075] The information fusion method provided by the embodiments of the present application can be applied to, for example, Figure 1 the application environment shown in the figure. Among them, the intelligent building automation system 101 is used to execute the information fusion method provided by the embodiments of the present application to integrate and manage various facilities in the target building through intelligent technologies; the building management system (BMS) 102 of the target building is used to monitor and adjust the electrical and mechanical equipment of the building, such as electrical systems, pipelines, fire alarm systems, heating, ventilation, air conditioning, power control, and lighting control; the low-voltage power distribution system 103 is a power distribution system that distributes and transmits electrical energy to each electrical equipment in the target building. Optionally, the intelligent building automation system 101 obtains the first attribute information of each electrical equipment in the low-voltage power distribution system in the target building, the second attribute information of each electrical equipment in the building management system BMS in the target building, and the equipment connection relationship between each electrical equipment; wherein, the equipment identifiers of each electrical equipment in the low-voltage power distribution system and the BMS are the same; according to the first attribute information and the second attribute information of each electrical equipment, determine the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment; according to the first attribute information and the second attribute information of each electrical equipment, and the equipment topology structure diagram of the target building, determine the attribute characteristics of each electrical equipment; according to the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment and the attribute characteristics of each electrical equipment, perform information fusion on the first attribute information and the second attribute information of each electrical equipment.

[0076] In an exemplary embodiment, as Figure 2 shown, an information fusion method is provided. Taking the intelligent building automation system 101 in Figure 1 as an example, the specific steps are as follows:

[0077] S201, obtain the first attribute information of each electrical equipment in the low-voltage power distribution system in the target building, the second attribute information of each electrical equipment in the building management system BMS in the target building, and the equipment connection relationship between each electrical equipment.

[0078] Among them, the target building can be any public building; the electrical equipment includes, but is not limited to, power generation equipment, voltage transformation equipment, power distribution equipment, etc. in the target building. The low-voltage power distribution system is a power distribution system that distributes and transmits electrical energy to each electrical equipment in the target building. The first attribute information of any electrical equipment is the attribute information stored by the electrical equipment in the low-voltage power distribution system; the second attribute information of any electrical equipment is the attribute information stored by the electrical equipment in the BMS; the equipment connection relationship between each electrical equipment represents the connection relationship between each electrical equipment.

[0079] It should be noted that in general practical applications, first, a detailed investigation is carried out on the system resources, network interfaces, and information attributes of the precision air conditioners and cooling automation control devices in the target building. By analyzing their functional modules, the precision air conditioners and cooling automation control devices are disassembled into components such as intelligent terminals, measuring intelligent electronic devices (Intelligent Electronic Device, IED) (such as intelligent electronic devices for measuring oil temperature and winding temperature), and monitoring IEDs (including intelligent electronic devices for monitoring equipment current, gas analysis, etc.), and are described in a standardized manner based on the Common Information Model (CIM). Subsequently, the device identifiers of heterogeneous electrical devices from the low-voltage power distribution and utilization system and the BMS are cleaned and aligned, and key attributes such as device identifiers, location codes, and communication protocol types are extracted to construct a unified device identifier index. That is to say, the device identifiers of each electrical device are the same in the low-voltage power distribution system and the BMS.

[0080] Optionally, the attribute information of each electrical device in the target building can be extracted from the storage system of the low-voltage power distribution system as the first attribute information of each electrical device; at the same time, the attribute information of each electrical device in the target building can be extracted from the storage system of the BMS as the second attribute information of each electrical device; in addition, the device connection relationships between each electrical device in the target building can also be obtained from the storage system of the BMS. In the embodiments of the present application, the first attribute information of each electrical device includes the first device name and the first device characteristics of the electrical device, and the second attribute information of each electrical device includes the second device name and the second device characteristics of the electrical device.

[0081] S202. Determine the similarity characteristics between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device.

[0082] Among them, the similarity characteristics between the first attribute information and the second attribute characterize the device attribute correlation of each electrical device in the low-voltage power distribution system and the BMS respectively. In the embodiments of the present application, the similarity characteristics between the first attribute information and the second attribute information of each electrical device include syntactic similarity characteristics and semantic similarity characteristics.

[0083] Optionally, a similarity characteristic extraction model can be pre-constructed, and the first attribute information and the second attribute information of each electrical device are input into the similarity characteristic extraction model so that the similarity characteristic extraction model extracts the similarity characteristics of the first attribute information and the second attribute information.

[0084] S203. Determine the attribute characteristics of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology structure diagram of the target building.

[0085] Among them, the equipment topology structure diagram of the target building characterizes the connection modes of various electrical equipment and communication media in the target building, can visually present the link relationships between the nodes in the electrical structure network of the target building, and reflect the structural relationships between the entities in the network. In the embodiments of the present application, each electrical equipment can be represented as each node in the equipment topology structure diagram, and the connection relationships between the electrical equipment are represented as the connection relationships between the nodes in the equipment topology structure diagram. The attribute characteristics of each electrical equipment characterize the relevance between the electrical equipment and the importance of each electrical equipment. In the embodiments of the present application, the attribute characteristics of each electrical equipment include but are not limited to relevance characteristics and importance characteristics. Among them, the relevance characteristics characterize to a certain extent the relevance between the electrical equipment; the importance characteristics characterize to a certain extent the importance of the electrical equipment in the target building structure.

[0086] Optionally, an attribute feature extraction model can be pre-constructed, and the first attribute information and the second attribute information of each electrical equipment are input into the attribute feature extraction model, so that the attribute feature extraction model extracts the attribute characteristics of each electrical equipment.

[0087] S204, according to the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment and the attribute characteristics of each electrical equipment, perform information fusion on the first attribute information and the second attribute information of each electrical equipment.

[0088] Optionally, a feature analysis model can be adopted, and the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment and the attribute characteristics of each electrical equipment are input into the feature analysis model, so that the feature analysis model performs feature analysis on the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment and the attribute characteristics of each electrical equipment, completes the disambiguation process of the first attribute information and the second attribute information, and completes the attribute association mapping of each electrical equipment in the low-voltage power distribution system and the BMS, so as to complete the information fusion process of the first attribute information and the second attribute information of each electrical equipment.

[0089] In the above information fusion method, after obtaining the first attribute information of each electrical device in the low-voltage power distribution system and the second attribute information in the BMS, as well as the device connection relationship between each electrical device, according to the first attribute information and the second attribute information of each electrical device, the similarity feature between the first attribute information and the second attribute information of each electrical device is determined; since the similarity feature can accurately map the difference degree between the first attribute information and the second attribute information, it can ensure the accurate disambiguation of the first attribute information and the second attribute information during the information fusion process; in addition, according to the first attribute information and the second attribute information of each electrical device, and the device topology structure diagram of the target building, the attribute feature of each electrical device is determined; since the attribute feature can characterize the association feature between each electrical device, therefore, the attribute mapping process in the subsequent information fusion process can be accurately realized, and therefore, the first attribute information and the second attribute information of each electrical device can be accurately and flexibly fused.

[0090] Optionally, in one embodiment, the first attribute information of each electrical device includes the first device name and the first device feature of the electrical device, the second attribute information of each electrical device includes the second device name and the second device feature of the electrical device, and the similarity feature includes a syntactic similarity feature and a semantic similarity feature; in this case, as Figure 3 shown, a method for determining the similarity feature between the first attribute information and the second attribute information of each electrical device is provided, which specifically includes the following steps:

[0091] S301, perform word segmentation processing on the first device name and the second device name of each electrical device respectively, to obtain the first device word segmentation and the second device word segmentation of each electrical device.

[0092] Wherein, the first device word segmentation of each electrical device is the word segmentation obtained by performing word segmentation processing on the first device name of the electrical device; the second device word segmentation of each electrical device is the word segmentation obtained by performing word segmentation processing on the second device name of the electrical device.

[0093] Optionally, a word segmentation algorithm can be used to perform word segmentation processing on the first device name and the second device name of each electrical device respectively, to obtain the first device word segmentation and the second device word segmentation of each electrical device. In the embodiments of the present application, an N-gram language model (N-gram language model) can be used to perform word segmentation processing on the first device name and the second device name of each electrical device respectively, and the N-gram language model, such as Bi-Gram or Tri-Gram. Exemplarily, assume that the first device name of a certain electrical device has m characters, and use , representing a word segment in a sentence of the first device name. In Bi-Gram, it is assumed that the probability of a word's occurrence depends only on the word that appears before it. In Tri-Gram, it is assumed that the probability of a word's occurrence depends on the two words that appear before it. In Bi-Gram and Tri-Gram The probability calculation methods for words are as follows:

[0094] (1)

[0095] (2)

[0096] Among them, identifies the th character, ; and respectively represent the probabilities that the word segments are words when using the Bi-Gram and Tri-Gram methods ; represents that when the th character is composed, the probability that the th character is also composed. Exemplarily, taking the first device name as "one-layer water heater" as an example, the results after using the Bi-Gram method = {one layer, water heater}, = {one, layer water heater}, = {one layer hot, water heater}. Among them, , , , among the three probabilities, "one layer" is more common in the corpus, so has a relatively high probability, and "one-layer water heater" will be segmented into "one layer" and "water heater".

[0097] S302. Perform word segmentation processing on the first device feature and the second device feature of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device.

[0098] Optionally, the same word segmentation method in S301 above can also be used to perform word segmentation processing on the first device feature and the second device feature of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device.

[0099] S303. Determine the syntactic similarity feature between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device.

[0100] Among them, the syntactic similarity feature between the first attribute information and the second attribute information of each electrical device characterizes the difference or similarity between the device naming rules of the electrical device in the low-voltage power distribution system and the BMS.

[0101] Optionally, a syntactic similarity feature extraction model based on word segmentation can be pre-constructed, and the first device word segmentation and the second device word segmentation of each electrical device are input into the syntactic similarity feature extraction model, so that the syntactic similarity feature extraction model extracts the syntactic similarity feature between the first attribute information and the second attribute information of each electrical device.

[0102] S304. Determine the semantic similarity feature between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device.

[0103] Among them, the semantic similarity feature between the first attribute information and the second attribute information of each electrical device characterizes the similarity or difference in the device attribute descriptions of the electrical device in the low-voltage power distribution system and the BMS.

[0104] Optionally, a semantic similarity feature extraction model based on word segmentation can be pre-constructed, and the first feature word segmentation and the second feature word segmentation of each electrical device are input into the semantic similarity feature extraction model, so that the semantic similarity feature extraction model extracts the semantic similarity feature between the first attribute information and the second attribute information of each electrical device.

[0105] In this embodiment, by performing word segmentation on the first device name and the second device name, as well as the first device feature and the second device feature, refined management of the first attribute information and the second attribute information is achieved; furthermore, the accuracy of the determined semantic similarity feature and syntactic similarity feature is ensured.

[0106] Optionally, in one embodiment, as Figure 4 shown, a method for determining the syntactic similarity feature between the first attribute information and the second attribute information of each electrical device is provided to refine the above S303, and specifically includes the following steps:

[0107] S401. For each electrical device, use the string length of the first device word segmentation of the electrical device as the first length.

[0108] It should be noted that the process of determining the syntactic similarity feature between the first attribute information and the second attribute information of each electrical device is to identify the device name information of the same electrical device from different systems or models, and avoid repeated calculation or data inconsistency caused by naming format or symbol differences.

[0109] Optionally, for each electrical device, the string length of the first device word segmentation of the electrical device can be calculated and analyzed through a character recognition model, and the recognized string length of the first device word segmentation of the electrical device is used as the first length.

[0110] S402, use the string length of the second device word segmentation of the electrical device as the second length.

[0111] Optionally, the string length of the second device word segmentation of the electrical device can also be calculated and analyzed through a character recognition model, and the recognized string length of the second device word segmentation of the electrical device is used as the second length.

[0112] S403, determine the edit distance between the first device name and the second device name of the electrical device according to the first length and the second length.

[0113] Among them, the edit distance between the first device name and the second device name of the electrical device is used to compare character sequences representing device information. The smaller the edit distance, the higher the corresponding syntactic similarity, indicating a stronger similarity between the two strings (the first device name and the second device name).

[0114] Optionally, the process of determining the edit distance between the first device name and the second device name of the electrical device according to the first length and the second length can be represented by the following formula (3):

[0115] (3)

[0116] Among them, is the first length; is the second length; is the minimum number of basic character editing operations (including deletion, addition, reordering, and replacement) required to transform the first device word segmentation into the second device word segmentation; is the maximum length of the first length and the second length.

[0117] S404, determine the syntactic similarity feature between the first attribute information and the second attribute information of the electrical device according to the edit distance.

[0118] Optionally, the normalized edit distance can be converted into the syntactic similarity between words, that is, the edit distance is normalized and converted into the syntactic similarity feature between the first attribute information and the second attribute information of the electrical device.

[0119] In this embodiment, the edit distance is introduced. By calculating the degree of proximity between the two strings of the first device name and the second device name, the difference between the first device name and the second device name is indirectly measured, thereby ensuring the accuracy of the syntactic similarity feature between the determined first attribute information and the second attribute information.

[0120] Optionally, in one embodiment, as Figure 5 shown, a method for determining the semantic similarity feature between the first attribute information and the second attribute information of each electrical device is provided to refine the above S304, which specifically includes the following steps:

[0121] S501. For each electrical device, using the bag-of-words model, the first feature of the electrical device is tokenized and represented as a vector to obtain the first attribute vector of the electrical device.

[0122] Optionally, in the process of determining the semantic similarity feature between the first attribute information and the second attribute information of each electrical device, disambiguation of the static features and association relationships of the electrical device is achieved, including metadata such as device type, function description, subordinate subsystem, communication interface, etc. By extracting device attribute features, the system can accurately understand the association between different electrical devices, thereby optimizing the accurate mapping of device identifiers.

[0123] Optionally, the first feature tokens of the electrical device can be input into the bag-of-words model, so that the bag-of-words model represents the first feature tokens of the electrical device as a vector to obtain the first attribute vector of the electrical device. Specifically, according to the first feature tokens, such as the features of attributes such as device type, function keywords, communication protocol, etc., a list containing the first feature tokens is constructed, and a vector with a dimension of n is generated. Each element in the vector corresponds to the occurrence frequency of the feature in the list.

[0124] S502. Using the bag-of-words model, the second feature tokens of the electrical device are tokenized and represented as a vector to obtain the second attribute vector of the electrical device.

[0125] Similarly, the second feature tokens of the electrical device can be input into the bag-of-words model, so that the bag-of-words model represents the second feature tokens of the electrical device as a vector to obtain the second attribute vector of the electrical device. Specifically, according to the second feature tokens, such as the features of attributes such as device type, function keywords, communication protocol, etc., a list containing the second feature tokens is constructed, and a vector with a dimension of n is generated. Each element in the vector corresponds to the occurrence frequency of the feature in the list.

[0126] S503. According to the cosine similarity between the first attribute vector and the second attribute vector of the electrical device, the semantic similarity feature between the first attribute information and the second attribute information of the electrical device is determined.

[0127] Optionally, the cosine similarity between two vectors can be calculated to quantify the semantic similarity feature between the first attribute information and the second attribute information of the electrical device. The formula is as follows:

[0128] (4)

[0129] Wherein, represents the first attribute vector; represents the second attribute vector; is the inner product of vectors; is the vector length of the first attribute vector; is the vector length of the second attribute vector.

[0130] In this embodiment, by introducing the bag-of-words model, the first feature word segmentation and the second feature word segmentation of the electrical device are represented as vectors, and by calculating the cosine similarity between the first attribute vector and the second attribute vector, the semantic similarity feature between the first attribute vector and the second attribute vector is indirectly measured, ensuring the accuracy of the determined semantic similarity feature.

[0131] Optionally, in one embodiment, the first attribute information of each electrical device includes the first device name and the first device features of the electrical device, and the second attribute information of each electrical device includes the second device name and the second device features of the electrical device. The attribute features include the relevance feature and the importance feature; in this case, as Figure 6 shown, a method for determining the attribute features of each electrical device is provided, which specifically includes the following steps:

[0132] S601, perform word segmentation processing on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device.

[0133] Optionally, the word segmentation method in S301 in the above embodiment can be adopted to perform word segmentation processing on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device.

[0134] S602, perform word segmentation processing on the first device features and the second device features of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device.

[0135] Optionally, the word segmentation method in S301 in the above embodiment can also be adopted to perform word segmentation processing on the first device features and the second device features of each electrical device respectively to obtain the first feature word segmentation and the second device word segmentation of each electrical device.

[0136] S603. Determine the location information of each electrical device in the target building and the device connection relationship between each electrical device according to the device topology structure diagram of the target building.

[0137] S604. Determine the correlation characteristics of each electrical device according to the device connection relationship between each electrical device and the first feature participles and second feature participles of each electrical device.

[0138] It can be understood that in the building automation scenario, if a certain candidate entity is unique in the knowledge base (such as a uniquely matched device ID), its correlation with the remaining candidate entities can be used as a key basis for disambiguation. Therefore, according to the device connection relationship between each electrical device and the first feature participles and second feature participles of each electrical device, a device association knowledge base of the target building can be constructed; according to the device association knowledge base, the correlation characteristics of each electrical device can be determined. Specifically, a device association knowledge base (including the "device identifier - affiliated system - location" triple) of each electrical device in the target building can be constructed, and the association strength of the first feature participles and second feature participles of each electrical device can be quantified. The specific process is as follows:

[0139] First, construct a device association knowledge base, define the topology relationship type and its basic weight to form a weight table. Exemplarily, as shown in Table 1.

[0140] Table 1 Weight Table

[0141]

[0142] Secondly, retrieve the association path between the first feature participles and second feature participles of each electrical device, record its association type and path length; then for each matched triple, calculate the influence feature according to the following formula:

[0143] (5)

[0144] Among them, is the path attenuation coefficient (default value is 0.5); is the weight; is the association path length. When directly associated ( =1), ; when indirectly associated ( ≥2), decays exponentially.

[0145] Finally, to avoid the feature value from being too large, normalize the total influence feature value:

[0146] (6)

[0147] The process of determining the correlation characteristics of each electrical device can solve semantic conflicts during cross-electrical-device mapping, ensure the consistency of device identification in the building automation system, and support the coordinated control of subsystems such as air conditioning, lighting, and security.

[0148] S605. Using a web page ranking algorithm, based on the location information of each electrical device in the target building, as well as the first feature word segmentation and the second feature word segmentation of each electrical device, determine the importance characteristics of each electrical device.

[0149] Optionally, in the intelligent building automation system, the importance characteristics of electrical devices are used to measure the criticality of electrical devices in the knowledge graph, ensuring that core devices obtain higher priority during entity disambiguation. For example, highly connected devices such as central controllers and main distribution circuit breakers play a decisive role in the stable operation of the building automation system due to their hub positions in the system topology. By calculating the importance characteristics of devices, the system can prioritize the identification mapping of key electrical devices, avoiding control failures or communication interruptions caused by disambiguation errors.

[0150] Optionally, in the embodiments of the present application, the importance of electrical devices can be calculated based on an improved web page ranking algorithm, such as the PageRank algorithm. Specifically, the transition probability in the PageRank calculation can be dynamically adjusted according to the device type and function of the electrical device. The formula is as follows:

[0151] (7)

[0152] Among them, represents the PageRank value of the electrical device , that is, the importance of the electrical device is the weight of the electrical device to (determined by the device type and status), is the set of all out-links to , is the number of out-links of is the total number of electrical devices in the device knowledge graph, is the damping coefficient, generally taken as 0.85.

[0153] In this embodiment, through the location information of each electrical device and the device connection relationship between each electrical device, the relationship between each electrical device and the important position of each electrical device in the target building can be more accurately and intuitively characterized, thereby ensuring the accuracy of the determined correlation characteristics and importance characteristics of each electrical device.

[0154] Optionally, in one embodiment, such as Figure 7As shown, a method for information fusion of the first attribute information and the second attribute information of each electrical device is provided, which specifically includes the following steps:

[0155] S701. For any electrical device, determine the first weight of the similarity feature between the first attribute information and the second attribute information of the electrical device, and the second weight of the attribute feature of the electrical device.

[0156] Optionally, for any electrical device, the first weight of the similarity feature between the first attribute information and the second attribute information of the electrical device, and the second weight of the attribute feature of the electrical device can be determined according to the similarity feature between the first attribute information and the second attribute information of the electrical device and the importance of the attribute feature of the electrical device. It should be noted that the sum of the first weight and the second weight is 1.

[0157] S702. Use the product of the first weight of the electrical device and the similarity feature as the first intermediate parameter.

[0158] In the embodiment of the present application, the similarity feature of the electrical device includes a semantic similarity feature and a syntactic similarity feature. In this case, after normalizing the semantic similarity feature and the syntactic similarity feature, a similarity feature is generated through linear combination , which can effectively solve the mapping conflict caused by the naming rule differences of electrical devices and improve the accuracy and consistency of cross-system device attribute association. Specifically, as shown in the following formula (8):

[0159] (8)

[0160] Wherein, is the coefficient of semantic similarity; is the coefficient of syntactic similarity; is the semantic similarity; is the syntactic similarity.

[0161] Furthermore, using the product of the first weight of the electrical device and the similarity feature as the first intermediate parameter is specifically:

[0162] (9)

[0163] Wherein, is the first intermediate parameter; is the first weight.

[0164] S703. Use the product of the second weight of the electrical device and the attribute feature as the second intermediate parameter.

[0165] Optionally, using the product of the second weight of the electrical device and the attribute feature as the second intermediate parameter is specifically:

[0166] (10)

[0167] Wherein, is the second intermediate parameter; is the second weight.

[0168] In the embodiments of the present application, the attribute characteristics of each electrical device include relevance characteristics and importance characteristics. In this case, the relevance characteristics and importance characteristics can be weighted and summed to obtain the attribute characteristics, specifically:

[0169] (11)

[0170] (12)

[0171] Wherein, is the importance characteristic; is the relevance characteristic; is the coefficient of the relevance characteristic; is the coefficient of the importance characteristic.

[0172] S704, Take the sum between the first intermediate parameter and the second intermediate parameter as the comprehensive characteristic of the electrical device.

[0173] Optionally, take the sum between the first intermediate parameter and the second intermediate parameter as the comprehensive characteristic of the electrical device, specifically:

[0174] (13)

[0175] Wherein, is the comprehensive characteristic.

[0176] S705, According to the comprehensive characteristic of the electrical device, perform information fusion on the first attribute information and the second attribute information of each electrical device.

[0177] Optionally, if the comprehensive characteristic is higher, disambiguation is preferentially performed to ensure that the system preferentially maps key devices, realizing information fusion of the first attribute information and the second attribute information of each electrical device, thereby ensuring the control stability of the building automation system.

[0178] In this embodiment, by introducing the first weight and the second weight, the accuracy of the determined comprehensive characteristic of the electrical device is ensured, and further the accuracy of the information fusion process is ensured.

[0179] Optionally, to verify the accuracy of the information fusion method provided in this application, the information fusion method can be applied to the attribute association mapping in the low-voltage power distribution system and the BMS, and verified through experiments. In one embodiment, the experiment is tested based on the public dataset CCKS2019-CKBQA and the customized cross-system device dataset. The customized dataset integrates the heterogeneous device identification data of the low-voltage power distribution system and the BMS, covering key attributes such as device identification, location coding, communication protocol type, and the subsystem to which it belongs, as shown in Table 2. The dataset size is 3000 training sets, 1000 validation sets, and 1000 test sets, covering device types such as air-conditioning units, power distribution terminals, and lighting controllers, and additionally annotating the cross-system entity references to evaluate the accuracy and consistency of attribute associations.

[0180] Table 2 Customized Dataset

[0181]

[0182] The experiment is carried out using PyCharm in the Windows10 environment, and the Pytorch version "Bert-Base-Chinese" of the Bert pre-trained model is used for word vector calculation. The encoder of the Bert model has 12 layers, the word vector dimension is 768 dimensions, and the maximum sentence length is set to 256. The experiment simulates the device identification mapping scenario between the low-voltage power distribution system and the BMS, and focuses on verifying the optimization effect of the multi-feature disambiguation technology on the cross-system attribute association. Suppose there are n entity references E N ={e1,e2,…,e n}, the predicted result E p ={e1,e2,…,e n}, and the target entity corresponding to the entity reference is E t ={e1,e2,…,e t}, then the accuracy rate P, recall rate R, and F value of entity disambiguation are respectively: . For the scenario of low-carbon buildings, the entity references in the experimental data have been secondarily annotated, so |E p |=|E t |, P = R = F, and the accuracy rate P is used as the main evaluation index of the experimental results to further evaluate the carbon emission optimization effect.

[0183] The experiment selects the following four features to verify their contributions to the attribute association mapping:

[0184] (1) Syntactic similarity feature ( ): Calculated based on the edit distance of device names to solve the identification conflicts caused by the naming rule differences between the low-voltage power distribution and utilization system and the BMS.

[0185] (2) Semantic similarity feature ( ): Analyze device attributes (such as type, function, communication protocol) through the bag-of-words model to quantify the semantic relevance of cross-system devices..

[0186] (3) Relevance feature ( ): Based on the topological relationships in the knowledge base (such as "subordinate to the power distribution system", "belonging to the same air conditioning unit"), enhance the mapping reliability of cross-system devices.

[0187] (4) Importance feature ( ): Identify core devices (such as the central controller, main power distribution breaker) by improving the PageRank algorithm to ensure the priority of key devices in cross-system mapping and enhance the overall control stability.

[0188] Entity disambiguation models with different numbers of features were constructed based on the above features, namely:

[0189] Model 1: Use semantic similarity feature ( ) and importance feature ( ).

[0190] Model 2: Use semantic similarity ( ), importance feature ( ) and syntactic similarity feature ( ).

[0191] Model 3: Use all features ( , , , ).

[0192] Model 4: CCKS-CKBQA evaluation benchmark model.

[0193] The experimental results are as Figure 8A shown. The accuracy of Model 3 is the highest (91.13%), which is 0.6% higher than that of Model 4. Especially in the cross-system device mapping scenario, Model 3 significantly reduces mapping conflicts by integrating topological relationships and device importance. Figure 8B It shows that the device relevance feature and topological importance feature contribute the most to the results, verifying the effectiveness of the collaborative disambiguation strategy.

[0194] Table 3 compares the influence of the word segmentation method. After using the N-Gram word segmentation method, the accuracy is improved by 0.08%, indicating that normalized word segmentation is crucial for the cleaning and alignment of heterogeneous identifiers. Table 4 shows that the cross-system mapping error rate of Model 3 is reduced by 12% compared with the benchmark model, and the daily average data conflict events are reduced by 9%, proving that this method can significantly improve the collaborative operation efficiency of the low-voltage power distribution and utilization system and the BMS.

[0195] Experiments have proved that through multi-feature fusion, the present invention effectively solves the problem of attribute association mapping between low-voltage power distribution and utilization systems and BMS models, providing reliable technical support for cross-system integration and collaborative control of intelligent buildings.

[0196] Table 3 Influence of Word Segmentation Methods

[0197] Candidate entity preprocessing method Entity disambiguation accuracy rate (%) Tokenization method using N-Gram model 91.13 Tokenization method without using N-Gram model 91.05

[0198] Table 4 Mapping Error Rates and Average Daily Data Conflict Events of Different Models

[0199] Method Mapping error rate (%) Average daily data conflict events BMS benchmark 15.2 127 Model 1 13.1 112 Model 2 10.8 95 Model 3 8.5 72 Model 4 9.1 81

[0200] Figure 9 FIG. is a schematic flow chart of an information fusion method in another embodiment. Based on the above embodiments, this embodiment provides an alternative example of an information fusion method. Combining Figure 9 , the specific implementation process is as follows:

[0201] S901, obtain the first attribute information of each electrical device in the low-voltage power distribution system in the target building, the second attribute information of each electrical device in the building management system BMS, and the device connection relationships between the electrical devices.

[0202] Among them, the device identifiers of each electrical device in the low-voltage power distribution system and BMS are the same.

[0203] S902, perform word segmentation processing on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device.

[0204] S903, perform word segmentation processing on the first device feature and the second device feature of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device.

[0205] S904, determine the syntactic similarity feature between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device.

[0206] Optionally, for each electrical device, take the string length of the first device word segmentation of the electrical device as the first length; take the string length of the second device word segmentation of the electrical device as the second length; determine the edit distance between the first device name and the second device name of the electrical device according to the first length and the second length; determine the syntactic similarity feature between the first attribute information and the second attribute information of the electrical device according to the edit distance.

[0207] S905. Determine the semantic similarity feature between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device.

[0208] Optionally, for each electrical device, use the bag-of-words model to represent the first feature word segmentation of the electrical device as a vector to obtain the first attribute vector of the electrical device; use the bag-of-words model to represent the second feature word segmentation of the electrical device as a vector to obtain the second attribute vector of the electrical device; determine the semantic similarity feature between the first attribute information and the second attribute information of the electrical device according to the cosine similarity between the first attribute vector and the second attribute vector of the electrical device.

[0209] S906. Determine the location information of each electrical device in the target building and the device connection relationship between each electrical device according to the device topology structure diagram of the target building.

[0210] S907. Determine the relevance feature of each electrical device according to the device connection relationship between each electrical device, and the first feature word segmentation and the second feature word segmentation of each electrical device.

[0211] Optionally, construct a device association knowledge base of the target building according to the device connection relationship between each electrical device, and the first feature word segmentation and the second feature word segmentation of each electrical device; determine the relevance feature of each electrical device according to the device association knowledge base.

[0212] S908. Use the page rank algorithm to determine the importance feature of each electrical device according to the location information of each electrical device in the target building, and the first feature word segmentation and the second feature word segmentation of each electrical device.

[0213] S909. Perform information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity feature between the first attribute information and the second attribute information of each electrical device and the attribute feature of each electrical device.

[0214] Optionally, for any electrical device, determine the first weight of the similarity feature between the first attribute information and the second attribute information of the electrical device, and the second weight of the attribute feature of the electrical device; take the product of the first weight and the similarity feature of the electrical device as the first intermediate parameter; take the product of the second weight and the attribute feature of the electrical device as the second intermediate parameter; take the sum between the first intermediate parameter and the second intermediate parameter as the comprehensive feature of the electrical device; perform information fusion on the first attribute information and the second attribute information of each electrical device according to the comprehensive feature of the electrical device.

[0215] The specific processes of the above S901 - S909 can refer to the descriptions of the above method embodiments, and their implementation principles and technical effects are similar, so they will not be elaborated here.

[0216] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the indications of the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments, and the execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of the steps or stages in other steps or other steps.

[0217] Based on the same inventive concept, an embodiment of the present application also provides an information fusion device for implementing the above-mentioned information fusion method. The solution provided by this device for solving problems is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the information fusion device provided below can refer to the limitations on the information fusion method in the above text, and will not be repeated here.

[0218] In an exemplary embodiment, as Figure 10 shown, an information fusion device 1000 is provided, including: an information acquisition module 1010, a first determination module 1020, a second determination module 1030, and an information fusion module 1040, where:

[0219] The information acquisition module 1010 is configured to acquire the first attribute information of each electrical device in the low-voltage power distribution system of the target building, the second attribute information of each electrical device in the building management system BMS, and the device connection relationship between each electrical device; wherein, the device identifiers of each electrical device in the low-voltage power distribution system and the BMS are the same.

[0220] The first determination module 1020 is configured to determine the similarity feature between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device.

[0221] The second determination module 1030 is configured to determine the attribute feature of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology structure diagram of the target building.

[0222] The information fusion module 1040 is configured to perform information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity feature between the first attribute information and the second attribute information of each electrical device and the attribute feature of each electrical device.

[0223] In one embodiment, the first attribute information of each electrical device includes the first device name and the first device features of the electrical device, the second attribute information of each electrical device includes the second device name and the second device features of the electrical device, and the similarity features include syntactic similarity features and semantic similarity features; the first determination module 1020 includes:

[0224] A first word segmentation unit, configured to perform word segmentation processing on the first device name and the second device name of each electrical device respectively, to obtain the first device word segmentation and the second device word segmentation of each electrical device.

[0225] A second word segmentation unit, configured to perform word segmentation processing on the first device features and the second device features of each electrical device respectively, to obtain the first feature word segmentation and the second feature word segmentation of each electrical device.

[0226] A first determination unit, configured to determine the syntactic similarity features between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device.

[0227] A second determination unit, configured to determine the semantic similarity features between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device.

[0228] In one embodiment, the first determination unit is specifically configured to:

[0229] For each electrical device, take the string length of the first device word segmentation of the electrical device as the first length; take the string length of the second device word segmentation of the electrical device as the second length; determine the edit distance between the first device name and the second device name of the electrical device according to the first length and the second length; determine the syntactic similarity features between the first attribute information and the second attribute information of the electrical device according to the edit distance.

[0230] In one embodiment, the second determination unit is specifically configured to:

[0231] For each electrical device, adopt the bag-of-words model, represent the first feature word segmentation of the electrical device as a vector to obtain the first attribute vector of the electrical device; adopt the bag-of-words model, represent the second feature word segmentation of the electrical device as a vector to obtain the second attribute vector of the electrical device; determine the semantic similarity features between the first attribute information and the second attribute information of the electrical device according to the cosine similarity between the first attribute vector and the second attribute vector of the electrical device.

[0232] In one embodiment, the first attribute information of each electrical device includes the first device name and the first device characteristics of the electrical device, the second attribute information of each electrical device includes the second device name and the second device characteristics of the electrical device, and the attribute characteristics include the relevance characteristics and the importance characteristics; the second determination module 1030 includes:

[0233] A third word segmentation unit, configured to perform word segmentation processing on the first device name and the second device name of each electrical device respectively, to obtain the first device word segmentation and the second device word segmentation of each electrical device.

[0234] A fourth word segmentation unit, configured to perform word segmentation processing on the first device characteristics and the second device characteristics of each electrical device respectively, to obtain the first characteristic word segmentation and the second characteristic word segmentation of each electrical device.

[0235] A relationship determination unit, configured to determine the location information of each electrical device in the target building and the device connection relationship between each electrical device according to the device topology structure diagram of the target building.

[0236] A third determination unit, configured to determine the relevance characteristics of each electrical device according to the device connection relationship between each electrical device and the first characteristic word segmentation and the second characteristic word segmentation of each electrical device.

[0237] A fourth determination unit, configured to use a page ranking algorithm to determine the importance characteristics of each electrical device according to the location information of each electrical device in the target building and the first characteristic word segmentation and the second characteristic word segmentation of each electrical device.

[0238] In one embodiment, the third determination unit is specifically configured to:

[0239] Construct a device association knowledge base of the target building according to the device connection relationship between each electrical device and the first characteristic word segmentation and the second characteristic word segmentation of each electrical device; determine the relevance characteristics of each electrical device according to the device association knowledge base.

[0240] In one embodiment, the information fusion module 1040 is specifically configured to:

[0241] For any electrical device, determine the first weight of the similarity characteristics between the first attribute information and the second attribute information of the electrical device, and the second weight of the attribute characteristics of the electrical device; use the product of the first weight and the similarity characteristics of the electrical device as the first intermediate parameter; use the product of the second weight and the attribute characteristics of the electrical device as the second intermediate parameter; use the sum of the first intermediate parameter and the second intermediate parameter as the comprehensive characteristic of the electrical device; perform information fusion on the first attribute information and the second attribute information of each electrical device according to the comprehensive characteristic of the electrical device.

[0242] Each module in the above information fusion device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above modules can be embedded in the processor of a computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0243] In an exemplary embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 11 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements an information fusion method.

[0244] Those skilled in the art can understand that Figure 11 the structure shown in

[0245] is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0246] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the information fusion method provided in the above embodiment.

[0247] In an embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, it implements the information fusion method provided in the above embodiment.

[0248] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data that have been authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0249] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the various embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the various embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the various embodiments provided in this application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0250] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the various technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0251] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. An information fusion method, characterized in that The method includes: Obtaining first attribute information of each electrical device in the low-voltage power distribution system of a target building, second attribute information of each electrical device in a building management system (BMS), and the device connection relationships between the electrical devices; wherein, the device identifiers of the electrical devices in the low-voltage power distribution system and the BMS are the same; Determining the similarity features between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device; Determining the attribute features of each electrical device according to the first attribute information and the second attribute information of each electrical device and the device topology diagram of the target building; Performing information fusion on the first attribute information and the second attribute information of each electrical device according to the similarity features between the first attribute information and the second attribute information of each electrical device and the attribute features of each electrical device.

2. The method according to claim 1, characterized in that, The first attribute information of each electrical device includes the first device name and the first device features of the electrical device, the second attribute information of each electrical device includes the second device name and the second device features of the electrical device, and the similarity features include syntactic similarity features and semantic similarity features; The determining the similarity features between the first attribute information and the second attribute information of each electrical device according to the first attribute information and the second attribute information of each electrical device includes: Performing word segmentation processing on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device; Performing word segmentation processing on the first device features and the second device features of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device; Determining the syntactic similarity features between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device; Determining the semantic similarity features between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device.

3. The method according to claim 2, wherein The determining the syntactic similarity features between the first attribute information and the second attribute information of each electrical device according to the first device word segmentation and the second device word segmentation of each electrical device includes: For each electrical device, taking the string length of the first device word segmentation of the electrical device as the first length; Taking the string length of the second device word segmentation of the electrical device as the second length; Determining the edit distance between the first device name and the second device name of the electrical device according to the first length and the second length; Determining the syntactic similarity features between the first attribute information and the second attribute information of the electrical device according to the edit distance.

4. The method according to claim 2, wherein The determining the semantic similarity features between the first attribute information and the second attribute information of each electrical device according to the first feature word segmentation and the second feature word segmentation of each electrical device includes: For each electrical device, using the bag-of-words model, segment the first feature of the electrical device into words and represent it as a vector to obtain the first attribute vector of the electrical device; Using the bag-of-words model, segment the second feature of the electrical device into words and represent it as a vector to obtain the second attribute vector of the electrical device; According to the cosine similarity between the first attribute vector and the second attribute vector of the electrical device, determine the semantic similarity feature between the first attribute information and the second attribute information of the electrical device.

5. The method according to claim 1, characterized in that, The first attribute information of each electrical device includes the first device name and the first device feature of the electrical device, the second attribute information of each electrical device includes the second device name and the second device feature of the electrical device, and the attribute feature includes the relevance feature and the importance feature; The determining of the attribute features of each electrical device according to the first attribute information and the second attribute information of each electrical device, and the device topology structure diagram of the target building includes: Perform word segmentation processing on the first device name and the second device name of each electrical device respectively to obtain the first device word segmentation and the second device word segmentation of each electrical device; Perform word segmentation processing on the first device feature and the second device feature of each electrical device respectively to obtain the first feature word segmentation and the second feature word segmentation of each electrical device; According to the device topology structure diagram of the target building, determine the location information of each electrical device in the target building and the device connection relationship between each electrical device; According to the device connection relationship between each electrical device, and the first feature word segmentation and the second feature word segmentation of each electrical device, determine the relevance feature of each electrical device; Adopt the PageRank algorithm, and according to the location information of each electrical device in the target building, and the first feature word segmentation and the second feature word segmentation of each electrical device, determine the importance feature of each electrical device.

6. The method according to claim 5, characterized in that, The determining of the relevance feature of each electrical device according to the device connection relationship between each electrical device, and the first feature word segmentation and the second feature word segmentation of each electrical device includes: According to the device connection relationship between each electrical device, and the first feature word segmentation and the second feature word segmentation of each electrical device, construct the device association knowledge base of the target building; According to the device association knowledge base, determine the relevance feature of each electrical device.

7. The method according to claim 1, characterized in that The information fusion of the first attribute information and the second attribute information of each electrical device according to the similarity feature between the first attribute information and the second attribute information of each electrical device and the attribute feature of each electrical device includes: For any electrical device, determine the first weight of the similarity feature between the first attribute information and the second attribute information of the electrical device, and the second weight of the attribute feature of the electrical device; Take the product of the first weight of the electrical device and the similarity feature as the first intermediate parameter; Take the product of the second weight of the electrical device and the attribute feature as the second intermediate parameter; Take the sum of the first intermediate parameter and the second intermediate parameter as the comprehensive feature of the electrical device; According to the comprehensive characteristics of the electrical equipment, information fusion is performed on the first attribute information and the second attribute information of each electrical equipment.

8. An information fusion device, characterized in that, The device includes: An information acquisition module, configured to acquire the first attribute information of each electrical equipment in the low-voltage power distribution system in a target building, the second attribute information of each electrical equipment in the building management system BMS, and the device connection relationship between the electrical equipment; wherein, the device identifiers of each electrical equipment in the low-voltage power distribution system and the BMS are the same; A first determination module, configured to determine the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment according to the first attribute information and the second attribute information of each electrical equipment; A second determination module, configured to determine the attribute characteristics of each electrical equipment according to the first attribute information and the second attribute information of each electrical equipment and the device topology structure diagram of the target building; An information fusion module, configured to perform information fusion on the first attribute information and the second attribute information of each electrical equipment according to the similarity characteristics between the first attribute information and the second attribute information of each electrical equipment and the attribute characteristics of each electrical equipment.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images