Gas safety knowledge accurate pushing method and system based on user feature analysis

Abstract

The invention relates to the technical field of fuel gas safety monitoring, in particular to a fuel gas safety knowledge accurate pushing method and system based on user feature analysis. The method comprises the steps that gas data and environment data are collected in real time, real-time data streams of time-space alignment are generated in combination with user attributes, a dynamic time warping algorithm is adopted to divide historical data streams into user gas use scenes, and a first model is trained based on the user gas use scenes and the historical data streams; importing historical graph node data and historical reference data corresponding to the historical security event, constructing a risk knowledge graph, and performing embedded learning on the second model; inputting the real-time data stream into the first model to obtain dynamic reference data, and obtaining a simulation result based on the difference vector of the real-time data stream and the dynamic reference data and the second model; and generating risk push control information according to the simulation result, and pushing risk prompt information to the user. According to the invention, the problem of insufficient pushing precision of the gas safety information is solved, and the pushing precision is improved.

Description

Technical Field

[0001] The present application relates to the field of gas safety monitoring technology, and in particular to a method and system for accurately pushing gas safety knowledge based on user feature analysis. Background Art

[0002] In the prior art, gas leak alarms and gas meters usually operate independently, with relatively simple monitoring functions, mainly used to detect whether the gas concentration exceeds the standard to trigger an alarm, and there are limitations in the overall safety monitoring of gas use. The monitoring methods of the prior art are often unable to comprehensively and accurately analyze the user's gas usage scenarios and potential risks, and it is difficult to provide personalized gas safety knowledge push based on user characteristics. . Similar prior art includes a Chinese patent with publication number CN116859019A, which proposes a safety detection method, system, device and storage medium for residential gas. The method includes: obtaining the structure of the house to be tested; the house to be tested is used to characterize a house that uses residential gas; based on the structure of the house to be tested, determining the trajectory to be tested; scanning and detecting the trajectory to be tested based on the flight equipment to obtain detection data; the detection data includes video data and concentration data; the concentration data is used to characterize the concentration of harmful gases in the measured environment; through the video data and the concentration data, the house to be tested is detected for safety hazards to obtain detection results. The embodiments of the present invention use the structure of the house to be tested to determine a test trajectory; and automatically detect the house to be tested using a flying device. This allows for safe detection of residential gas use in any situation, improving detection accuracy and finding widespread application in the field of gas technology. Similar prior art includes U.S. Patent Publication No. US20230419811A1, which provides an alarm-based intelligent gas safety risk prevention and control method and Internet of Things system, including: obtaining gas monitoring data according to a data acquisition instruction to determine whether a gas leak has occurred; in response to determining a gas leak, generating a control instruction to control fan operation according to a fan operation strategy, and obtaining gas monitoring data under the fan operation strategy, which is used to remove the leaked gas and assist in determining the type of gas leak; generating a notification instruction based on the gas monitoring data and the operating strategy of the indicator light, and controlling the indicator light to issue an alarm notification according to the notification instruction. Although both of these patents address the issue of gas safety alarms, they struggle to accurately identify different gas usage scenarios, fail to effectively leverage historical maintenance records and safety event data for risk prediction, and lack adaptability to emerging gas usage scenarios, preventing timely updates to the risk assessment model. In addition, existing technologies also have shortcomings in the accuracy of risk warnings and the personalization of push methods, and are unable to optimize the push strategy of risk warning information based on different user characteristics and historical feedback records. Summary of the Invention

[0003] This application provides a method and system for accurately pushing gas safety knowledge based on user feature analysis. Through core technologies such as multimodal data integration, dynamic scene division, knowledge graph embedding, and closed-loop feedback optimization, the system achieves accurate monitoring of gas usage and risk warning. Its core advantages lie in dynamic benchmark adaptation, scene adaptive expansion, and personalized alarms, which significantly reduce false alarm rates and improve gas safety management efficiency. The method includes: collecting gas data and environmental data in real time, and generating a spatiotemporally aligned real-time data stream in combination with user types; Using a dynamic time warping (DTW) algorithm to divide the historical data stream into user gas usage scenarios, and training a first model based on the user gas usage scenarios and the corresponding historical data stream; Extract graph nodes based on historical maintenance records, import historical graph node data and historical benchmark data corresponding to historical security events, build a risk knowledge graph, and embed learning of the second model based on the risk knowledge graph; inputting the real-time data stream into the first model to obtain dynamic benchmark data and gas usage scenarios, and obtaining simulation results based on the real-time data stream, the dynamic benchmark data, the gas usage scenarios, and the second model; A push unit is used to determine whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector, and generate risk push control information based on the simulation result.

[0004] As a preferred technical solution of the present invention, the training of the first model includes: aligning historical gas data, historical environmental data and historical user attributes in the historical multimodal data stream according to the collection time to obtain a historical time series data stream, drawing a gas flow curve based on the gas data in the historical time series data stream of different dates, obtaining multiple groups of curve segments in any two flow curves whose similarity is greater than a set threshold through a dynamic time warping (DWT) algorithm, generating label data corresponding to the curve segment in each cluster cluster based on each group of the curve segments, the corresponding environmental data and user attributes through a clustering algorithm, and using the time series data segments corresponding to the curve segments in the time series data stream and the label data corresponding to the curve segments as training data to train the first model.

[0005] As a preferred technical solution of the present invention, the construction of the risk knowledge graph and the training of the second model include: Graph entities are extracted from the historical maintenance records through semantic analysis, and the graph entities include at least gas entities, environment entities, user entities, scenario entities, benchmark entities, safety event entities and safety factor entities. The graph entities are used as graph nodes, and the correlation between the graph nodes is obtained through semantic analysis of the historical maintenance records. Edge connections are performed based on the correlation to construct the risk knowledge graph. The historical graph node data corresponding to historical safety events in the historical maintenance records are imported into the risk knowledge graph. The second model is embedded and learned based on the historical graph node data, historical dynamic benchmark data, the historical safety events and the causal factors leading to the historical safety events in the risk knowledge graph.

[0006] As a preferred technical solution of the present invention, the acquisition of the simulation results includes: The real-time data stream is input into the first model to obtain the gas usage scenario and the corresponding dynamic benchmark data, the real-time data stream, the dynamic benchmark data and the gas usage scenario are updated to the corresponding node of the risk knowledge graph, and the difference vector between the gas data in the real-time data stream and the dynamic benchmark data is calculated. When the length of the difference vector is less than the set value, the gas usage is normal; when the length of the difference vector is greater than or equal to the set value, the second model simulates the difference vector based on the real-time data stream, the gas usage scenario and the dynamic benchmark data in the risk knowledge graph to obtain the simulation result and the weight of the target graph node corresponding to the simulation result, and displays each target graph node in a different color according to the weight size, wherein the larger the weight, the more eye-catching the color of the corresponding target graph node, and the target graph node is a graph node whose weight value is greater than the set threshold.

[0007] As a preferred technical solution of the present invention, judging whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector includes: When the length of the difference vector between the gas data and the dynamic benchmark data in the real-time data stream is greater than or equal to a set value, and the simulation result corresponding to the difference vector cannot be obtained based on the risk knowledge graph and the second model, the real-time data stream corresponding to the user is collected in real time, and the real-time data stream is mapped to the time coordinate axis, and the target data stream corresponding to the new gas usage scenario is extracted based on the dynamic time warping DWT algorithm, and the first model is trained based on the target data stream and the corresponding new gas usage scenario; otherwise, the gas usage scenario is not a new gas usage scenario; Risk simulation is also performed on the basis of the dynamic benchmark data corresponding to the new gas usage scenario to obtain multiple groups of risk events and corresponding risk data, and the data are added to the corresponding graph node positions in the risk knowledge graph. The second model is trained based on the multiple groups of risk events and the corresponding risk data.

[0008] As a preferred technical solution of the present invention, generating risk information push control information according to the simulation results includes: Based on the weight distribution of the graph nodes in the corresponding risk knowledge graph in the second model corresponding to the simulation results, the risk level of the risk warning information is determined, and according to user attributes and historical feedback records, the push method of the risk warning information is determined, and the weight of the corresponding graph node in the second model is adjusted according to the user's feedback on the pushed risk warning information.

[0009] As a preferred technical solution of the present invention, in the heating gas usage scenario, a temperature-flow compensation curve is generated through historical indoor and outdoor temperature data and gas flow data, and the dynamic benchmark data is corrected based on the temperature-flow compensation curve; when the risk knowledge graph detects a change in user attributes, the first model training cycle is reset and real-time data stream is collected, and the first model is timely trained based on the real-time data stream.

[0010] As a preferred technical solution of the present invention, the gas data includes gas flow, gas concentration data and gas pressure data, and the environmental data includes environmental temperature, humidity and ventilation status data.

[0011] The present invention also provides a gas safety knowledge accurate push system based on user feature analysis, which is used to implement the above method. The system includes: The collection unit is used to collect gas data and environmental data in real time and generate a real-time data stream aligned in time and space based on user types; A training unit, configured to divide the historical data stream into user gas usage scenarios using a dynamic time warping (DTW) algorithm, and train a first model based on the user gas usage scenarios and the corresponding historical data stream; A construction unit is used to extract graph nodes based on historical maintenance records, import historical graph node data and historical benchmark data corresponding to historical security events, construct a risk knowledge graph, and embed learning of the second model based on the risk knowledge graph; a simulation unit, configured to input the real-time data stream into the first model, obtain dynamic benchmark data and a gas usage scenario, and obtain a simulation result based on the real-time data stream, the dynamic benchmark data, the gas usage scenario, and the second model; A push unit is used to determine whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector, and generate risk push control information based on the simulation result.

[0012] The present invention also provides a computer-readable storage medium having instructions stored thereon, and the above-mentioned method is implemented when the instructions are executed by a processor.

[0013] Effect The present invention generates a real-time data stream by collecting gas and environmental data in real time and combining it with user attributes, divides the user's gas usage scenarios into two categories and trains a first model using a dynamic time warping algorithm, and simultaneously constructs a risk knowledge graph for embedded learning, thereby achieving accurate identification and risk prediction of gas usage scenarios. In real-time monitoring, the real-time data stream is input into the model to obtain the user's gas usage scenario and the corresponding dynamic benchmark data. The difference vector between the real-time data and the dynamic benchmark data is calculated, and the risk knowledge graph is used to simulate the difference vector based on the second model to obtain the simulation result. In addition, according to the difference vector and the simulation result, safety events are discovered in a timely manner and it is determined whether it is a new gas usage scenario. If the scenario is not new, risk control information is generated and a prompt is sent, effectively improving the timeliness and accuracy of gas safety monitoring. If it is a new gas usage scenario, new gas usage scenario data is promptly collected and the first model is trained in a timely manner to improve the recognition ability of new gas usage scenarios. After identifying a safety incident through the above simulation results, risk prompts are accurately sent based on different user characteristics, improving users' understanding and handling of gas risks, reducing false alarm rates, and enhancing the proactive and reliable management of gas safety. In heating scenarios, dynamic baseline data is corrected using a temperature-flow compensation curve to reduce false alarms caused by ambient temperature fluctuations. When a change in user attributes is detected, the model training cycle is reset and training is carried out in a timely manner, allowing the model to quickly adapt to the new gas usage pattern and ensure the continued accuracy of risk identification. Overall, this patent comprehensively improves the effectiveness of gas safety monitoring and management from multiple aspects, including data collection, scenario recognition, risk prediction, and information push, providing gas users with more accurate and efficient safety protection services and effectively preventing gas accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0015] Figure 1 This is a flowchart of the method for accurately pushing gas safety knowledge based on user feature analysis in this application; Figure 2 This is a flow chart of the first model training method of this application; Figure 3 Flowchart of the method for obtaining simulation results; Figure 4 This is the structural diagram of the gas safety knowledge accurate push system based on user feature analysis for this application. DETAILED DESCRIPTION

[0016] The embodiments of the present application provide a method and system for accurately pushing gas safety knowledge based on user feature analysis. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0017] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 As shown, an embodiment of the method for accurately pushing gas safety knowledge based on user feature analysis in the embodiment of the present application includes: Step S1: Generate a spatiotemporally aligned multimodal data stream by collecting gas data and environmental data in real time and combining them with user attributes; Specifically, gas flow, concentration, and pressure data are obtained in real time through gas meters, gas leak alarms, and other equipment. Combined with environmental data such as temperature, humidity, and ventilation status, user attributes, such as household users and elderly people living alone, are associated with the collected data. Multi-source data are synchronized by timestamp and spatial location to form a unified multimodal data stream, integrating multi-dimensional information on gas, environment, and users to provide complete input for scene recognition, avoid misjudgment caused by data dislocation, and improve the accuracy of subsequent analysis.

[0018] Step S2: using a dynamic time warping (DTW) algorithm to divide the historical data stream into user gas usage scenarios, and training a first model based on the user gas usage scenarios and the corresponding historical data streams; Specifically, historical gas flow curves are analyzed to identify similar time segments, such as peak gas consumption at breakfast and dinner. Environmental data and user attributes, such as the number of family members, are combined to divide gas usage scenarios, such as cooking and heating, through clustering algorithms. Time series models such as LSTM are used to train the first model with scene labels and time series data, and output gas usage scenarios and dynamic benchmark data. The above technical solution solves the time series alignment problem through the DTW algorithm and improves the scene classification accuracy. The first model outputs an adaptive benchmark value based on the scene to reduce false alarms caused by environmental interference.

[0019] Step S3: extracting graph nodes based on historical maintenance records, importing historical graph node data and historical benchmark data corresponding to historical security events, constructing a risk knowledge graph, and embedding learning of the second model based on the risk knowledge graph; Specifically, by extracting entities from historical maintenance records, establishing graph nodes, and through semantic analysis, associating nodes, such as "gas leakage" and "valve failure", a risk knowledge graph is formed. The second model is embedded and learned based on historical graph node data, historical dynamic benchmark data, historical safety events and the causes of the above historical safety events. The above technical solution intuitively displays the relationship between risk factors through the risk knowledge graph, assists in quickly locating the root cause of the problem, and improves the model's reasoning ability for complex risk scenarios through embedded learning combined with graph relationships.

[0020] Step S4: inputting the real-time data stream into the first model, obtaining dynamic benchmark data and gas usage scenarios, and obtaining simulation results based on the real-time data stream, the dynamic benchmark data, the gas usage scenarios, and the second model; Specifically, by inputting the above-mentioned real-time data stream into the above-mentioned first model, the user's current gas usage scenario and the gas data during normal gas usage under the current gas usage scenario, namely the above-mentioned dynamic benchmark data, are obtained, and then the difference vector between the gas data in the real-time data stream and the dynamic benchmark data is calculated. According to the length of the above-mentioned difference vector, namely the Euclidean distance, it is judged whether the gas usage status of the above-mentioned user is normal. When the length of the above-mentioned difference vector is less than the set value, the gas usage is normal. When the above-mentioned difference vector is greater than or equal to the above-mentioned set value, the second model simulates the above-mentioned difference vector based on the real-time data stream, the current gas usage scenario, and the above-mentioned dynamic benchmark data in the above-mentioned risk knowledge graph and obtains the simulation results and the weights of the target graph nodes corresponding to the above-mentioned simulation results. Through the above-mentioned technical solution, real-time risk monitoring of the gas usage process can be realized, potential safety hazards can be warned in advance, and a basis for taking corresponding measures can be provided, effectively improving the safety of gas use.

[0021] Step S5: a push unit, configured to determine whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector, and generate risk push control information based on the simulation result.

[0022] Specifically, based on the simulation results output by the second model and the difference vector between the real-time data stream and the dynamic benchmark data, it is determined whether the current gas usage scenario is different from the previously known gas usage scenario. If it is determined that it is not a new gas usage scenario, the simulation results are further analyzed, and risk push control information is generated based on the risk information contained in the results. The method, content and level of risk warnings are determined, and the risk warning information is sent to the user through appropriate methods, such as SMS, APP message push, etc. Through the above technical solution, new gas usage scenarios can be identified in a timely and accurate manner to avoid misjudgment and false alarm of new scenarios. At the same time, for abnormal situations under known gas usage scenarios, risk warning information can be quickly generated and pushed, so that users can timely understand the potential risks in the process of gas use and take corresponding measures to deal with them, thereby effectively reducing the probability of gas accidents and protecting the lives and property of users.

[0023] Furthermore, if Figure 2 As shown, the training of the first model includes: The historical gas data, historical environmental data, and historical user attributes in the historical multimodal data are aligned according to the collection time to obtain a historical time series data stream. A gas flow curve is drawn based on the gas data of different dates in the historical time series data stream. A dynamic time warping (DWT) algorithm is used to obtain multiple groups of curve segments with a similarity greater than a set threshold in any two of the flow curves. A clustering algorithm is used based on each group of the curve segments, the corresponding environmental data, and the user attributes to generate label data corresponding to the curve segment in each cluster cluster. The time series data segments corresponding to the curve segments in the time series data stream and the label data corresponding to the curve segments are used as training data to train the first model.

[0024] Specifically, the above-mentioned historical time series data is historical data under normal user use, the above-mentioned gas data includes gas flow data and pressure data collected by the gas meter and gas concentration data collected by the gas leak alarm, the above-mentioned environmental data includes air temperature and humidity data and ventilation status data, and the above-mentioned user attributes include young families or elderly people living alone. A flow curve is drawn according to the gas flow in the above-mentioned gas data on different dates in the historical time series data stream, and multiple groups of curve segments with a similarity greater than a set threshold in any two of the above-mentioned flow curves are obtained through the above-mentioned dynamic time warping DWT algorithm, wherein the change trend of the gas flow in each group of curve segments is the same, that is, different curve segment groupings reflect different application scenarios of users, and also according to the above-mentioned environmental data and user corresponding to the curve segment in each group. The attributes are clustered unsupervisedly, and the flow data characteristics are analyzed, such as gas usage time, gas usage duration and gas usage flow, to generate label data corresponding to each cluster cluster. The above-mentioned label data at least includes breakfast, lunch and dinner gas usage scenarios, bathing gas usage scenarios, heating gas usage scenarios or rest gas shutdown scenarios, and the time series data segments and corresponding label data in the time series data stream corresponding to each of the above curve segments are used as training data, and the above-mentioned first model is trained based on the above-mentioned training data. The above-mentioned first model is an LSTM model, wherein the above-mentioned time series data segments are the same as the time periods corresponding to the above-mentioned curve segments. Through the above-mentioned technical solution, an accurate gas usage scenario recognition model, namely the above-mentioned first model, can be obtained, and the above-mentioned first model can also be used to obtain the dynamic control reference value of the gas data corresponding to each of the above-mentioned gas usage scenarios.

[0025] Furthermore, the construction of the risk knowledge graph and the training of the second model include: Graph entities are extracted from the historical maintenance records through semantic analysis, and the graph entities include at least gas entities, environment entities, user entities, scenario entities, benchmark entities, safety event entities and safety factor entities. The graph entities are used as graph nodes, and the correlation between the graph nodes is obtained through semantic analysis of the historical maintenance records. Edge connections are performed based on the correlation to construct the risk knowledge graph. The historical graph node data corresponding to historical safety events in the historical maintenance records are imported into the risk knowledge graph. The second model is embedded and learned based on the historical graph node data, historical dynamic benchmark data, the historical safety events and the causal factors leading to the historical safety events in the risk knowledge graph.

[0026] Specifically, since the historical maintenance records record the gas data, environmental data and corresponding scenario modes under the historical safety incident status, the historical dynamic benchmark data under the above scenario modes, user behavior, user attributes and the causes of the above historical safety incidents, therefore, by extracting the atlas entities from the historical maintenance records, the above-mentioned atlas entities include gas entities, environmental entities, user entities, scenario entities, benchmark entities, safety event entities and safety factor entities. The above-mentioned gas entities include gas valve status entities, gas flow, pressure data and gas concentration data of gas leak alarms. The above-mentioned environmental data entities include real-time temperature, humidity, ventilation status and the operating status of ventilation equipment. The above-mentioned real-time temperature is outdoor temperature, indoor temperature and outdoor and indoor humidity. The above-mentioned scenario entity is the user's gas usage scenario type. The above-mentioned benchmark entity is the normal gas usage corresponding to the user's current gas usage scenario. The threshold range of the parameter, the above-mentioned user entity includes user attributes, the above-mentioned security event entity is the type of security event, and the above-mentioned security factor entity is the cause of the above-mentioned security event. By using the gas data, environmental data, user attributes, gas usage scenarios and historical security event types and the cause of the historical security events corresponding to the occurrence of the above-mentioned historical security events as historical graph node data, and saving them to the storage location of the corresponding entity in the above-mentioned risk knowledge graph, and based on the difference vector between the above-mentioned historical graph node data and the corresponding gas data and the historical dynamic benchmark data, the above-mentioned risk prediction model is embedded and learned. Through the above-mentioned technical solution, the above-mentioned risk prediction model can be obtained accurately, which lays the foundation for further predicting the cause of risk based on the graph node data obtained in real time and the dynamic benchmark data in the corresponding scenario, wherein the above-mentioned graph node data is the above-mentioned real-time data stream.

[0027] Furthermore, if Figure 3 As shown, the acquisition of the simulation results includes: The real-time data stream and the user attributes are input into the first model to obtain the gas usage scenario and the corresponding dynamic benchmark data. The real-time data stream, the dynamic benchmark data and the gas usage scenario are updated to the corresponding node of the risk knowledge graph, and the difference vector between the gas data in the real-time data stream and the dynamic benchmark data is calculated. When the length of the difference vector is less than the set value, the gas usage is normal; when the length of the difference vector is greater than or equal to the set value, the second model simulates the difference vector based on the real-time data stream, the gas usage scenario and the dynamic benchmark data in the risk knowledge graph to obtain the simulation result and the weight of the target graph node corresponding to the simulation result, and displays each target graph node in a different color according to the weight size, wherein the larger the weight, the more eye-catching the color of the corresponding target graph node, and the target graph node is a graph node whose weight value is greater than the set threshold.

[0028] Specifically, by inputting the above-mentioned real-time data stream and user attributes into the above-mentioned first model, the gas usage scenario of the user output by the above-mentioned first model and the dynamic benchmark data of the gas data in the gas usage scenario are obtained. The above-mentioned dynamic benchmark data is dynamic gas data, which at least includes gas flow data, gas pressure data and gas concentration data. The above-mentioned real-time data, the above-mentioned gas usage scenario and the above-mentioned dynamic benchmark data are updated to the above-mentioned risk knowledge graph, and the difference vector of the gas data in the above-mentioned real-time data stream and the corresponding data of the above-mentioned dynamic benchmark data is calculated. The above-mentioned difference vector reflects the user's deviation from the benchmark value in the current gas usage scenario. The length of the above-mentioned difference vector, that is, the Euclidean distance between the gas data in the above-mentioned real-time data stream and the above-mentioned dynamic benchmark data is calculated. When the length of the above-mentioned difference vector is less than the above-mentioned set value, it means that the gas data in the above-mentioned real-time data stream is closer to the above-mentioned dynamic benchmark data. At this time, it is considered that the user is in a normal gas usage state. Conversely, when the length of the above-mentioned difference vector is less than the above-mentioned set value, it means that the gas data in the above-mentioned real-time data stream is closer to the above-mentioned dynamic benchmark data. When the length of the quantity is greater than or equal to the above-mentioned set value, it means that the gas data in the real-time data stream deviates more from the above-mentioned dynamic benchmark data. At this time, it is considered that the above-mentioned user may have an unsafe gas usage risk. The environmental data and user attributes are obtained through the above-mentioned real-time data stream. The above-mentioned second model simulates the above-mentioned difference vector, that is, the above-mentioned unsafe gas usage situation based on the above-mentioned gas data and dynamic benchmark data under the above-mentioned environmental data, the above-mentioned user attributes and the above-mentioned gas usage scenario conditions, and obtains the above-mentioned simulation results and the corresponding target graph node weights that cause the above-mentioned unsafe gas usage risks, such as: gas pressure graph nodes, gas flow nodes, etc., wherein the above-mentioned simulation results are unsafe gas usage types, for example: there are small cracks in the gas pipeline, which affect the above-mentioned gas pressure and gas flow, and different colors are displayed according to the weight of the above-mentioned target graph nodes, thereby forming a risk process diagram of the above-mentioned unsafe gas usage risks through the above-mentioned risk knowledge graph, thereby laying the foundation for accurately pushing safety information to users.

[0029] Furthermore, judging whether the gas usage scenario is a new gas usage scenario according to the simulation result and the difference vector includes: When the length of the difference vector between the gas data and the dynamic benchmark data in the real-time data stream is greater than or equal to a set value, and the simulation result corresponding to the difference vector cannot be obtained based on the risk knowledge graph and the second model, the real-time data stream corresponding to the user is collected in real time, and the real-time data stream is mapped to the time coordinate axis, and the target data stream corresponding to the new gas usage scenario is extracted based on the dynamic time warping DWT algorithm, and the first model is trained based on the target data stream and the corresponding new gas usage scenario; otherwise, the gas usage scenario is not a new gas usage scenario; Risk simulation is also performed on the basis of the dynamic benchmark data corresponding to the new gas usage scenario to obtain multiple groups of risk events and corresponding risk data, and the data are added to the corresponding graph node positions in the risk knowledge graph. The second model is trained based on the multiple groups of risk events and the corresponding risk data.

[0030] Specifically, when the length of the difference vector between the gas data and the dynamic benchmark data in the real-time data stream is greater than or equal to the set value, and the risk knowledge graph and the second model cannot simulate the simulation result corresponding to the difference vector, it is possible that the user's gas usage scenario has changed. For example, the gas outlet diameter of the user's cooking stove has changed, causing the gas flow rate to change, which is no longer compatible with the original cooking gas usage scenario, or the user is heating or taking a bath while cooking, so that the dynamic benchmark data of the gas usage scenario before the new gas usage scenario has a large deviation, but does not belong to any unsafe gas usage risk situation. Therefore, in order to improve the recognition accuracy of unsafe gas usage risk situations, the real-time data stream of the user is collected in real time. Since the gas data corresponding to different time periods of different gas usage scenarios are different, the real-time data stream is mapped to the corresponding time axis, and the target data stream corresponding to the new gas usage scenario is extracted based on the dynamic time warping DWT algorithm. The first model is trained based on the target data stream and the corresponding new gas usage scenario, wherein the target data stream includes the environmental data, gas data and user attributes under the new gas usage scenario. Otherwise, when the length of the difference vector is less than the set value, or the length of the length difference vector is greater than or equal to the set value and the simulation result can be obtained, the gas usage scenario is not a new gas usage scenario. Since there is a lack of historical safety event data when the new gas usage scenario just appears, unsafe gas usage occurs in the new gas usage scenario, and the cause of the unsafe gas usage cannot be obtained in time. Therefore, the dynamic baseline data under the new gas usage scenario can be obtained through the first model, and on the basis of the dynamic baseline data, multiple groups of risk events and corresponding risk data are simulated, and the second model is trained. Through the technical solution, the new gas usage scenario can be automatically learned in time, and the unsafe gas usage risk situation can be identified in time, thereby improving the accuracy of safety information push.

[0031] Furthermore, generating risk information push control information according to the simulation results includes: Based on the weight distribution of the graph nodes in the corresponding risk knowledge graph in the second model corresponding to the simulation results, the risk level of the risk warning information is determined, and according to user attributes and historical feedback records, the push method of the risk warning information is determined, and the weight of the corresponding graph node in the second model is adjusted according to the user's feedback on the pushed risk warning information.

[0032] Specifically, the second model analyzes the weight distribution of the corresponding nodes in the risk knowledge graph based on the simulation results. The weights reflect the contribution of different risk factors to the current risk, such as abnormal gas pressure, abnormal gas concentration, and abnormal gas flow. The system automatically classifies risk levels based on the weight distribution, such as high, medium, and low. For example, abnormal gas concentration has the highest weight, indicating greater danger and a higher risk level, ensuring that the priority of risk warning information matches the actual threat level. Secondly, based on user attributes, such as elderly people living alone and family users, and historical feedback records, such as the frequency of users' responses to past push notifications and the effectiveness of their actions, an appropriate push notification method is selected, such as SMS emergency notifications, app pop-ups, or voice reminders, to improve information reach efficiency. Finally, user feedback on push notifications, such as confirmation, ignore, or false alarm marking, will reversely adjust the weight of the graph node. If the user promptly handles the risk, the corresponding node weight is reduced; if the feedback is a false alarm, the model's sensitivity to the relevant features is optimized. This closed-loop mechanism achieves continuous optimization of risk prediction, enhances push notification accuracy and user trust, and reduces false alarm rates by dynamically learning user behavior patterns, improving the proactive and reliable management of gas safety.

[0033] Furthermore, in the heating gas scenario, a temperature-flow compensation curve is generated through historical indoor and outdoor temperature data and gas flow data, and the dynamic benchmark data is corrected based on the temperature-flow compensation curve; when the risk knowledge graph detects a change in user attributes, the first model training cycle is reset and real-time data stream is collected, and the first model is trained in a timely manner based on the real-time data stream.

[0034] Specifically, in the heating scenario, gas usage is highly correlated with indoor and outdoor temperatures. By analyzing historical indoor and outdoor temperature data and gas flow data, the system establishes a temperature-gas flow correlation curve and a temperature-flow compensation curve. The above temperature is the indoor temperature. The curve reflects the normal gas usage pattern under different temperature conditions. For example, at low temperatures, the gas flow increases to maintain the indoor temperature. The dynamic benchmark data was originally based on general scenario settings, but in the heating scenario, temperature changes will cause the benchmark value to deviate from actual needs. The system uses the temperature-flow compensation curve and combines real-time temperature data to correct the dynamic benchmark data to make it more in line with the normal gas usage status in the current environment. Through the temperature compensation curve, the dynamic benchmark data is more accurate in the heating scenario, reducing false alarms caused by ambient temperature fluctuations, such as misjudging normal high heating gas consumption as leaks. When the knowledge graph detects changes in user attributes, such as changes in family population, house renovations, etc., the system determines that the original model may not be able to adapt to the new gas usage pattern. At this time, the training cycle of the first model is reset and the real-time data stream is re-collected. Based on the new data, the first model re-divides the gas usage scenarios and updates the dynamic baseline data through the dynamic time warping (DTW) algorithm. This mechanism ensures that the model can quickly learn the user's new habits and avoid risk misjudgment due to attribute changes. When user attributes change, the model is retrained in a timely manner to ensure the continued accuracy of risk identification. For example, when an elderly person living alone becomes a multi-person household, the system can quickly adapt to the new gas usage peak and avoid frequent false alarms.

[0035] Furthermore, the gas data includes gas flow, gas concentration data and gas pressure data, and the environmental data includes ambient temperature, humidity and ventilation status data.

[0036] The present invention also provides a gas safety knowledge accurate push system based on user feature analysis, which is used to implement the above method, such as Figure 4 As shown, the system includes: The collection unit is used to collect gas data and environmental data in real time and generate a real-time data stream aligned in time and space based on user types; A training unit, configured to divide the historical data stream into user gas usage scenarios using a dynamic time warping (DTW) algorithm, and train a first model based on the user gas usage scenarios and the corresponding historical data stream; A construction unit is used to extract graph nodes based on historical maintenance records, import historical graph node data and historical benchmark data corresponding to historical security events, construct a risk knowledge graph, and embed learning of the second model based on the risk knowledge graph; a simulation unit, configured to input the real-time data stream into the first model, obtain dynamic benchmark data and a gas usage scenario, and obtain a simulation result based on the real-time data stream, the dynamic benchmark data, the gas usage scenario, and the second model; A push unit is used to determine whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector, and generate risk push control information based on the simulation result.

[0037] The present invention also provides a computer-readable storage medium having instructions stored thereon, and the above-mentioned method is implemented when the instructions are executed by a processor.

[0038] In summary, the present invention generates a real-time data stream by collecting gas and environmental data in real time and combining it with user attributes, divides the user's gas usage scenarios into two categories and trains a first model using a dynamic time warping algorithm, and constructs a risk knowledge graph for embedded learning, thereby achieving accurate identification and risk prediction of gas usage scenarios. In real-time monitoring, the real-time data stream is input into the model to obtain the user's gas usage scenario and the corresponding dynamic benchmark data. The difference vector between the real-time data and the dynamic benchmark data is calculated, and the risk knowledge graph is used to simulate the above difference vector based on the second model to obtain the simulation result. In addition, according to the above difference vector and the above simulation result, safety incidents are discovered in time and it is determined whether it is a new gas usage scenario. If it is not a new scenario, risk push control information is generated and a prompt is pushed, which effectively improves the timeliness and accuracy of gas safety monitoring. If it is a new gas usage scenario, the new gas usage scenario data is collected in time, and the simulation result is corrected in time. The above-mentioned first model is trained to improve the ability to recognize new gas usage scenarios. After identifying safety incidents through the above simulation results, risk prompts are accurately pushed according to different user characteristics, which improves users' awareness and handling capabilities of gas risks, reduces false alarm rates, and enhances the initiative and reliability of gas safety management. In the heating scenario, the dynamic baseline data is corrected through the temperature-flow compensation curve to reduce false alarms caused by ambient temperature fluctuations. When changes in user attributes are detected, the model training cycle is reset and trained in time, so that the model can quickly adapt to new gas usage patterns and ensure the continued accuracy of risk identification. Overall, this patent starts from data collection, scene recognition, risk prediction to information push and other aspects, and comprehensively improves the efficiency of gas safety monitoring and management, providing gas users with more accurate and efficient safety protection services, and effectively preventing the occurrence of gas accidents.

[0039] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0040] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0041] As described above, the above 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 above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, 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 embodiments of the present application.

Claims

1. A method for accurately pushing gas safety knowledge based on user feature analysis, characterized in that: The method comprises: By collecting gas data and environmental data in real time and combining them with user types, a real-time data stream aligned in time and space is generated; Using a dynamic time warping (DTW) algorithm to divide the historical data stream into user gas usage scenarios, and training a first model based on the user gas usage scenarios and the corresponding historical data stream; Extract graph nodes based on historical maintenance records, import historical graph node data and historical benchmark data corresponding to historical security events, build a risk knowledge graph, and embed learning of the second model based on the risk knowledge graph; inputting the real-time data stream into the first model to obtain dynamic benchmark data and gas usage scenarios, and obtaining simulation results based on the real-time data stream, the dynamic benchmark data, the gas usage scenarios, and the second model; A push unit is used to determine whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector, and generate risk push control information based on the simulation result.

2. The method according to claim 1, characterized in that The training of the first model includes: aligning historical gas data, historical environmental data, and historical user attributes in a historical multimodal data stream according to the collection time to obtain a historical time series data stream, drawing a gas flow curve based on the gas data in the historical time series data stream of different dates, obtaining multiple groups of curve segments in any two flow curves whose similarity is greater than a set threshold through a dynamic time warping (DWT) algorithm, generating label data corresponding to the curve segment in each cluster cluster based on each group of the curve segments, the corresponding environmental data, and the user attributes through a clustering algorithm, and using the time series data segments corresponding to the curve segments in the time series data stream and the label data corresponding to the curve segments as training data to train the first model.

3. The method according to claim 1, characterized in that The construction of the risk knowledge graph and the training of the second model include: Graph entities are extracted from the historical maintenance records through semantic analysis, and the graph entities include at least gas entities, environment entities, user entities, scenario entities, benchmark entities, safety event entities and safety factor entities. The graph entities are used as graph nodes, and the correlation between the graph nodes is obtained through semantic analysis of the historical maintenance records. Edge connections are performed based on the correlation to construct the risk knowledge graph. The historical graph node data corresponding to historical safety events in the historical maintenance records are imported into the risk knowledge graph. The second model is embedded and learned based on the historical graph node data, historical dynamic benchmark data, the historical safety events and the causal factors leading to the historical safety events in the risk knowledge graph.

4. The method according to claim 1, wherein The acquisition of the simulation results includes: inputting the real-time data stream into the first model, obtaining the gas usage scenario and the corresponding dynamic benchmark data, updating the real-time data stream, the dynamic benchmark data and the gas usage scenario to the corresponding node of the risk knowledge graph, calculating the difference vector between the gas data in the real-time data stream and the dynamic benchmark data, and when the length of the difference vector is less than the set value, the gas usage is normal; when the length of the difference vector is greater than or equal to the set value, the second model simulates the difference vector based on the real-time data stream, the gas usage scenario, and the dynamic benchmark data in the risk knowledge graph to obtain the simulation result and the weight of the target graph node corresponding to the simulation result, and displays each target graph node in different colors according to the weight size, wherein the larger the weight, the more eye-catching the color of the corresponding target graph node, and the target graph node is a graph node whose weight value is greater than the set threshold.

5. The method according to claim 1, characterized in that Determining whether the gas usage scenario is a new gas usage scenario according to the simulation result and the difference vector includes: When the length of the difference vector between the gas data and the dynamic benchmark data in the real-time data stream is greater than or equal to a set value, and the simulation result corresponding to the difference vector cannot be obtained based on the risk knowledge graph and the second model, the real-time data stream corresponding to the user is collected in real time, and the real-time data stream is mapped to the time coordinate axis, and the target data stream corresponding to the new gas usage scenario is extracted based on the dynamic time warping DWT algorithm, and the first model is trained based on the target data stream and the corresponding new gas usage scenario; otherwise, the gas usage scenario is not a new gas usage scenario; Risk simulation is also performed on the basis of the dynamic benchmark data corresponding to the new gas usage scenario to obtain multiple groups of risk events and corresponding risk data, and the data are added to the corresponding graph node positions in the risk knowledge graph. The second model is trained based on the multiple groups of risk events and the corresponding risk data.

6. The method according to claim 1, characterized in that Generate risk information push control information based on the simulation results, including: Based on the weight distribution of the graph nodes in the corresponding risk knowledge graph in the second model corresponding to the simulation results, the risk level of the risk warning information is determined, and according to user attributes and historical feedback records, the push method of the risk warning information is determined, and the weight of the corresponding graph node in the second model is adjusted according to the user's feedback on the pushed risk warning information.

7. The method according to claim 1, characterized in that In the heating gas usage scenario, a temperature-flow compensation curve is generated using historical indoor and outdoor temperature data and gas flow data, and the dynamic baseline data is corrected based on the temperature-flow compensation curve; When the risk knowledge graph detects a change in user attributes, the first model training cycle is reset and real-time data stream is collected, and the first model is trained in a timely manner based on the real-time data stream.

8. The method according to claim 1, characterized in that The gas data includes gas flow, gas concentration data and gas pressure data, and the environmental data includes environmental temperature, humidity and ventilation status data.

9. A gas safety knowledge accurate push system based on user feature analysis, used to implement the method according to any one of claims 1 to 8, characterized in that: The system comprises: The collection unit is used to collect gas data and environmental data in real time and generate a real-time data stream aligned in time and space based on user types; A training unit, configured to divide the historical data stream into user gas usage scenarios using a dynamic time warping (DTW) algorithm, and train a first model based on the user gas usage scenarios and the corresponding historical data stream; A construction unit is used to extract graph nodes based on historical maintenance records, import historical graph node data and historical benchmark data corresponding to historical security events, construct a risk knowledge graph, and embed learning of the second model based on the risk knowledge graph; a simulation unit, configured to input the real-time data stream into the first model, obtain dynamic benchmark data and a gas usage scenario, and obtain a simulation result based on the real-time data stream, the dynamic benchmark data, the gas usage scenario, and the second model; A push unit is used to determine whether the gas usage scenario is a new gas usage scenario based on the simulation result and the difference vector, and generate risk push control information based on the simulation result.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images