A method and system for tracing abnormality of fire-fighting equipment

Abstract

The present invention relates to a method and system for tracing the source of abnormalities of fire-fighting equipment, belonging to the field of data processing technology. The method comprises: extracting traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment; constructing a knowledge graph based on the traceability data related to the fire-fighting equipment; determining multiple event paths from the knowledge graph, each event path including at least one node corresponding to the fire-fighting equipment; calculating the failure transmission probability of each event path based on the timestamp; calculating the path priori probability of the starting point nodes in the multiple event paths based on the failure transmission probability of each event path; and determining the abnormal traceability node of the fire-fighting equipment based on the path priori probability of the starting point node. The present application realizes accurate traceability and responsibility identification of abnormalities of fire-fighting equipment through knowledge graph construction, multi-path backtracing, failure transmission probability calculation, and priori probability aggregation, saving manpower and material resources.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an abnormality tracing method and system for fire-fighting equipment. Background Art

[0002] With the continuous strengthening of fire awareness, various types of fire-fighting equipment have begun to emerge, and fire-fighting equipment has become more intelligent. The emergence of fire-fighting equipment has ensured the safety of residents to a certain extent.

[0003] The study found that problems may occur in any link of the production, installation, and use of fire-fighting equipment. For example, irregular behavior, human damage, and production defects may cause the fire-fighting equipment to fail to perform its corresponding functions, and then cause a small fire to become a large fire, resulting in irreparable losses.

[0004] However, the existing method of tracing the abnormality of fire-fighting equipment usually adopts manual investigation, such as retrieving monitoring, questioning the manufacturer, investigating the fire-fighting system log, etc. This method is time-consuming and labor-intensive, with high labor costs and low tracing efficiency. Summary of the Invention

[0005] In order to solve the above-mentioned problems in the prior art, the present invention provides a method and system for tracing the source of abnormalities in fire-fighting equipment.

[0006] In the first aspect, the present application provides a method for tracing the abnormality of fire-fighting equipment, comprising: extracting traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment; constructing a knowledge graph based on the traceability data related to the fire-fighting equipment; wherein the knowledge graph includes nodes, edges and timestamps; the types of nodes include at least two of the production subject, installation subject, maintenance subject, user subject and equipment sub-component; the types of edges include at least two of manufacturing, installation, maintenance, use and supervision; the timestamps correspond to the edges; and multiple event paths are determined from the knowledge graph, each event path includes at least A node corresponding to the fire-fighting equipment; calculating the failure transmission probability of each event path based on the timestamp; wherein the failure transmission probability represents the possibility of the event path transmitting a fault; calculating the path prior probability of the starting point node in the multiple event paths based on the failure transmission probability of each event path; wherein the path prior probability represents the prior probability value of the starting point node as the responsible entity; each node in the knowledge graph is used as a starting point node at least once, and the path prior probabilities of the same starting point node are accumulated; based on the path prior probability of the starting point node, determining the abnormal tracing node of the fire-fighting equipment.

[0007] Optionally, the abnormal tracing node of the fire-fighting equipment is determined based on the path prior probability of the starting point node, including: outputting the semantic embedding vector of each node in the knowledge graph based on a heterogeneous graph neural network; mapping the abnormality-related data of the fire-fighting equipment into a query vector via a multi-layer perceptron; calculating the similarity between the semantic embedding vector of each node and the query vector; and determining the abnormal tracing node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector, as well as the path prior probability of the starting point node.

[0008] Optionally, the method of determining the abnormal traceability node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector, and the path prior probability of the starting node, includes: constructing a credibility correction term; wherein the credibility correction term is related to at least one of the historical number of violations of the subject corresponding to the node, the equipment batch defect rate, and the warranty amount; constructing a likelihood term based on the similarity between the semantic embedding vector of each node and the query vector; determining the posterior probability of each node based on the credibility correction term, the likelihood term, and the path prior probability of the starting node; and determining the abnormal traceability node of the fire-fighting equipment based on the posterior probability of each node.

[0009] Optionally, the abnormal tracing node of the fire-fighting equipment is determined based on the similarity between the semantic embedding vector of each node and the query vector, and the path prior probability of the starting point node, including: obtaining a first weight value and a second weight value; wherein the first weight value corresponds to the similarity, and the second weight value corresponds to the path prior probability; based on the similarity between the semantic embedding vector of each node and the query vector, the first weight value, the path prior probability of the starting point node and the second weight value, the abnormal tracing node of the fire-fighting equipment is determined.

[0010] Optionally, determining multiple event paths from the knowledge graph includes: determining relevant nodes within a preset time from the abnormal moment of the fire-fighting equipment based on timestamps; and determining multiple event paths from the knowledge graph based on the relevant nodes.

[0011] Optionally, before extracting the traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment, the method further includes: obtaining historical data of each fire-fighting equipment at the production end, installation end, maintenance end and use end; the historical data at the use end includes the detection data of the fire-fighting equipment during use; the historical data of each fire-fighting equipment at the production end, installation end, maintenance end and use end are stored on the chain through the blockchain; accordingly, extracting the traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment includes: extracting the traceability data related to the fire-fighting equipment from the blockchain in response to an abnormality in the fire-fighting equipment.

[0012] Optionally, before responding to an abnormality in the fire-fighting equipment, the method further includes: obtaining detection data of the fire-fighting equipment in the previous L time steps; L is a positive integer greater than zero; inputting the detection data of the fire-fighting equipment in the previous L time steps into a bidirectional LSTM network, and outputting prediction data for the next step; based on the historical standard deviation of the fire-fighting equipment under normal working conditions, the detection data of the Lth time step and the prediction data for the next step, determining whether there is an abnormality in the fire-fighting equipment.

[0013] Optionally, the historical data of the fire-fighting equipment at the production end includes a production quality score of the fire-fighting equipment at the production end; the production quality score of the fire-fighting equipment at the production end is calculated based on different indicators of the fire-fighting equipment.

[0014] Optionally, the fire-fighting equipment is provided with an identification code; wherein, the identification code is configured to be scanned to obtain historical data of the fire-fighting equipment at the production end, installation end, maintenance end and use end.

[0015] On the second aspect, the present application provides an abnormality tracing system for fire-fighting equipment, including: an extraction module for extracting traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment; a construction module for constructing a knowledge graph based on the traceability data related to the fire-fighting equipment; wherein the knowledge graph includes nodes, edges and timestamps; the types of nodes include at least two of the production subject, installation subject, maintenance subject, user subject and equipment sub-components; the types of edges include at least two of manufacturing, installation, maintenance, use and supervision; the timestamps correspond to the edges; a path determination module for determining multiple event paths from the knowledge graph, each event path includes at least one event path related to the A node corresponding to the fire-fighting equipment; a first calculation module, used to calculate the failure transmission probability of each event path based on the timestamp; wherein, the failure transmission probability represents the possibility of the event path transmitting a fault; a second calculation module, used to calculate the path prior probability of the starting node in the multiple event paths based on the failure transmission probability of each event path; wherein, the path prior probability represents the prior probability value of the starting node as the responsible entity; each node in the knowledge graph is used as a starting node at least once, and the path prior probabilities of the same starting node are accumulated; a tracing determination module, used to determine the abnormal tracing node of the fire-fighting equipment based on the path prior probability of the starting node.

[0016] The beneficial effects of the present invention include: the present application provides a method for tracing the source of abnormalities of fire-fighting equipment, the core of which is that when it is determined that there is an abnormality in the fire-fighting equipment, the traceability data related to the fire-fighting equipment can be extracted and a knowledge graph can be constructed based on this, then the event path is extracted, and the failure transmission probability of each path is calculated based on the timestamp of the event, and finally, the path prior probability of each node is determined in combination with the starting point node, so as to screen out the abnormal traceability node. That is, the present application realizes the accurate traceability and responsibility identification of the abnormalities of fire-fighting equipment through the construction of knowledge graph, multi-path backtracing, failure transmission probability calculation and prior probability aggregation. The nodes and edges of the knowledge graph in the present application are semantically defined, and the path backtracing results are directly mapped to the responsibility chain; by extracting multiple event paths and calculating the failure transmission probability based on the timestamp, combined with the consideration of the time of event occurrence, the possibility of the fault being transmitted on the responsibility chain is quantified, and then by accumulating the failure probability of all paths with the same node as the starting point, the path prior probability is generated, and the possibility of each subject being the source of responsibility is quantified, so as to provide an efficient and accurate fire-fighting equipment abnormality traceability strategy, saving manpower and material resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of a method for tracing the source of an abnormality in fire-fighting equipment provided by an embodiment of the present invention;

[0018] Figure 2A flowchart of another method for tracing the source of abnormalities in fire-fighting equipment provided by an embodiment of the present invention;

[0019] Figure 3 A module block diagram of an abnormality tracing system for fire-fighting equipment provided by an embodiment of the present invention;

[0020] Figure 4 This is a module block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In the following description, specific details such as specific system structures and technologies are provided for illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present application with unnecessary details.

[0022] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0023] The study found that problems may occur in any link of the production, installation, and use of fire-fighting equipment. For example, irregular behavior, human damage, and production defects may cause the fire-fighting equipment to fail to perform its corresponding functions, and then cause a small fire to become a large fire, resulting in irreparable losses.

[0024] However, the existing method of tracing the abnormality of fire-fighting equipment usually adopts manual investigation, such as retrieving monitoring, questioning the manufacturer, investigating the fire-fighting system log, etc. This method is time-consuming and labor-intensive, with high labor costs and low tracing efficiency.

[0025] In view of the above problems, the present application proposes the following embodiments to solve the above technical problems.

[0026] See also Figure 1 , an embodiment of the present application provides a method for tracing the source of abnormalities in fire-fighting equipment, including: steps 101 to 106.

[0027] Step 101: In response to an abnormality in fire-fighting equipment, extract traceability data related to the fire-fighting equipment.

[0028] When the fire-fighting equipment has an abnormality, the abnormal data of the fire-fighting equipment may be detected. Alternatively, when the fire-fighting equipment has an abnormality, the abnormality may be predicted to occur soon.

[0029] Among them, traceability data related to fire-fighting equipment is pre-stored for direct extraction.

[0030] In the embodiment of the present application, the fire fighting equipment may be but is not limited to

[0031] Step 102: Build a knowledge graph based on the traceability data related to firefighting equipment.

[0032] Among them, the knowledge graph includes nodes, edges and timestamps; the types of nodes include at least two of the production entity, installation entity, maintenance entity, user entity and equipment subcomponents; the types of edges include at least two of manufacturing, installation, maintenance, use and supervision; the timestamps correspond to the edges.

[0033] Specifically, after obtaining the traceability data related to the fire-fighting equipment, data mapping is performed to form a knowledge graph corresponding to the life cycle of the fire-fighting equipment.

[0034] Among them, the knowledge graph can be expressed as G=(V, E, T), where V represents a node, and its type can be production entity, installation entity, maintenance entity, user entity, and equipment sub-component.

[0035] It should be noted that the production entity can correspond to the manufacturer or production unit of the fire-fighting equipment. The installation entity can correspond to the installation department or installation unit of the fire-fighting equipment. The maintenance entity can correspond to the maintenance department or maintenance unit of the fire-fighting equipment. The user entity corresponds to the enterprise, department, or unit that uses the fire-fighting equipment. The equipment subcomponent can be a key component of the fire-fighting equipment, or a monitoring component for the fire-fighting equipment, such as a detector, fire extinguisher valve, water pump motor, temperature and humidity sensor, etc.

[0036] E represents an edge, which can be a semantic edge, including the following types: "manufacture-of", "install-by", "maintain-by", "use-by", and "supervisory-of".

[0037] T corresponds to an edge and can specifically represent the timestamp of each event. For example, if the timestamp of the edge between the maintenance subject and the fire equipment is 20250102, it means that the maintenance subject performed a maintenance event on the fire equipment on January 2, 2025.

[0038] In one embodiment, the knowledge graph can also be expressed as G (V, E, R, T), and the knowledge graph G adds a weight term R in the aforementioned example, which can be used to characterize the reliability weight of each edge.

[0039] Step 103: Determine multiple event paths from the knowledge graph, each event path including at least one node corresponding to fire-fighting equipment.

[0040] Then, based on the constructed knowledge graph, all event paths associated with firefighting equipment are determined.

[0041] For example, path 1 may be manufacturing enterprise A—maintenance entity A—firefighting equipment.

[0042] Path 2 can be manufacturing enterprise A—maintenance entity B—user A—firefighting equipment.

[0043] Path 3 may be to use subject B - firefighting equipment.

[0044] Path 4 can be valve-firefighting equipment.

[0045] Step 104: Calculate the failure transmission probability of each event path based on the timestamp.

[0046] The failure transmission probability represents the possibility of the event path transmitting a fault.

[0047] Among them, the failure transmission probability The calculation formula can be expressed as:

[0048] ;

[0049] In the above formula, Indicates the event paths; represents the total number of edges, Indicates the sides, Indicates the The first event path sides; Represents an edge The attenuation coefficient, For the edge The time interval between the most recent exception time node and the current exception time node. Indicates the abnormal time node of fire protection equipment.

[0050] Step 105: Based on the failure transmission probability of each event path, calculate the path priori probabilities of the starting nodes in the multiple event paths.

[0051] Among them, the path prior probability represents the prior probability value of the starting node as the responsible party; each node in the knowledge graph is used as a starting node at least once, and the path prior probabilities of the same starting node are accumulated.

[0052] The path prior probability of the above starting node The formula for obtaining can be expressed as:

[0053] ;

[0054] The above formula, Indicates node, summation formula subscript Characterization Is the path The first node.

[0055] Step 106: Determine the abnormal tracing node of the fire-fighting equipment based on the path prior probability of the starting point node.

[0056] Finally, the abnormal tracing node of the firefighting equipment can be determined based on the path prior probability of the starting node. Specifically, the abnormal tracing node can be determined based on the size of the path prior probability. The larger the path prior probability, the greater the probability of being the responsible party.

[0057] In summary, the embodiment of the present application provides a method for tracing the abnormality of fire-fighting equipment. The core of the present application is that when it is determined that there is an abnormality in the fire-fighting equipment, the traceability data related to the fire-fighting equipment can be extracted and a knowledge graph can be constructed based on this. Then, the event path is extracted, and the failure transmission probability of each path is calculated based on the timestamp of the event. Finally, the path prior probability of each node is determined in combination with the starting point node to screen out the abnormal traceability node. That is, the present application realizes the accurate traceability and responsibility identification of the abnormality of fire-fighting equipment through the construction of knowledge graph, multi-path backtracking, failure transmission probability calculation and prior probability aggregation. The nodes and edges of the knowledge graph in the present application are semantically defined, and the path backtracking results are directly mapped to the responsibility chain; by extracting multiple event paths and calculating the failure transmission probability based on the timestamp, combined with the consideration of the time of event occurrence, the possibility of fault transmission in the responsibility chain is quantified, and then by accumulating the failure probability of all paths with the same node as the starting point, the path prior probability is generated, and the possibility of each subject as the source of responsibility is quantified, so as to provide an efficient and accurate fire-fighting equipment abnormality traceability strategy, saving manpower and material resources.

[0058] See also Figure 2 Optionally, the above steps determine the abnormal tracing node of the fire-fighting equipment based on the path prior probability of the starting node, including: steps 201 to 204.

[0059] Step 201: Based on the heterogeneous graph neural network, output the semantic embedding vector of each node in the knowledge graph.

[0060] That is, load the heterogeneous graph neural network and generate semantic embedding vectors of all nodes.

[0061] Step 202: Mapping the abnormality-related data of the fire-fighting equipment into a query vector via a multi-layer perceptron.

[0062] The abnormality-related data here can be detected abnormal indicators, alarm types, alarm locations, etc.

[0063] Step 203: Calculate the similarity between the semantic embedding vector of each node and the query vector.

[0064] In an embodiment of the present application, an attention mechanism may be used to calculate the similarity between the semantic embedding vector of each node and the query vector.

[0065] The similarity calculation formula can be specifically expressed as:

[0066] ;

[0067] in, Representation node The similarity between it and the query vector q. Representation node The semantic embedding vector of Representation node The semantic embedding vector of Represents a collection of nodes.

[0068] Step 204: Based on the similarity between the semantic embedding vector of each node and the query vector, and the path prior probability of the starting node, the abnormal tracing node of the fire-fighting equipment is determined.

[0069] Finally, the abnormal tracing node of the fire-fighting equipment can be determined by combining the similarity between the semantic embedding vector of each node and the query vector, as well as the path prior probability of the starting node.

[0070] In this embodiment, a heterogeneous graph neural network and a fusion analysis of abnormality-related data of fire-fighting equipment are introduced, and real-time data information is introduced to assist in responsibility judgment, thereby quantifying the degree of correlation between each node and the current abnormality-related data, thereby significantly improving the accuracy and reliability of abnormal tracing nodes.

[0071] Optionally, the above steps determine the abnormal traceability node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector, and the path prior probability of the starting node, including: constructing a credibility correction term; wherein the credibility correction term is related to at least one of the historical number of violations of the subject corresponding to the node, the equipment batch defect rate, and the warranty amount; constructing a likelihood term based on the similarity between the semantic embedding vector of each node and the query vector; determining the posterior probability of each node based on the credibility correction term, the likelihood term, and the path prior probability of the starting node; and determining the abnormal traceability node of the fire-fighting equipment based on the posterior probability of each node.

[0072] The calculation formula of the credibility correction term in one embodiment can be expressed as:

[0073] ;

[0074] in, represents the Sigmoid function, Indicates that for the node Credibility modifier of ; Representation node The number of historical violations, Representation node Equipment batch defect rate, Representation node warranty amount. , , , Obtained through offline training of the log-likelihood of historical data.

[0075] The formula for the posterior probability of the final responsible party can be expressed as:

[0076] ;

[0077] In the above formula, Representation node The corresponding posterior probability; Representation node Corresponding likelihood term; Representation node Prior probability of the path; Representation node Corresponding likelihood term; Representation node Prior probability of the path; Represents a collection of nodes.

[0078] This implementation further introduces a credibility modifier and posterior probability calculation mechanism. By constructing a credibility modifier related to the number of historical violations, equipment batch defect rate, and warranty amount, the historical behavioral reputation and objective risk indicators of the responsible party are quantified. The posterior probability is calculated by layering and integrating three types of evidence: real-time observation matching (likelihood term), historical path responsibility (prior probability), and subject credibility (credibility modifier), further improving the accuracy and reliability of subsequent abnormal tracing node determination.

[0079] Specifically, the likelihood term (similarity) reflects the direct correlation between the current accident characteristics and the node (e.g., a valve closure alarm matches a maintenance record). The prior probability reflects the node's exposure to liability in the historical path. The credibility correction term balances the credibility of the subject's historical behavior (e.g., if a manufacturer has a high path probability but good historical quality, a correction is made).

[0080] In summary, the embodiments of the present application establish a dynamic knowledge graph by performing semantic extraction and temporal embedding on multidimensional data related to fire-fighting equipment, and immediately activate the above-mentioned responsibility inference algorithm strategy after an accident occurs or when the risk of an accident arises, thereby reducing labor costs and improving accountability efficiency.

[0081] Optionally, the above steps determine the abnormal tracing node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector, and the path prior probability of the starting node, including: obtaining a first weight value and a second weight value; wherein the first weight value corresponds to the similarity, and the second weight value corresponds to the path prior probability; based on the similarity between the semantic embedding vector of each node and the query vector, the first weight value, the path prior probability of the starting node and the second weight value, determine the abnormal tracing node of the fire-fighting equipment.

[0082] Specifically, the weight ratio of real-time detection evidence and historical path responsibility can be flexibly adjusted through the first weight value (similarity weight) and the second weight value (prior probability weight).

[0083] For example, for sudden failures (such as human operational errors), the first weight value (similarity-dominated) can be increased to quickly match real-time abnormal features.

[0084] For cumulative failures (such as equipment aging), the second weight value can be increased (dominated by prior probability) to strengthen historical path analysis.

[0085] It can be seen that by introducing a dynamic weight distribution mechanism of similarity and prior probability, the responsibility determination can flexibly adjust the contribution ratio of real-time data and historical evidence according to scenario requirements, taking into account the agility, configurability and anti-interference ability of traceability.

[0086] Optionally, multiple event paths are determined from the knowledge graph, including: based on the timestamp, determining relevant nodes within a preset time from the abnormal moment of the fire-fighting equipment; based on the relevant nodes, determining multiple event paths from the knowledge graph.

[0087] The above preset time can be set according to needs, such as three months, five months, one year, or two years.

[0088] That is, in this embodiment, a time window screening mechanism is introduced to retain only the nodes and paths close to the abnormal moment, eliminating the interference of premature historical data or obsolete data, and ensuring the timeliness and rationality of the responsibility source tracing.

[0089] Optionally, in response to an abnormality in fire-fighting equipment, before extracting traceability data related to the fire-fighting equipment, the method also includes: obtaining historical data of each fire-fighting equipment at the production end, installation end, maintenance end, and use end; the historical data at the use end includes detection data of the fire-fighting equipment during use; and storing the historical data of each fire-fighting equipment at the production end, installation end, maintenance end, and use end on the blockchain.

[0090] Correspondingly, the above steps extract the traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment, including: extracting the traceability data related to the fire-fighting equipment from the blockchain in response to an abnormality in the fire-fighting equipment.

[0091] Optionally, the fire-fighting equipment is provided with an identification code; wherein, the identification code is configured to be scanned to obtain historical data of the fire-fighting equipment at the production end, installation end, maintenance end and use end.

[0092] Specifically, the identification code can be a unique, resolvable, and verifiable digital identity assigned to the device the moment it leaves the factory, and ensure that the identity is always single-mapped, cannot be forged, and cannot be lost throughout the entire process of production, installation, use, maintenance, and even scrapping.

[0093] The identification code can include: region code, company code, equipment type code, batch code, and timestamp. The identification code can also be attached to firefighting equipment in the form of a QR code or NFC, used to link information throughout the four stages of "production, installation, use, and maintenance."

[0094] The following examples illustrate the historical data collection at the production, installation, maintenance, and usage ends of each link.

[0095] Production side: By combining machine vision with sensors, real-time, online quality inspections can be performed on key processes of firefighting equipment, and the inspection results can be quantified into a unified production quality score. :

[0096] ;

[0097] in, Indicates the total number of quality indicators included in this test. Indicates the The standardized score of the indicator, value Specifically, if the indicator is positive (the larger the value, the better, such as coating thickness uniformity), it can be expressed by the following formula:

[0098] ;

[0099] If the indicator is inverse (the smaller the value, the better, such as defect rate), it can be expressed by the following formula:

[0100] ;

[0101] in, is the measured value, The minimum value is given by industry standards or enterprise internal control standards. The maximum value is given by industry standards or enterprise internal control standards.

[0102] Indicates the The weight coefficient of each indicator is set according to its impact on the risk of fire equipment failure, and .

[0103] The present application embodiment provides the following practical workflow:

[0104] Step 1: The industrial camera and 3D structured light sensor are triggered synchronously; process parameters (temperature, speed, gas flow) are entered into the same data frame;

[0105] Step 2: Run the YOLOv9+Unet model to output structured results such as defect masks and size differences.

[0106] Step 3: Instant Computing at the Edge and , blockchain for storage.

[0107] That is, the historical data of the fire-fighting equipment at the production end may include the production quality score of the fire-fighting equipment at the production end; the production quality score of the fire-fighting equipment at the production end is calculated based on different indicators of the fire-fighting equipment.

[0108] Installation end: After the operator scans the identification code, the terminal automatically obtains the product installation method and guides step by step on the AR interface: After the user scans the QR code on the fire-fighting equipment, select "Installation Information Entry"; the terminal module automatically obtains the installation method corresponding to the identification code from the production supervision module; the user installs the equipment according to the installation method displayed on the terminal module and manually gives the installation score.

[0109] User end: The embedded gateway collects pressure, temperature, humidity, current, voltage, vibration and other data every 60 seconds (can be set according to needs) .

[0110] Maintenance end: Maintenance personnel can also scan the identification code to enter "Information Entry" and automatically pull the product maintenance method: scan the code to enter the maintenance interface; the terminal displays the maintenance points; maintenance personnel maintain and upload images and parameters according to the process; after the maintenance is completed, the platform generates and stores maintenance information.

[0111] Optionally, before responding to an abnormality in the fire-fighting equipment, the method further includes: obtaining detection data of the fire-fighting equipment in the previous L time steps; L is a positive integer greater than zero; inputting the detection data of the fire-fighting equipment in the previous L time steps into a bidirectional LSTM network, and outputting prediction data for the next step; based on the historical standard deviation of the fire-fighting equipment under normal working conditions, the detection data of the Lth time step and the prediction data for the next step, determining whether there is an abnormality in the fire-fighting equipment.

[0112] Input the detection data of the last L time steps into the bidirectional LSTM network and output the predicted data for the next step. The expression can be:

[0113] ;

[0114] Represents the predicted data for the next step, represents a bidirectional LSTM network, express The detection data corresponding to the moment, Indicates the detection data corresponding to the time.

[0115] By comparing indicator fluctuations to determine whether there is abnormal risk:

[0116] ;

[0117] Indicates indicator fluctuations; is the historical standard deviation of the equipment under normal working conditions. <0.70 is considered "needing attention", <0.50 was considered “abnormal”.

[0118] That is, the technical solution provided by this embodiment can predict anomalies through a bidirectional LSTM network, identify potential fault risks in advance through time series data analysis, achieve early warning of fire-fighting equipment anomalies, provide richer fault development process data for subsequent tracing, and form a full-chain technical closed loop of "prediction-warning-tracing".

[0119] See also Figure 3 Based on the same inventive concept, the present application provides a fire-fighting equipment abnormality tracing system 300, comprising:

[0120] An extraction module 301 is configured to extract traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment;

[0121] A construction module 302 is configured to construct a knowledge graph based on the traceability data related to the firefighting equipment; wherein the knowledge graph includes nodes, edges, and timestamps; the node types include at least two of the production entity, installation entity, maintenance entity, user entity, and equipment subcomponent; the edge types include at least two of the manufacturing, installation, maintenance, use, and supervision; and the timestamps correspond to the edges;

[0122] A path determination module 303 is configured to determine a plurality of event paths from the knowledge graph, each event path including at least one node corresponding to the fire-fighting equipment;

[0123] A first calculation module 304 is configured to calculate a failure transmission probability of each event path based on the timestamp; wherein the failure transmission probability represents the possibility of the event path transmitting a fault;

[0124] A second calculation module 305 is configured to calculate a path prior probability of a starting point node in the plurality of event paths based on the failure transmission probability of each event path; wherein the path prior probability represents a prior probability value of the starting point node as a responsible entity; each node in the knowledge graph is used as a starting point node at least once, and the path prior probabilities of the same starting point node are accumulated;

[0125] The source tracing determination module 306 is configured to determine the abnormal source node of the fire-fighting equipment based on the path prior probability of the starting point node.

[0126] See also Figure 4 Based on the same inventive concept, the present invention provides a module frame of an electronic device 400 that applies the above method. The electronic device 400 includes: at least one processor 401 ( Figure 4 Only one is shown), a memory 402, a computer program 403 stored in the memory 402 and executable on at least one processor 401, wherein the processor 401 implements the steps of the method in any of the foregoing embodiments when executing the computer program 403.

[0127] The electronic device 400 can be a server, a personal computer, a notebook computer, etc.

[0128] Those skilled in the art will understand that Figure 4 This is merely an example of the electronic device 400 and does not constitute a limitation on the electronic device 400 . The electronic device 400 may include more or fewer components than shown in the figure, or may combine certain components, or may include different components.

[0129] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or any conventional processor.

[0130] In some embodiments, the memory 402 may be an internal storage unit of the electronic device 400, such as a hard disk or memory of the electronic device 400. In other embodiments, the memory 402 may also be an external storage device of the electronic device 400, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the electronic device 400. Furthermore, the memory 402 may include both an internal storage unit of the electronic device 400 and an external storage device.

[0131] It should be noted that the above-mentioned systems, equipment, etc. are based on the same concept as the method embodiments of this application. The modules designed for the systems, the steps performed by the equipment, and the technical effects brought about can all be found in the method embodiment section and will not be repeated here.

[0132] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0133] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned various method embodiments can be implemented.

[0134] An embodiment of the present application provides a computer program product. When the computer program product is run on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned various method embodiments when executing the computer program product.

[0135] If the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process steps in the above-mentioned method embodiments by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to the camera / electronic device, a recording medium, computer memory, read-only memory (ROM), random access memory (RAM), an electrical carrier signal, a telecommunications signal, and a software distribution medium. Examples include a USB flash drive, a removable hard drive, a magnetic disk, or an optical disk.

[0136] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0137] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0138] In the embodiments provided in this application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0139] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0140] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for tracing the source of abnormalities in fire-fighting equipment, characterized in that: include: In response to an abnormality in fire-fighting equipment, extracting traceability data related to the fire-fighting equipment; A knowledge graph is constructed based on the traceability data related to the fire-fighting equipment; wherein the knowledge graph includes nodes, edges, and timestamps; the types of nodes include at least two of the production entity, installation entity, maintenance entity, user entity, and equipment subcomponent; the types of edges include at least two of the manufacturing, installation, maintenance, use, and supervision; and the timestamps correspond to the edges; Determining a plurality of event paths from the knowledge graph, each event path including at least one node corresponding to the fire-fighting equipment; Calculate the failure transmission probability of each event path based on the timestamp; wherein the failure transmission probability represents the possibility of the event path transmitting a fault; Based on the failure transmission probability of each event path, calculate the path prior probability of the starting node in the multiple event paths; wherein the path prior probability represents the prior probability value of the starting node as the responsible party; each node in the knowledge graph is used as a starting node at least once, and the path prior probabilities of the same starting node are accumulated; Determining the abnormal tracing node of the fire-fighting equipment based on the path prior probability of the starting point node; The method of determining the abnormal tracing node of the fire-fighting equipment based on the path prior probability of the starting point node includes: outputting the semantic embedding vector of each node in the knowledge graph based on a heterogeneous graph neural network; mapping the abnormality-related data of the fire-fighting equipment into a query vector via a multi-layer perceptron; calculating the similarity between the semantic embedding vector of each node and the query vector; and determining the abnormal tracing node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector, as well as the path prior probability of the starting point node.

2. The abnormality tracing method of fire-fighting equipment according to claim 1, characterized in that: The determining of the abnormal tracing node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector and the path prior probability of the starting point node includes: Constructing a credibility correction item; wherein the credibility correction item is related to at least one of the number of historical violations of the subject corresponding to the node, the equipment batch defect rate, and the warranty amount; Constructing a likelihood term based on the similarity between the semantic embedding vector of each node and the query vector; Determining the posterior probability of each node based on the credibility correction term, the likelihood term, and the path prior probability of the starting node; Based on the posterior probability of each node, the abnormal tracing node of the fire-fighting equipment is determined.

3. The abnormality tracing method of fire-fighting equipment according to claim 1, characterized in that: The determining of the abnormal tracing node of the fire-fighting equipment based on the similarity between the semantic embedding vector of each node and the query vector and the path prior probability of the starting point node includes: Obtaining a first weight value and a second weight value; wherein the first weight value corresponds to the similarity, and the second weight value corresponds to the path prior probability; Based on the similarity between the semantic embedding vector of each node and the query vector, the first weight value, the path prior probability of the starting point node and the second weight value, the abnormal tracing node of the fire-fighting equipment is determined.

4. The abnormality tracing method of fire-fighting equipment according to claim 1, characterized in that: Determining multiple event paths from the knowledge graph includes: Based on the timestamp, determine the relevant nodes within a preset time from the abnormal moment of the fire-fighting equipment; Based on the relevant nodes, multiple event paths are determined from the knowledge graph.

5. The abnormality tracing method of fire-fighting equipment according to claim 1, characterized in that: Before extracting traceability data related to the fire-fighting equipment in response to the fire-fighting equipment being abnormal, the method further includes: Obtain historical data of each fire-fighting equipment at the production, installation, maintenance, and user ends; the historical data at the user end includes the detection data of the fire-fighting equipment during use; The historical data of each fire-fighting equipment at the production, installation, maintenance and use ends are stored on the blockchain; Accordingly, in response to the fire-fighting equipment being abnormal, extracting traceability data related to the fire-fighting equipment includes: In response to an abnormality in fire-fighting equipment, traceability data related to the fire-fighting equipment is extracted from the blockchain.

6. The abnormality tracing method for fire-fighting equipment according to claim 5, characterized in that: Before responding to the abnormality of the fire-fighting equipment, the method further includes: Obtain detection data of the fire-fighting equipment in the previous L time steps; L is a positive integer greater than zero; Input the detection data of the fire-fighting equipment in the first L time steps into the bidirectional LSTM network and output the prediction data for the next step; Based on the historical standard deviation of the fire-fighting equipment under normal working conditions, the detection data of the Lth time step and the predicted data of the next step, it is determined whether the fire-fighting equipment has an abnormality.

7. The method for tracing the source of abnormality of fire-fighting equipment according to claim 5, characterized in that: Historical data of the fire-fighting equipment at the production end, including the production quality score of the fire-fighting equipment at the production end; The production quality score of the fire-fighting equipment at the production end is calculated based on different indicators of the fire-fighting equipment.

8. The method for tracing the source of abnormality of fire-fighting equipment according to claim 5, characterized in that: The fire-fighting equipment is provided with an identification code; The identification code is configured to be scanned to obtain historical data of the fire-fighting equipment at the production end, installation end, maintenance end and use end.

9. An abnormality tracing system for fire-fighting equipment, characterized in that: include: An extraction module, configured to extract traceability data related to the fire-fighting equipment in response to an abnormality in the fire-fighting equipment; A construction module is configured to construct a knowledge graph based on the traceability data related to the firefighting equipment; wherein the knowledge graph includes nodes, edges, and timestamps; the node types include at least two of the production entity, installation entity, maintenance entity, user entity, and equipment subassembly; the edge types include at least two of the manufacturing, installation, maintenance, use, and supervision; and the timestamps correspond to the edges; a path determination module, configured to determine a plurality of event paths from the knowledge graph, each event path including at least one node corresponding to the fire-fighting equipment; A first calculation module is configured to calculate a failure transmission probability of each event path based on a timestamp; wherein the failure transmission probability represents a possibility of a failure being transmitted by the event path; A second calculation module is configured to calculate a path prior probability of a starting point node in the plurality of event paths based on the failure transmission probability of each event path; wherein the path prior probability represents a prior probability value of the starting point node as a responsible entity; each node in the knowledge graph is used as a starting point node at least once, and the path prior probabilities of the same starting point node are accumulated; The tracing determination module is used to determine the abnormal tracing node of the fire-fighting equipment based on the path prior probability of the starting point node; specifically, it is used to output the semantic embedding vector of each node in the knowledge graph based on the heterogeneous graph neural network; map the abnormality-related data of the fire-fighting equipment into a query vector through a multi-layer perceptron; calculate the similarity between the semantic embedding vector of each node and the query vector; based on the similarity between the semantic embedding vector of each node and the query vector, and the path prior probability of the starting point node, determine the abnormal tracing node of the fire-fighting equipment.

Images