An AI intelligent data processing method for the fire protection field

Abstract

The present invention belongs to the field of fire protection data processing, and in particular relates to an AI intelligent data processing method for the fire protection field. This method first collects multi-domain and multi-source data in the fire protection field to construct a multi-associated knowledge graph of fire protection; secondly, uses a configured multi-input intention inference model to parse the semantic information and scenario background input by the user and determine the main - collaborative knowledge sub-graph; thirdly, constructs a multi-graph joint inference model, fuses the main - collaborative knowledge sub-graph with the user input semantic information to generate a set of user fire protection demand plans, and continuously optimizes the model performance by constructing an optimal inference search path and an inference approximation loss function; finally, according to the real-time user input and scenario feature space, uses the trained multi-graph joint inference model to quickly generate user-customized fire protection demand plans, and feeds this information back into the multi-associated knowledge graph of fire protection to achieve continuous update and accumulation of knowledge, providing strong support for intelligent decision-making in the fire protection field.

Description

Technical Field

[0001] The present invention belongs to the field of fire protection data processing, and particularly relates to an AI intelligent data processing method for the fire protection field. Background Art

[0002] In the field of fire protection, traditional fire protection management methods mainly rely on manual inspections and basic fire protection facilities, such as smoke detectors and sprinkler systems. Although these methods can cope with fire accidents to a certain extent, they have obvious deficiencies in risk prediction, rapid response, and optimal allocation of resources; for example, traditional methods often rely on simple resampling or a small amount of data augmentation techniques for fire risk data augmentation, which easily leads to overfitting and insufficient generalization ability; existing technologies may not fully utilize the internal structure of fire risk data, resulting in a time-consuming and inefficient feature extraction and dimensionality reduction process; when evaluating high-risk fire events, existing technologies may be difficult to achieve high accuracy and sensitivity due to limitations in model design and loss functions.

[0003] For example, the Chinese patent application with the publication number CN118152939A discloses an intelligent fire protection data processing method and device. The method includes: obtaining target fire protection data to be processed and corresponding data identification parameters; inputting the data identification parameters into a pre-trained neural network model to obtain the predicted data level and predicted data type of the output; determining the data processing rules and data forwarding rules corresponding to the target fire protection data according to the predicted data level and predicted data type, and a preset type-rule correspondence; and performing data processing operations and forwarding operations on the target fire protection data according to the data processing rules and data forwarding rules.

[0004] For example, the Chinese patent application with the publication number CN114764743A discloses a fire protection data processing method and a smart fire protection platform; based on fire protection management objects, fire protection data is divided into N preset categories; the categories include: personnel, substances, fire sources, building facilities, fire protection systems, and fire protection facilities; collecting the fire protection data of enterprises, institutions, and their buildings and inputting them into a network database, and initializing the fire protection data; establishing corresponding data update rules based on the categories, and enterprises and institutions dynamically update the fire protection data according to the data update rules.

[0005] The existing technologies have the following problems: The existing technologies cannot provide professional and accurate fire assessment plans and overall professional fire protection strategy plans for different scenarios in the fire protection field, and cannot quickly find corresponding answers to core problems, resulting in search delays and inability to give accurate answers. In addition, the fire protection field is a large field including multiple different small fields, and each small field involves huge amounts of data. How to coordinate and manage these data is also one of the difficulties of the existing technologies. Therefore, the present invention provides an AI intelligent data processing method for the fire protection field. Summary of the Invention

[0006] In view of the deficiencies of the existing technologies, the present invention proposes an AI intelligent data processing method for the fire protection field. This method first collects multi-domain and multi-source data in the fire protection field to construct a multi-correlated knowledge graph of fire protection; secondly, uses a configured multi-input intention inference model to analyze the semantic information and scene feature space input by the user, and determine the main - collaborative knowledge sub-graph; thirdly, constructs a multi-graph joint inference model, fuses the main - collaborative knowledge sub-graph with the user input semantic information to generate a set of user fire protection demand plans, and continuously optimizes the model performance by constructing an optimal inference search path and an inference approximation loss function; finally, according to the real-time user input and scene feature space, uses the trained multi-graph joint inference model to quickly generate user-customized fire protection demand plans, and feeds this information back into the multi-correlated knowledge graph of fire protection to achieve continuous update and accumulation of knowledge, providing strong support for intelligent decision-making in the fire protection field.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] An AI intelligent data processing method for the fire protection field, comprising:

[0009] S1. Obtain building fire protection layout data, fire protection law data, fire and corresponding treatment information data, and fire protection simulation and management data in different fields, and construct a multi-correlated knowledge graph of fire protection through natural language algorithms and graph databases;

[0010] S2. Through a configured multi-input intention inference model, obtain the semantic information input by the user and the corresponding scene feature space, and determine the main knowledge sub-graph and the collaborative knowledge sub-graph in the multi-correlated knowledge graph of fire protection through the semantic information input by the user and the corresponding scene feature space and the multi-correlated knowledge graph of fire protection, and perform main - collaborative knowledge sub-graph switching according to the scene feature space through a configured main - collaboration switching node;

[0011] S3. Construct a multi-graph joint reasoning model. Input the information of the main knowledge sub-graph and the collaborative knowledge sub-graph, the user input semantic information, and the scenario feature space into the multi-graph joint reasoning model to obtain a set of user fire demand plans. At the same time, use the set of user fire demand plans to construct an optimal reasoning search path and a reasoning approximation loss function to train the multi-graph joint reasoning model, and obtain a trained multi-graph joint reasoning model;

[0012] S4. According to the user input semantic information and the corresponding scenario feature space obtained in real time, through the trained multi-graph joint reasoning model, obtain the user fire demand plan, and at the same time save the real-time user input semantic information, the scenario feature space, and the corresponding user fire demand plan into the fire multi-associated knowledge graph.

[0013] Specifically, the fire multi-associated knowledge graph includes a first knowledge sub-graph, a second knowledge sub-graph, a third knowledge sub-graph, and a fourth knowledge sub-graph; the specific steps for constructing the fire multi-associated knowledge graph are as follows:

[0014] S101. Set the building fire layout data in different fields as the first fire sub-information saved in the first knowledge sub-graph, the fire law data as the second fire sub-information saved in the second knowledge sub-graph, the fire and corresponding treatment information data as the third fire sub-information saved in the third knowledge sub-graph, and the fire simulation and management data as the fourth fire sub-information saved in the fourth knowledge sub-graph;

[0015] S102. According to the first fire sub-information, the second fire sub-information, the third fire sub-information, and the fourth fire sub-information, use the set classification criteria and classification algorithms to classify each fire sub-information hierarchically to obtain multi-level first fire sub-information, second fire sub-information, third fire sub-information, and fourth fire sub-information;

[0016] S103. According to the classified fire sub-information corresponding to the j-th level in the i-th fire sub-information obtained, use the entity extraction algorithm to obtain ; where represents the entity triple set corresponding to the classified fire sub-information at the j-th level in the i-th fire sub-information;

[0017] S104. Repeat the process of S103 to obtain the entity triple set corresponding to each level in the i-th fire sub-information, and perform hierarchical label annotation on each entity triple set;

[0018] S105. According to the entity triple set in the i-th fire sub-information after hierarchical label annotation, obtain the knowledge sub-graph corresponding to the i-th fire sub-information. Repeat the process of S103 - S105 to obtain the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph.

[0019] Specifically, the specific steps for constructing the fire multi - related knowledge graph also include:

[0020] S106. According to the fire sub - information saved in the j - th level corresponding node in the i - th knowledge sub - graph and the fire sub - information saved in the j - th level corresponding node in the k - th knowledge sub - graph, through a similarity matching algorithm, obtain two nodes with the highest node matching degree and the lowest load in the j - th level and ; where represents the h - th node in the j - th level of the i - th knowledge sub - graph, represents the g - th node in the j - th level of the k - th knowledge sub - graph;

[0021] S107. Utilize and and their corresponding node matching degrees to construct the graph connection of the i - th knowledge sub - graph and the k - th knowledge sub - graph at the j - th level;

[0022] S108. Repeat the process of S106 - S107 to obtain the graph connections of the i - th knowledge sub - graph and the k - th knowledge sub - graph at the same level;

[0023] S109. Repeat the process of S106 - S108 to obtain the graph connections corresponding to each pair of the first knowledge sub - graph, the second knowledge sub - graph, the third knowledge sub - graph, and the fourth knowledge sub - graph;

[0024] S110. According to the first knowledge sub - graph, the second knowledge sub - graph, the third knowledge sub - graph, the fourth knowledge sub - graph, and all the corresponding graph connections, construct the fire multi - related knowledge graph.

[0025] Specifically, the steps for determining the main knowledge sub - graph and the collaborative knowledge sub - graph include:

[0026] S201. Obtain the user's historical input semantic information, and through a scene classification algorithm, obtain the scene type corresponding to each user input semantic information;

[0027] S202. According to the user input semantic information under the corresponding scene type, the historical access number and search frequency of each knowledge sub - graph node, and the fire multi - related knowledge graph, obtain the correlation matching degrees between each scene type and the first knowledge sub - graph, the second knowledge sub - graph, the third knowledge sub - graph, and the fourth knowledge sub - graph through a correlation matching algorithm;

[0028] S203. Take the knowledge sub - graph corresponding to the largest correlation matching degree as the main knowledge sub - graph, and take the remaining knowledge sub - graphs as the collaborative knowledge sub - graphs.

[0029] Specifically, the main knowledge sub-graph adopts a full-node search method; the collaborative knowledge sub-graph adopts a differential search method; the steps of the differential search include:

[0030] S2031. Obtain the proportion of the number of search nodes corresponding to each collaborative knowledge sub-graph and the proportion of the search frequency of a single node in each collaborative knowledge sub-graph according to the historical access number and search frequency of each knowledge sub-graph node in the collaborative knowledge sub-graph under each scenario type;

[0031] S2032. Obtain the search importance degree corresponding to each collaborative knowledge sub-graph according to the product of the correlation matching degree between each collaborative knowledge sub-graph and the corresponding scenario type and the proportion of the corresponding number of search nodes, and construct the first differential index information between each scenario type and each collaborative knowledge sub-graph based on the construction principle of the database index by using the search importance degree.

[0032] Specifically, the steps of the differential search further include:

[0033] S2033. Obtain the search importance degree of the corresponding single node in each collaborative knowledge sub-graph according to the product of the correlation matching degree between each collaborative knowledge sub-graph and the corresponding scenario type, the proportion of the corresponding number of search nodes, and the proportion of the search frequency of a single node, and construct the second differential index information by using the search importance degree of a single node in each collaborative knowledge sub-graph and the corresponding scenario type;

[0034] S2034. Configure a main-collaborative switching node in the fire multi-correlation knowledge graph, save the first differential index information and the second differential index information into the main-collaborative switching node, and perform main-collaborative knowledge sub-graph switching and fire sub-information differential search through the main-collaborative switching node according to the user input semantic information obtained in real time and the scenario type in the corresponding scenario feature space.

[0035] Specifically, the steps for constructing and training the multi-graph joint reasoning model include:

[0036] S301. Construct an inference input sequence according to the user input semantic information, the corresponding scenario feature space, and the fire multi-correlation knowledge graph information;

[0037] S302. Determine the main knowledge sub-graph and collaborative knowledge sub-graph information according to the scenario feature space in the inference input sequence through the information invocation layer, and perform main-collaborative knowledge sub-graph switching on the fire multi-correlation knowledge graph through the determined information by invoking the main-collaborative switching node, and obtain the first differential index information and the second differential index information saved in the main-collaborative switching node;

[0038] S303. Input the switching information for the primary - collaborative knowledge sub - graph switching, the user - input semantic information, the first difference index information, and the second difference index information into the search path layer to obtain the full - search main path, the first difference search path, the second difference search path, and the third difference search path;

[0039] S304. Input the full - search main path, the first difference search path, the second difference search path, the third difference search path, and the determined primary - collaborative knowledge sub - graph information into the matching search layer to conduct an initial fire - fighting plan search for the corresponding scenario, and obtain the fire - fighting plan sequence for the corresponding scenario, as well as the corresponding search matching degree and search matching loss.

[0040] Specifically, the steps for constructing and training the multi - graph joint reasoning model further include:

[0041] S305. Set a search matching degree threshold. When at least one of the search matching degrees of the fire - fighting plans is greater than the search matching degree threshold, then use the fire - fighting plan corresponding to the maximum search matching degree that meets the threshold condition as the fire - fighting plan for the current scenario;

[0042] S306. Set a fire - fighting plan screening threshold. When all the search matching degrees of the fire - fighting plans are less than or equal to the search matching degree threshold, use the fire - fighting plan information with a search matching degree greater than the fire - fighting plan screening threshold in the fire - fighting plan sequence for the corresponding scenario as the fire - fighting plan inference information sequence through the fire - fighting plan screening threshold;

[0043] S307. Input the corresponding scenario feature space, the fire - fighting plan inference information sequence, and the full - search main path, the first difference search path, the second difference search path, and the third difference search path into the fusion path inference layer constructed by the convolutional attention algorithm and the path generation algorithm to obtain the corresponding full - search inference main path, the first difference inference path, the second difference inference path, and the third difference inference path, and embed the search matching degree threshold into all the inference paths;

[0044] S308. Input the full - search inference main path, the first difference inference path, the second difference inference path, the third difference inference path, and the corresponding primary - collaborative knowledge sub - graph information into the scenario inference layer constructed by the multi - modal feature fusion - based scenario text recognition algorithm to obtain each fire - fighting plan inference information and the corresponding inference search matching degree on the full - search inference main path and each difference inference path;

[0045] S309. According to the inference search matching degree corresponding to each fire - fighting plan inference on the full - search inference main path and each difference inference path and the search matching degree threshold embedded in the corresponding inference, obtain the inference search loss corresponding to each inference process on the full - search inference main path and each difference inference path;

[0046] S310. Set an approximation loss threshold, and construct an inference approximation loss function using the error between the sum of the inference search losses of each inference process corresponding to the inference path and the approximation loss threshold.

[0047] S311. Set a training loss threshold and a training period, and construct a multi-graph joint inference model training loss function using the sum of the inference approximation loss function and the search matching loss. When the inference search loss of each inference path is less than the approximation loss threshold within the training period and the multi-graph joint inference model training loss function is less than the training loss threshold, obtain the trained multi-graph joint inference model.

[0048] Specifically, the full search main path in S303 corresponds to the main knowledge sub-graph; the first differential search path corresponds to the knowledge sub-graph one included in the collaborative knowledge sub-graph; the second differential search path corresponds to the knowledge sub-graph two included in the collaborative knowledge sub-graph; the third differential search path corresponds to the knowledge sub-graph three included in the collaborative knowledge sub-graph; the knowledge sub-graph one, knowledge sub-graph two, and knowledge sub-graph three included in the collaborative knowledge sub-graph are the remaining three knowledge sub-graphs among the first knowledge sub-graph, second knowledge sub-graph, third knowledge sub-graph, and fourth knowledge sub-graph after removing the one used as the main knowledge sub-graph, and the knowledge sub-graph one, knowledge sub-graph two, and knowledge sub-graph three correspond one-to-one with the remaining three knowledge sub-graphs.

[0049] An electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements an AI intelligent data processing method for the fire protection field.

[0050] A computer-readable storage medium stores computer instructions, and when the computer instructions run, they execute an AI intelligent data processing method for the fire protection field.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] In view of the deficiencies of the prior art, the present invention constructs a multi - related knowledge graph for fire protection, integrates building fire protection layout, fire protection laws, fire handling information and simulation management data, and uses natural language processing and graph database technology to achieve in - depth understanding and rapid response to complex fire protection scenarios; with the help of a multi - input intention reasoning model, it accurately captures the semantic information and scene features of user input, dynamically switches the main - collaborative knowledge sub - graphs, ensuring the accuracy and timeliness of providing professional fire assessment plans and fire protection strategy plans for different scenarios. At the same time, switching the main - collaborative knowledge sub - graphs for different scenarios and combining the built - in differential search index can increase the accuracy of search, reduce the scope and amount of search data, and reduce search latency; the introduction of the multi - graph joint reasoning model not only optimizes the path of plan generation, but also continuously improves the prediction accuracy through machine learning, solving the problems of search delay and inaccurate answers in the prior art; in addition, the multi - related knowledge graph for fire protection and the built - in switching nodes and differential search methods in the present invention are also applicable to large - scale fire protection scenarios covering multiple sub - fields, capable of efficiently managing and collaboratively processing massive data, providing instant, professional and personalized fire protection solutions for users, and greatly improving the efficiency and reliability of dealing with emergencies such as fires. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flowchart of an AI intelligent data processing method for the fire protection field in Embodiment 1 of the present invention;

[0054] Figure 2 It is a structural diagram of the multi - graph joint reasoning model in Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] Embodiment 1

[0056] Please refer to Figure 1 , an embodiment provided by the present invention: an AI intelligent data processing method for the fire protection field, the steps include:

[0057] S1. Obtain building fire protection layout data, fire protection law data, fire and corresponding handling information data, and fire protection simulation and management data in different fields, and construct a multi - related knowledge graph for fire protection through natural language algorithms and graph databases;

[0058] Further, the multi - related knowledge graph for fire protection in this embodiment includes a first knowledge sub - graph, a second knowledge sub - graph, a third knowledge sub - graph and a fourth knowledge sub - graph; the specific steps for constructing the multi - related knowledge graph for fire protection include:

[0059] S101. Set the first fire sub - information saved in the first knowledge sub - graph for the building fire - fighting layout data in different fields, the second fire sub - information saved in the second knowledge sub - graph for the fire - fighting legal data, the third fire sub - information saved in the third knowledge sub - graph for the fire and corresponding handling information data, and the fourth fire sub - information saved in the fourth knowledge sub - graph for the fire - fighting simulation and management data;

[0060] S102. According to the first fire sub - information, the second fire sub - information, the third fire sub - information and the fourth fire sub - information, through the set classification criteria combined with the classification algorithm, classify each fire sub - information hierarchically to obtain multi - level first fire sub - information, second fire sub - information, third fire sub - information and fourth fire sub - information;

[0061] Further, the classification criteria specifically include:

[0062] Conduct data analysis on the building fire - fighting layout data in different fields, the fire - fighting legal data, the fire and corresponding handling information data, and the fire - fighting simulation and management data respectively to obtain the corresponding second - level classification data under each major category, and so on, to obtain the hierarchical data corresponding to each level, where the first - level classification is the building fire - fighting layout data in different fields, the fire - fighting legal data, the fire and corresponding handling information data, and the fire - fighting simulation and management data.

[0063] Exemplarily, the first - level classification:

[0064] The classification at the first level is directly based on the four major types of basic data sources, and each category constitutes the basic structure of the knowledge sub - graph;

[0065] Building fire - fighting layout data in different fields (the first knowledge sub - graph):

[0066] Data types: Include fire - fighting design drawings of various buildings (such as residential, commercial, industrial, etc.), safety passage layouts, locations of fire - fighting facilities (such as fire extinguishers, fire hydrants, sprinkler systems, etc.), emergency exit signs, etc.; Specifically, for example, the specific types corresponding to the second level can be different major types of buildings, such as residential, commercial, industrial, etc.; At the third level, the residential, commercial, and industrial are further refined. For example, residential buildings are divided into high - end residential, mid - range residential, and low - end residential or population - dense or sparsely populated according to housing prices, geographical locations, and population density; Industrial buildings are divided into different grades according to the safety levels required by the planned building. For example, the higher the fire - fighting grade for chemical plants, and the lower the fire - fighting grade for ordinary small factories; The classification levels are divided specifically according to specific actual needs; The following fire - fighting legal data, fire and corresponding handling information data, and fire - fighting simulation and management data are also divided similarly according to this process;

[0067] Data information: Specific fire safety requirements, evacuation route planning, fire compartment division, etc. for each type of building.

[0068] Fire law data (the second knowledge sub-graph):

[0069] Data types: National or regional fire regulations, industry standards, building codes, etc.

[0070] Data information: Specific clauses of laws and regulations, update records, scope of application, legal responsibilities, etc.

[0071] Fire and corresponding handling information data (the third knowledge sub-graph):

[0072] Data types: Historical fire cases, fire cause analysis, fire fighting operation reports, casualty situations, etc.

[0073] Data information: Time and location of the fire, type of ignition source, loss assessment caused, key decision-making points during the rescue process, etc.

[0074] Fire simulation and management data (the fourth knowledge sub-graph):

[0075] Data types: Results of fire simulation experiments, records of emergency plan drills, risk management strategies, etc.

[0076] Data information: Simulation condition settings, comparison of expected and actual results, problems found during the drill and improvement measures, etc.

[0077] In the process of formulating the classification criteria corresponding to the fire law data, fire and corresponding handling information data, and fire simulation and management data, refer to the classification process of building fire layout data in different fields. Technical personnel in this field refine and layer specifically according to the major categories set in this embodiment. The specific number of layers for layering is specifically set by technical personnel in this field according to the actual data storage requirements and hardware facility configurations. The more detailed the data division, the more accurate the subsequent search for corresponding answers.

[0078] S103. According to the classified fire sub-information corresponding to the j-th level in the i-th fire sub-information obtained, through the entity extraction algorithm, obtain ; where represents the set of entity triples corresponding to the classified fire sub-information of the j-th level in the i-th fire sub-information;

[0079] S104. Repeat the process of S103 to obtain the set of entity triples corresponding to each level in the i-th fire sub-information, and perform hierarchical label annotation on each set of entity triples; further, the hierarchical label annotation is to label the corresponding hierarchical level for each set of entity triples.

[0080] S105. Obtain the knowledge sub-graph corresponding to the \(i\)-th fire sub-information according to the entity triple set in the \(i\)-th fire sub-information after being labeled with the classification label. Repeat the process of S103 - S105 to obtain the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph;

[0081] S106. According to the fire sub-information saved in the \(j\)-th level corresponding node in the \(i\)-th knowledge sub-graph and the fire sub-information saved in the \(j\)-th level corresponding node in the \(k\)-th knowledge sub-graph, through the similarity matching algorithm, obtain two nodes with the highest node matching degree and the lowest load in the \(j\)-th level and ; where represents the \(h\)-th node in the \(j\)-th level of the \(i\)-th knowledge sub-graph, represents the \(g\)-th node in the \(j\)-th level of the \(k\)-th knowledge sub-graph;

[0082] S107. Utilize and and the corresponding node matching degrees to construct the graph connection of the \(i\)-th knowledge sub-graph and the \(k\)-th knowledge sub-graph at the \(j\)-th level;

[0083] S108. Repeat the process of S106 - S107 to obtain the graph connections corresponding to the same levels of the \(i\)-th knowledge sub-graph and the \(k\)-th knowledge sub-graph;

[0084] S109. Repeat the process of S106 - S108 to obtain the graph connections corresponding to each pair of the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph;

[0085] S110. According to the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, the fourth knowledge sub-graph, and all the corresponding graph connections, construct the fire multi-associated knowledge graph.

[0086] By finely constructing a multi - related knowledge graph for fire protection, the efficient organization and utilization of complex information in the fire protection field have been achieved. Specifically, first, through multi - level classification of building fire layouts, fire laws, fire handling information, and simulation management data in different fields, the systematicness and orderliness of the data are ensured. This structured classification method enables each level of data to be accurately refined into the most required application scenarios. For example, corresponding fire protection strategies are formulated for different types of buildings or safety levels. In further data processing, entity extraction algorithms are used to obtain entity triple sets from each level of classification and perform hierarchical label annotation on them. This not only improves the readability of the data but also provides a solid foundation for subsequent knowledge graph construction. Through similarity matching algorithms, the best - matching nodes are found and connections are established between different knowledge sub - graphs. This process effectively integrates scattered information resources, forms an organic whole, and enhances the connectivity and practicality of the knowledge graph. The finally constructed multi - related knowledge graph for fire protection can achieve cross - domain and cross - level data collaboration, greatly improving the user query efficiency and accuracy. It can quickly locate the core problem, provide professional and accurate fire assessment plans and fire protection strategy plans, reducing search latency and inaccurate answer problems. In addition, this graph has high flexibility and scalability and can be continuously updated and improved as new data is added, ensuring the timeliness and reliability of information. This efficient resource integration and intelligent analysis ability provide strong support for fire protection management and emergency response, significantly enhancing the ability and effect of dealing with emergencies such as fires.

[0087] S2. Through the configured multi - input intention inference model, obtain the user - input semantic information and the corresponding scenario feature space, and through the user - input semantic information, the corresponding scenario feature space, and the multi - related knowledge graph for fire protection, determine the main knowledge sub - graph and the collaborative knowledge sub - graph in the multi - related knowledge graph for fire protection, and perform main - collaborative knowledge sub - graph switching according to the scenario feature space through the configured main - collaboration switching node;

[0088] Furthermore, in this embodiment, the multi - input intention inference model is composed of a pre - trained scenario inference model and a user demand inference library constructed from inference process data such as voice, text, gestures, and lip - shapes during historical inferences. The user demand inference library includes information such as voice, text, gestures, and lip - shapes input by the user, and the corresponding inferred user demand text information and the corresponding scenario space. For example, when the user asks: Please output a fire prevention plan according to the characteristics of the building. It is inferred from the text information that the user's demand is a fire prevention plan, and the corresponding scenario is the building scenario. The pre - trained scenario inference model is constructed by combining existing voice, text, gesture, and lip - shape recognition algorithms and deep inference algorithms, and is used to obtain the user demand intention and the corresponding scenario space characteristics according to the user - input information.

[0089] Further, the steps for determining the main knowledge sub-graph and the collaborative knowledge sub-graph in this embodiment include:

[0090] S201. Obtain the semantic information of the user's historical input, and through the scene classification algorithm, obtain the scene type corresponding to each user input semantic information;

[0091] S202. According to the user input semantic information under the corresponding scene type, the historical access count and search frequency of each knowledge sub-graph node, and the fire multi-association knowledge graph, obtain the association matching degrees corresponding to each scene type and the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph through the association matching algorithm;

[0092] S203. Take the knowledge sub-graph corresponding to the largest association matching degree as the main knowledge sub-graph, and take the remaining knowledge sub-graphs as the collaborative knowledge sub-graphs;

[0093] Further, in this embodiment, the main knowledge sub-graph adopts the full-node search method; the collaborative knowledge sub-graph adopts the differential search method; further, the steps of the differential search in this embodiment include:

[0094] S2031. According to the historical access count and search frequency of each knowledge sub-graph node in the collaborative knowledge sub-graph under each scene type, obtain the proportion of the corresponding search node count of each collaborative knowledge sub-graph and the proportion of the single-node search frequency in each collaborative knowledge sub-graph;

[0095] S2032. According to the product of the association matching degree of each collaborative knowledge sub-graph and the corresponding scene type and the proportion of the corresponding search node count, obtain the search importance degree corresponding to each collaborative knowledge sub-graph, and based on the construction principle of the database index, construct the first differential index information between each scene type and each collaborative knowledge sub-graph;

[0096] S2033. According to the product of the association matching degree of each collaborative knowledge sub-graph and the corresponding scene type, the proportion of the corresponding search node count, and the proportion of the single-node search frequency, obtain the search importance degree of the corresponding single node in each collaborative knowledge sub-graph, and construct the second differential index information using the search importance degree of the single node in each collaborative knowledge sub-graph and the corresponding scene type;

[0097] S2034. Configure a main-collaborative switching node in the fire multi-association knowledge graph, save the first differential index information and the second differential index information to the main-collaborative switching node, and through the main-collaborative switching node, perform the switching between the main and collaborative knowledge sub-graphs and the differential search of the fire sub-information according to the real-time obtained user input semantic information and the scene type in the corresponding scene feature space.

[0098] This process significantly improves the accuracy and response speed of fire information query through a multi-input intention inference model and a master-collaborative knowledge sub-graph switching mechanism. Specifically, the multi-input intention inference model combines a pre-trained scenario inference model and a user demand inference library, and can accurately infer the actual needs of users and the corresponding scenario feature space from various input methods such as voice, text, gestures, and lip movements, ensuring a more comprehensive and accurate understanding of user intentions; in the process of determining the master knowledge sub-graph and the collaborative knowledge sub-graph, the system first identifies the scenario type corresponding to the user input semantic information through a scenario classification algorithm, and then uses an association matching algorithm to calculate the association matching degree between each scenario type and each knowledge sub-graph, and selects the one with the highest matching degree as the master knowledge sub-graph, and the rest as the collaborative knowledge sub-graphs. This design enables the system to quickly focus on the most relevant fire information for a specific scenario and provide professional and targeted solutions; in particular, the master knowledge sub-graph uses full-node search to ensure that no key information is missed; while the collaborative knowledge sub-graphs optimize through differential search, and construct the first and second differential index information based on the number of historical visits and search frequencies, improving the search efficiency. The existence of the master-collaborative switching node enables the system to flexibly switch between different scenarios, ensuring the coherence and accuracy of information retrieval; in summary, this process not only solves the problem in the prior art that it is impossible to provide accurate solutions for specific scenarios, but also greatly reduces the search latency and provides instant and personalized fire protection plans. By continuously updating the user demand inference library and optimizing the differential index, the system can continuously improve its performance, adapt to new fire challenges, and thus provide users with more intelligent and efficient fire safety services. This method effectively promotes the scientific and intelligent development of fire protection work and enhances the ability to respond to emergencies such as fires.

[0099] S3. Construct a multi-graph joint inference model, input the master knowledge sub-graph and collaborative knowledge sub-graph information, user input semantic information and scenario feature space into the multi-graph joint inference model to obtain a set of user fire demand plans, and at the same time use the set of user fire demand plans to construct an optimal inference search path and an inference approximation loss function to train the multi-graph joint inference model to obtain a trained multi-graph joint inference model;

[0100] Further, please refer to Figure 2 , in this embodiment, the steps of constructing and training the multi-graph joint inference model include:

[0101] S301. Construct an inference input sequence according to the user input semantic information, the corresponding scenario feature space and the fire multi-association knowledge graph information;

[0102] S302. According to the scene feature space in the inference input sequence, through the information invocation layer, determine the main knowledge sub-graph and co-knowledge sub-graph information, and through the determined information, invoke the main-co switching node to perform the main-co knowledge sub-graph switching on the fire multi-association knowledge graph, and obtain the first difference index information and the second difference index information saved in the main-co switching node; further, the information invocation layer in this embodiment is constructed by the SAC algorithm;

[0103] S303. Input the switching information of the main-co knowledge sub-graph switching, the user input semantic information, the first difference index information and the second difference index information into the search path layer to obtain the full search main path, the first difference search path, the second difference search path, and the third difference search path; the search path layer in this embodiment is constructed by the path generation algorithm;

[0104] Further, the full search main path in this step corresponds to the main knowledge sub-graph; the first difference search path corresponds to the knowledge sub-graph one included in the co-knowledge sub-graph; the second difference search path corresponds to the knowledge sub-graph two included in the co-knowledge sub-graph; the third difference search path corresponds to the knowledge sub-graph three included in the co-knowledge sub-graph; the knowledge sub-graph one, knowledge sub-graph two, and knowledge sub-graph three included in the co-knowledge sub-graph are the remaining three knowledge sub-graphs after removing the one used as the main knowledge sub-graph from the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph, and the knowledge sub-graph one, knowledge sub-graph two, and knowledge sub-graph three correspond one-to-one with the remaining three knowledge sub-graphs.

[0105] S304. Input the full search main path, the first difference search path, the second difference search path, the third difference search path, and the determined main-co knowledge sub-graph information into the matching search layer to perform the initial fire protection plan search for the corresponding scene, and obtain the corresponding scene fire protection plan sequence and the corresponding search matching degree and search matching loss; further, the matching search layer in this embodiment is constructed by the attention matching algorithm;

[0106] S305. Set the search matching degree threshold. When at least one of the search matching degrees of the fire protection plans is greater than the search matching degree threshold, then use the fire protection plan corresponding to the maximum search matching degree that satisfies the threshold condition as the fire protection plan corresponding to the current scene;

[0107] S306. Set the fire protection plan screening threshold. When the search matching degrees of all the fire protection plans are less than or equal to the search matching degree threshold, use the fire protection plan information with a search matching degree greater than the fire protection plan screening threshold in the corresponding scene fire protection plan sequence as the fire protection plan inference information sequence through the fire protection plan screening threshold;

[0108] S307. Input the corresponding scenario feature space, the fire protection plan inference information sequence, the full search main path, the first differential search path, the second differential search path, and the third differential search path into the fusion path inference layer constructed by the convolutional attention algorithm and the path generation algorithm to obtain the corresponding full search inference main path, the first differential inference path, the second differential inference path, and the third differential inference path, and embed the search matching degree threshold into all inference paths;

[0109] Further, in this embodiment, the full search inference main path is obtained from the full search main path through the fusion path inference layer; the first differential inference path is obtained from the first differential search path through the fusion path inference layer; the second differential inference path is obtained from the second differential search path through the fusion path inference layer; the third differential inference path is obtained from the third differential search path through the fusion path inference layer.

[0110] S308. Input the full search inference main path, the first differential inference path, the second differential inference path, the third differential inference path, and the corresponding master - collaborative knowledge sub - graph information into the scenario inference layer constructed by the multi - modal feature fusion scenario text recognition algorithm to obtain each fire protection plan inference information and the corresponding inference search matching degree on the full search inference main path and each differential inference path;

[0111] Further, in this embodiment, each fire protection plan inference information and the corresponding inference search matching degree on each differential inference path refer to the inference search matching degree obtained by single - time inference through information inference on the first differential inference path, the second differential inference path, and the third differential inference path respectively.

[0112] S309. According to the inference search matching degree corresponding to each fire protection plan inference on the full search inference main path and each differential inference path and the search matching degree threshold built into the corresponding inference, obtain the inference search loss corresponding to each inference process on the full search inference main path and each differential inference path;

[0113] S310. Set an approximation loss threshold, and construct an inference approximation loss function using the error between the sum of the inference search losses of the inference paths corresponding to each inference process and the approximation loss threshold;

[0114] S311. Set a training loss threshold and a training cycle, and construct a multi - graph joint inference model training loss function using the sum of the inference approximation loss function and the search matching loss. When the inference search loss of each inference path is less than the approximation loss threshold within the training cycle and the multi - graph joint inference model training loss function is less than the training loss threshold, obtain the trained multi - graph joint inference model.

[0115] This process significantly improves the efficiency, accuracy and intelligence level of fire demand plan generation by constructing a multi-graph joint reasoning model. Specifically, the system first constructs a reasoning input sequence based on the user input semantic information and scene feature space, and uses the SAC algorithm to determine the main knowledge subgraph and the collaborative knowledge subgraph in the information call layer to realize the main-cooperative switching node configuration, thereby obtaining differential index information. Then, the full search main path and multiple differential search paths are generated through the path generation algorithm to ensure comprehensive coverage of different knowledge subgraphs. The matching search layer uses the attention matching algorithm to evaluate the matching degree and loss of the initial fire plan, sets the threshold to screen the optimal plan, and improves the accuracy of plan selection; the fusion path reasoning layer combines the convolutional attention algorithm and the path generation algorithm to optimize the reasoning path, and has a built-in search matching degree threshold to improve accuracy, ensuring that the most suitable solution can be found in each reasoning process. The scene reasoning layer further refines the fire plan reasoning information on each path through the scene text recognition algorithm with multimodal feature fusion, enhancing the adaptability and flexibility of the system. Based on this, the system can provide personalized and professional fire plans according to the characteristics of different scenarios, greatly shortening the response time and reducing human errors; in addition, by constructing an inference approximation loss function and setting a training loss threshold, the system can automatically adjust parameters, optimize model performance, and ensure the quality of the output plan. This method not only improves the speed and accuracy of fire plan generation, but also enhances the robustness and learning ability of the system; continuous training allows the model to continuously adapt to new data and maintain its advancement and practicality; ultimately, the invention provides users with an intelligent and efficient fire safety service platform, significantly improving the ability to respond to emergencies such as fires, and protecting the lives and property of the public. This efficient, accurate, and intelligent fire plan generation mechanism provides strong technical support for fire management and emergency response, and promotes the scientific and intelligent development of firefighting work.

[0116] S4. Based on the real-time acquired user input semantic information and the corresponding scene feature space, the user fire demand plan is obtained through the trained multi-graph joint reasoning model, and the real-time user input semantic information, scene feature space and the corresponding user fire demand plan are saved in the fire multi-association knowledge graph.

[0117] Furthermore, the fire data processing method proposed in this embodiment can quickly and accurately provide fire emergency plans and related legal and regulatory issues related to corresponding fields and corresponding scenarios.

[0118] For example: The user asks: Please output a fire prevention plan according to the characteristics of the building. According to the content of this question, it is about the building fire protection plan. The main corresponding domain data is the first fire protection sub-information content in the first knowledge sub-graph, that is, the first knowledge sub-graph is the main knowledge sub-graph. Through the main knowledge sub-graph, the corresponding building fire protection layout, personnel gathering distribution and building pattern are comprehensively obtained. Through the collaborative third fire protection sub-information content in the third knowledge sub-graph and the fourth fire protection sub-information content in the fourth knowledge sub-graph, a corresponding building fire prevention plan can be quickly generated and corresponding specific examples can be given accordingly;

[0119] Answer: According to data statistics, 8 households have moved into the building, and each household's room has a smoke-triggered sprinkler device. After monitoring, the health status of the smoke-triggered sprinkler device is good and it can be triggered at any time. The specific usage method and corresponding distribution of the smoke-triggered sprinkler device.

[0120] Question: What is the latest fire protection policy for chemical plants and what should we do? According to this question, its main field still corresponds to the fire protection plan in the building field, that is, the first knowledge sub-graph is the main knowledge sub-graph. Through the main knowledge sub-graph, the corresponding building fire protection layout is comprehensively obtained; through the fire protection layout, building pattern and safety requirement information regarding chemical plant buildings in the lower-level classification within the first fire protection sub-information content in the main knowledge sub-graph, and by collaborating with the fire protection laws and regulations information in the second fire protection sub-information in the second knowledge sub-graph and the historical fire and corresponding handling situation information in the third fire protection sub-information in the third knowledge sub-graph, a corresponding chemical plant building fire prevention plan can be quickly generated and corresponding specific examples can be given accordingly;

[0121] Answer: According to the relevant policies issued on November 28, 2024, the latest chemical fire protection policy shows a tightening trend and is more stringent than in the past. Based on this trend change, the layout of fire protection products can be carried out for our company.

[0122] Embodiment 2

[0123] An electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements an AI intelligent data processing method for the fire protection field.

[0124] A computer-readable storage medium stores computer instructions, and when the computer instructions run, they execute an AI intelligent data processing method for the fire protection field.

[0125] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments without departing from the spirit and scope of the present invention as protected by the claims. These all fall within the protection scope of the present invention.

[0126] If the technical solution of the present disclosure involves personal information, before the product applying the technical solution of the present disclosure processes personal information, the rules for processing personal information have been clearly informed and the personal's autonomous consent has been obtained. If the technical solution of the present disclosure involves sensitive personal information, before the product applying the technical solution of the present disclosure processes sensitive personal information, the personal's separate consent has been obtained and the requirements of "express consent" have been met at the same time. For example, at a personal information collection device such as a camera, a clear and prominent sign is set to inform that the personal information collection scope has been entered and personal information will be collected. If the personal voluntarily enters the collection scope, it is regarded as consenting to the collection of their personal information; or on the device for processing personal information, when the rules for processing personal information are informed by obvious signs / information, personal authorization is obtained through pop-up messages or by asking the personal to upload their personal information by themselves; among them, the rules for processing personal information may include information such as the personal information processor, the purpose of processing personal information, the processing method, and the types of personal information processed.

Claims

1. An AI intelligent data processing method for fire protection, characterized in that the steps include: S1. Obtain building fire layout data, fire law data, fire and corresponding processing information data, and fire simulation and management data in different fields, and construct a fire protection multi-association knowledge graph through natural language algorithms and graph databases; S2. Obtain user input semantic information and corresponding scene feature space through the configured multi-input intention reasoning model, and determine the main knowledge subgraph and collaborative knowledge subgraph in the fire protection multi-association knowledge graph through the user input semantic information, the corresponding scene feature space and the fire protection multi-association knowledge graph, and perform the main-collaborative knowledge subgraph switching according to the scene feature space through the configured main-cooperative switching node; S3. Construct a multi-graph joint reasoning model, input the main knowledge sub-graph and the collaborative knowledge sub-graph information, the user input semantic information and the scene feature space into the multi-graph joint reasoning model, obtain the user fire demand plan set, and use the user fire demand plan set to construct the optimal reasoning search path and reasoning approximation loss function to train the multi-graph joint reasoning model, and obtain the trained multi-graph joint reasoning model; S4. According to the real-time acquired user input semantic information and the corresponding scene feature space, the user fire demand plan is obtained through the trained multi-graph joint reasoning model, and the real-time user input semantic information, the scene feature space and the corresponding user fire demand plan are saved in the fire multi-association knowledge graph; The fire protection multi-association knowledge graph includes a first knowledge sub-graph, a second knowledge sub-graph, a third knowledge sub-graph and a fourth knowledge sub-graph; The steps of determining the main knowledge sub-graph and the collaborative knowledge sub-graph include: S201, obtaining user historical input semantic information, and obtaining the scene type corresponding to each user input semantic information through a scene classification algorithm; S202, according to the semantic information input by the user under the corresponding scene type, the number of historical visits and search frequency of each knowledge sub-graph node, and the fire protection multi-association knowledge graph, obtain the association matching degree corresponding to each scene type and the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph through an association matching algorithm; S203, taking the knowledge sub-graph corresponding to the maximum correlation matching degree as the main knowledge sub-graph, and taking the remaining knowledge sub-graphs as collaborative knowledge sub-graphs; The main knowledge subgraph adopts a full-node search method; the collaborative knowledge subgraph adopts a difference search method; the difference search steps include: S2031. According to the number of historical visits and search frequency of each knowledge sub-graph node in the collaborative knowledge sub-graph under each scenario type, obtain the proportion of the number of search nodes corresponding to each collaborative knowledge sub-graph and the proportion of the search frequency of a single node in each collaborative knowledge sub-graph; S2032, obtaining the search importance corresponding to each collaborative knowledge subgraph according to the product of the association matching degree between each collaborative knowledge subgraph and the corresponding scene type and the proportion of the corresponding search node number, and constructing the first difference index information between each scene type and each collaborative knowledge subgraph based on the construction principle of the search importance based on the database index; S2033, obtaining the search importance of the corresponding single node in each collaborative knowledge subgraph according to the product of the association matching degree between each collaborative knowledge subgraph and the corresponding scene type, the proportion of the number of corresponding search nodes and the proportion of the search frequency of a single node, and constructing the second difference index information using the search importance of the single node in each collaborative knowledge subgraph and the corresponding scene type; S2034, configuring a main-cooperative switching node in the fire protection multi-association knowledge graph, and saving the first difference index information and the second difference index information in the main-cooperative switching node, and performing main-cooperative knowledge sub-graph switching and fire protection sub-information difference search through the main-cooperative switching node according to the user input semantic information obtained in real time and the scene type in the corresponding scene feature space; The steps of constructing and training the multi-graph joint reasoning model include: S301, constructing a reasoning input sequence according to user input semantic information, corresponding scene feature space, and fire protection multi-association knowledge graph information; S302, according to the scene feature space in the reasoning input sequence, determine the main knowledge subgraph and the collaborative knowledge subgraph information through the information call layer, and call the main-cooperative switching node through the determined information to switch the fire protection multi-association knowledge graph between the main and collaborative knowledge subgraphs, and obtain the first difference index information and the second difference index information stored in the main-cooperative switching node; S303, input the switching information of the main-collaborative knowledge sub-graph switching, the user input semantic information, the first difference index information and the second difference index information into the search path layer to obtain the full search main path and the first difference search path, the second difference search path, and the third difference search path; S304, inputting the full search main path, the first difference search path, the second difference search path, the third difference search path and the determined main-collaborative knowledge sub-graph information into the matching search layer to search for the initial fire plan for the corresponding scene, and obtaining the fire plan sequence for the corresponding scene and the corresponding search matching degree and search matching loss; S305, setting a search matching degree threshold, when at least one of the search matching degrees of the fire emergency plans is greater than the search matching degree threshold, the fire emergency plan that meets the threshold condition and corresponds to the maximum search matching degree is used as the fire emergency plan corresponding to the current scene; S306, setting a fire plan screening threshold. When the search matching degrees of the fire plans are all less than or equal to the search matching degree threshold, the fire plan information with the corresponding search matching degree greater than the fire plan screening threshold in the fire plan sequence of the corresponding scene is used as the fire plan reasoning information sequence through the fire plan screening threshold; S307, input the corresponding scene feature space, the fire plan reasoning information sequence and the full search main path, the first difference search path, the second difference search path, and the third difference search path into the fusion path reasoning layer constructed by the convolutional attention algorithm and the path generation algorithm, obtain the corresponding full search reasoning main path, the first difference reasoning path, the second difference reasoning path, and the third difference reasoning path, and build the search matching degree threshold into all the reasoning paths; S308, input the full search reasoning main path, the first difference reasoning path, the second difference reasoning path, the third difference reasoning path and the corresponding main-collaborative knowledge sub-graph information into the scene reasoning layer constructed by the multimodal feature fusion scene text recognition algorithm, and obtain each fire plan reasoning information on the full search reasoning main path and each difference reasoning path and the corresponding reasoning search matching degree; S309, according to the reasoning search matching degree corresponding to each fire plan reasoning on the full search reasoning main path and each difference reasoning path and the search matching degree threshold built in the corresponding reasoning, obtain the reasoning search loss corresponding to each reasoning process on the full search reasoning main path and each difference reasoning path; S310, setting an approximation loss threshold, and constructing an inference approximation loss function using the error between the sum of the inference search loss corresponding to the inference path of each inference process and the approximation loss threshold; S311, setting a training loss threshold and a training cycle, using the sum of the inference approximation loss function and the search matching loss to construct a multi-graph joint reasoning model training loss function, when the inference search loss of each inference path in the training cycle is less than the approximation loss threshold and the multi-graph joint reasoning model training loss function is less than the training loss threshold, a trained multi-graph joint reasoning model is obtained; The full search main path in S303 corresponds to the main knowledge subgraph; the first difference search path corresponds to the knowledge subgraph one contained in the collaborative knowledge subgraph; the second difference search path corresponds to the knowledge subgraph two contained in the collaborative knowledge subgraph; the third difference search path corresponds to the knowledge subgraph three contained in the collaborative knowledge subgraph; the knowledge subgraph one, knowledge subgraph two and knowledge subgraph three contained in the collaborative knowledge subgraph are the remaining three knowledge subgraphs of the first knowledge subgraph, the second knowledge subgraph, the third knowledge subgraph and the fourth knowledge subgraph excluding the remaining three knowledge subgraphs used as the main knowledge subgraph, and the knowledge subgraph one, the knowledge subgraph two and the knowledge subgraph three correspond one-to-one to the remaining three knowledge subgraphs.

2. The AI ​​intelligent data processing method for fire protection field according to claim 1, characterized in that: The specific steps of constructing the fire protection multi-association knowledge graph include: S101, setting the fire protection layout data of buildings in different fields as the first fire protection sub-information stored in the first knowledge sub-graph, the fire protection law data as the second fire protection sub-information stored in the second knowledge sub-graph, the fire and corresponding processing information data as the third fire protection sub-information stored in the third knowledge sub-graph, and the fire protection simulation and management data as the fourth fire protection sub-information stored in the fourth knowledge sub-graph; S102, according to the first fire sub-information, the second fire sub-information, the third fire sub-information and the fourth fire sub-information, hierarchically classify each fire sub-information by using a set classification criterion combined with a classification algorithm to obtain multi-level first fire sub-information, multi-level second fire sub-information, multi-level third fire sub-information and multi-level fourth fire sub-information; S103, according to the classified fire protection sub-information corresponding to the jth level in the obtained i-th fire protection sub-information, obtain the ;in represents the entity triple set corresponding to the j-th level classification fire protection sub-information in the i-th fire protection sub-information; S104, repeating the process of S103 to obtain the entity triple set corresponding to each level in the i-th fire protection sub-information, and marking each entity triple set with a hierarchical label; S105. Obtain the knowledge sub-graph corresponding to the i-th fire protection sub-information according to the set of entity triples in the i-th fire protection sub-information annotated with hierarchical labels, and repeat the processes S103-S105 to obtain the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph and the fourth knowledge sub-graph.

3. The AI ​​intelligent data processing method for fire protection field as claimed in claim 2, characterized in that: The specific steps of constructing the fire protection multi-association knowledge graph also include: S106, according to the fire protection sub-information stored in the node corresponding to the jth level in the i-th knowledge sub-graph and the fire protection sub-information stored in the node corresponding to the jth level in the k-th knowledge sub-graph, through a similar matching algorithm, obtain the two nodes with the highest node matching degree and the lowest load in the j-th level. and ;in represents the hth node in the jth level in the i-th knowledge subgraph, represents the g-th node in the j-th level in the k-th knowledge subgraph; S107, Utilization and and the corresponding node matching degree, construct the graph connection between the i-th knowledge subgraph and the k-th knowledge subgraph at the j-th level; S108, repeat the process of S106-S107 to obtain the graph connection of the same level corresponding to the i-th knowledge sub-graph and the k-th knowledge sub-graph; S109, repeating the process of S106-S108 to obtain the corresponding graph connections between any two knowledge sub-graphs in the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph, and the fourth knowledge sub-graph; S110, construct a fire protection multi-association knowledge graph based on the first knowledge sub-graph, the second knowledge sub-graph, the third knowledge sub-graph and the fourth knowledge sub-graph and all corresponding graph connections.

4. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, an AI intelligent data processing method for the fire protection field as described in any one of claims 1 to 3 is executed.

Images