Gas leak intelligent early warning method and Internet of Things system for smart gas

Abstract

The embodiments of this specification provide a smart gas leak intelligent early warning method and Internet of Things system for smart gas, which is executed by a smart gas indoor safety management sub-platform, and includes: obtaining monitoring information and user description information, wherein the monitoring information includes alarm information and gas terminal monitoring data, and the user description information includes customized alarm information uploaded by the user; determining the cause of the gas leak based on the monitoring information and the user description information; determining candidate solutions based on the cause of the gas leak; determining the effective solution rate of the candidate solutions for the cause of the gas leak and the candidate solutions based on an effective solution rate determination model, and determining the target solution from the candidate solutions based on the effective solution rate, wherein the effective solution rate determination model is a machine learning model.

Description

[0001] Description of the case

[0002] This application is a divisional application filed in response to the Chinese application with application date of October 20, 2022, application number 202211283273.2, and invention name “Smart gas terminal linkage disposal method and Internet of Things system for achieving indoor safety”. Technical Field

[0003] This specification relates to the field of smart gas, and in particular to a gas leak intelligent early warning method and Internet of Things system for smart gas. Background Art

[0004] Gas systems are commonly used indoor energy supply systems, for example, providing fuel for water heaters and stoves. However, with the widespread use of gas systems, the number of gas accidents has also been increasing, particularly those caused by gas leaks. Because gas is flammable, explosive, fluid, and easily diffused, leaks can easily cause explosions and fires when exposed to open flames or sparks.

[0005] In view of this, it is hoped to provide a gas leak intelligent early warning method and Internet of Things system for smart gas, which can automatically determine the cause of gas leaks and provide solutions while improving the efficiency of troubleshooting the cause of gas leaks. Summary of the Invention

[0006] One or more embodiments of this specification provide a smart gas leak intelligent early warning method. The method is executed by a smart gas indoor safety management sub-platform and includes: obtaining monitoring information and user description information, wherein the monitoring information includes alarm information and gas terminal monitoring data, and the user description information includes customized alarm information uploaded by the user; determining the cause of the gas leak based on the monitoring information and the user description information; determining candidate solutions based on the cause of the gas leak; determining the effective solution rate of the candidate solutions based on the gas leak cause and the candidate solutions based on an effective solution rate determination model; and determining the target solution from the candidate solutions based on the effective solution rate, wherein the effective solution rate determination model is a machine learning model.

[0007] One or more embodiments of this specification provide an intelligent gas leak early warning Internet of Things system for smart gas. The Internet of Things system includes a smart gas user platform, a smart gas service platform, a smart gas safety management platform, a smart gas sensor network platform, and a smart gas object platform. The smart gas safety management platform includes a smart gas indoor safety management sub-platform and a smart gas data center. The smart gas indoor safety management sub-platform is configured to perform the following operations:

[0008] Acquire monitoring information and user description information, wherein the monitoring information includes alarm information and gas terminal monitoring data, and the user description information includes customized alarm information uploaded by the user; determine the cause of the gas leakage based on the monitoring information and the user description information; determine candidate solutions based on the cause of the gas leakage; determine the effective solution rate of the candidate solutions for the cause of the gas leakage and the candidate solutions based on an effective solution rate determination model, and determine the target solution from the candidate solutions based on the effective solution rate, wherein the effective solution rate determination model is a machine learning model.

[0009] One or more embodiments of this specification provide a gas leakage intelligent early warning device for smart gas, including a processor, wherein the processor is used to execute the above-mentioned gas leakage intelligent early warning method for smart gas.

[0010] One or more embodiments of this specification provide a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the above-mentioned intelligent gas leak warning method for smart gas. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein:

[0012] Figure 1 This is an exemplary structural diagram of the Internet of Things structure of the smart gas terminal linkage disposal system according to some embodiments of this specification;

[0013] Figure 2 is an exemplary flow chart of a smart gas terminal linkage disposal method according to some embodiments of this specification;

[0014] Figure 3 is a schematic diagram of a gas leakage cause prediction model according to some embodiments of this specification;

[0015] Figure 4 is a schematic diagram of the structure of a gas leakage cause prediction model according to some embodiments of this specification;

[0016] Figure 5 is an exemplary flow chart of a method for determining a target solution according to some embodiments of this specification. DETAILED DESCRIPTION

[0017] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0018] It should be understood that the terms "system," "device," "unit," and / or "module" used herein are a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other terms can achieve the same purpose, the terms may be replaced by other expressions.

[0019] Unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not refer to the singular but include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0020] Flowcharts are used throughout this specification to illustrate the operations performed by systems according to embodiments of this specification. It should be understood that preceding or following operations do not necessarily need to be performed in exact order. Instead, the steps may be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0021] Figure 1 This is an exemplary structural diagram of the Internet of Things structure of the smart city gas terminal linkage disposal system shown in some embodiments of this specification.

[0022] It should be understood that the smart city gas terminal linkage disposal system 100 can be implemented in various ways. Figure 1 As shown, the smart city gas terminal linkage disposal system can include a smart gas user platform 110, a smart gas service platform 120, a smart gas safety management platform 130, a smart gas sensor network platform 140 and a smart gas object platform 150.

[0023] The smart gas user platform 110 can be a user-driven, interactive platform that can be configured as a terminal device to provide users with gas anomaly information and corresponding solutions. The gas anomaly information can include at least gas system monitoring information and user-uploaded customized alarm information.

[0024] In some embodiments, the smart gas user platform 110 may include a gas user sub-platform and a supervisory user sub-platform. The gas user sub-platform is a sub-platform for gas users that can interact with and exchange data with the smart gas service sub-platform, providing gas users with gas-related data and gas problem solutions. The supervisory user sub-platform can be a scoring platform for supervisory users, allowing supervisory users to monitor the operation of the entire IoT system.

[0025] In some embodiments, the smart gas user platform can exchange data with the gas service sub-platform of the smart gas service platform 120. For example, the smart gas user platform 110 can issue a household gas safety information query instruction to the smart gas service platform 120. For another example, the smart gas user platform 110 can receive household gas safety information uploaded by the smart gas service platform 120. Indoor gas safety information may include gas anomaly information (for example, gas leaks, etc.) and corresponding solutions. Specifically, the supervisory user can issue a query instruction to the smart supervisory service sub-platform through the supervisory user sub-platform to obtain the gas safety status of the relevant jurisdiction of the smart supervisory service sub-platform; gas users can obtain safety reminder information from the smart gas service sub-platform through the gas user sub-platform.

[0026] The smart gas service platform 120 may be a platform that provides users with safe gas use and safety supervision services.

[0027] In some embodiments, the smart gas service platform 120 may include a smart gas usage service sub-platform and a smart supervision service sub-platform. The smart gas usage service sub-platform may correspond to the gas user sub-platform and be used to provide safe gas usage services to gas users. The smart supervision service sub-platform may correspond to the supervision user sub-platform and be used to provide safety supervision services to supervised users.

[0028] In some embodiments, the smart gas service platform 120 can exchange data with the smart gas user platform 110 and the smart gas safety management platform 130. For example, the smart gas service platform 120 can issue a household gas safety information query instruction to the smart gas safety management platform 130 and receive household gas safety information uploaded by the smart gas safety management platform 130. For another example, the smart gas service platform can receive a household gas safety information query instruction issued by the smart gas user platform 110 and upload the household gas safety information from the smart gas safety management platform 130.

[0029] The smart gas safety management platform 130 coordinates and coordinates the connections and collaborations between various functional platforms, aggregating all data and information from the smart city gas terminal linkage and disposal system 100. The management platform provides data management, control management, and data analysis capabilities for the operation of the smart city gas terminal linkage and disposal system 100.

[0030] In some embodiments, the management platform may be a remote platform controlled by a manager, artificial intelligence, or preset rules.

[0031] In some embodiments, the smart gas safety management platform 130 may include a smart gas data center and at least one smart gas indoor safety management sub-platform. Each smart gas indoor safety management sub-platform may correspond to a gas system used by a user. The smart gas indoor safety management sub-platform interacts bidirectionally with the smart gas data center. The smart gas indoor safety management sub-platform obtains and returns safety management data of indoor gas terminal devices from the smart gas data center. The data center can aggregate and store all operational data of the smart city gas terminal linkage disposal system 100.

[0032] In some embodiments, the smart gas indoor safety management sub-platform may include at least one safety management module. Among them, at least one safety management module may include an intrinsic safety monitoring and management module, an information security monitoring and management module, and a functional monitoring and management module. The intrinsic safety monitoring and management module can monitor data on explosion-proof safety such as mechanical leakage, electrical power consumption (intelligent control power consumption, communication power consumption), and valve control; the information security monitoring and management module can monitor data anomalies, illegal device information, illegal access, etc. Functional monitoring management includes monitoring of functional safety such as long-term non-use, continuous flow timeout, flow overload, abnormally large flow, abnormally small flow, low air pressure, strong magnetic interference, and low voltage; the functional monitoring and management module can monitor functional safety such as long-term non-use, continuous flow timeout, flow overload, abnormally large flow, abnormally small flow, low air pressure, strong magnetic interference, and low voltage.

[0033] In some embodiments, the smart gas data center identifies the safety parameter category and sends the acquired safety data to the corresponding safety management module. Each safety management module has a preset safety threshold. When the safety data exceeds the safety threshold, the smart gas safety management platform can automatically issue an alarm and can choose to automatically push the alarm information to gas users and / or supervisory users.

[0034] In some embodiments, the smart gas safety management platform 130 can exchange data with the smart gas sensor network platform 140 and the smart gas service platform 120. For example, the smart gas safety management platform 130 can issue instructions for indoor gas safety-related data to the smart gas sensor network platform 140 and receive indoor gas safety-related data uploaded by the smart gas sensor network platform 140. In another example, the smart gas safety management platform 130 can receive instructions for querying indoor gas safety-related data issued by the smart gas service platform 120 and upload the indoor gas safety-related data to the smart gas service platform 120.

[0035] In some embodiments, data exchange between the smart gas safety management platform 130, the smart gas service platform 120, and the smart gas sensor network platform 140 is conducted through the smart gas data center. For example, the smart gas data center receives a query instruction for indoor gas anomaly information from the smart gas service platform 120; the smart gas data center issues an instruction to obtain data related to indoor gas anomalies to the smart gas sensor network platform 140; the smart gas data center transmits the indoor gas anomaly data from the smart gas sensor network platform 140 to the smart gas safety management sub-platform for analysis and processing. The smart gas indoor safety management sub-platform transmits the processed data to the smart gas data center; and the smart gas data center transmits the aggregated and processed data (e.g., the cause of the gas anomaly and the corresponding solution) to the smart gas service platform 120.

[0036] In some embodiments, the smart gas safety management platform 130 can be used to determine the cause of gas leakage based on the monitoring information of the gas system and the user description information, and further determine the target solution based on the cause of gas leakage. For more information on determining the cause of gas leakage and the target solution, please refer to this specification. Figures 2 to 5 and its related descriptions.

[0037] The smart gas sensor network platform 140 may be a functional platform for managing sensor communications and may be configured as a communication network and a gateway.

[0038] In some embodiments, the smart gas sensor network platform 140 may include at least one sensor network sub-platform. This at least one sensor network sub-platform may include sub-platforms for network management, protocol management, command management, and data analysis, responsible for managing networks, protocols, and commands, and parsing relevant data. A sensor network sub-platform may correspond to a user's gas system and a management sub-platform of the smart gas safety management platform 130. For example, a sensor network sub-platform may transmit monitoring data of a user's indoor gas system, acquired through the smart gas object platform 150, to the corresponding management sub-platform.

[0039] In some embodiments, the smart gas sensor network platform 140 can exchange data with the smart gas safety management platform 130 and the smart gas object platform 150 to implement sensing information sensing and communication and control information sensing and communication. For example, the smart gas sensor network platform 140 can issue a command to obtain indoor gas system monitoring data to the smart gas object platform 150 and receive the indoor gas system monitoring data uploaded by the smart gas object platform 150. For another example, the smart gas sensor network platform 140 can receive a command to obtain indoor gas system monitoring data from the smart gas data center of the smart gas safety management platform 130 and upload the indoor gas system monitoring data to the smart gas data center.

[0040] The smart gas object platform 150 can be a functional platform for generating sensory information and can be configured as various devices, such as indoor gas equipment and safety detection equipment. In some embodiments, the object platform can obtain information, such as monitoring information of a user's gas system.

[0041] In some embodiments, the smart gas object platform 150 can obtain gas safety-related information through at least one sub-platform. The at least one sub-platform can include a fair metering device object sub-platform, a safety monitoring device object sub-platform, and a safety valve control device object sub-platform. The safety monitoring device object sub-platform can be configured as a gas concentration detection device to detect gas leaks.

[0042] In some embodiments, the smart gas object platform 150 can exchange data with the smart gas sensor network platform 140. For example, the smart gas object platform 150 can receive instructions from the smart gas sensor network platform 140 to obtain gas safety-related data and upload the household gas safety-related data to the smart gas sensor network platform 140.

[0043] It should be noted that the above description of the smart city gas terminal linkage disposal system and its different platforms is for convenience only and does not limit this specification to the scope of the embodiments. It is understandable that for those skilled in the art, after understanding the principles of the system, they may arbitrarily combine the modules or form subsystems to connect with other modules without deviating from the principles. For example, Figure 1 The smart gas safety management platform, smart gas user platform, smart gas service platform, smart gas sensor network platform and smart gas object platform disclosed in the document can be different platforms in a system or the same platform in a system.

[0044] Figure 2This is an exemplary flow chart of the smart city gas terminal linkage disposal method shown in some embodiments of this specification. Figure 2 As shown, the process 200 includes the following steps. In some embodiments, the process 200 can be executed by the smart gas safety management platform 130.

[0045] Step 210: Acquire monitoring information and user description information.

[0046] Monitoring information refers to monitoring information of the gas system, which may include information related to the gas system status that is monitored and recorded in real time by the gas system. In some embodiments, the monitoring information of the gas system may include alarm information and gas terminal monitoring data.

[0047] Alarm information refers to alarm information related to abnormal conditions in the gas system. For example, alarm information may include excessive gas concentration (for example, gas concentration exceeding 25% of the lower explosive limit) or multiple ignition failures within a short period of time (for example, ignition failures three or more times within one minute).

[0048] Gas terminal monitoring data refers to various monitoring indicator data within a preset time period (e.g., the last 3 minutes) from the current moment. In some embodiments, the monitoring indicator data may include gas concentration, the number of times the stove is opened or closed, the number of times the water heater valve is opened or closed, whether the ignition is successful, the fire power, etc.

[0049] In some embodiments, gas terminal monitoring data can be obtained by sensors on the smart gas object platform 150. Sensors can be installed at various locations within the gas system or indoors, such as inside pipes, on ceilings, etc. There can be multiple sensors to monitor multiple locations.

[0050] In some embodiments, the alarm information can be determined by the smart gas indoor safety management sub-platform. For example, the smart gas indoor safety management sub-platform can obtain gas concentration information based on the gas concentration sensor and determine whether the gas concentration is higher than a preset concentration threshold (for example, the gas concentration is higher than 25% of the lower explosion limit concentration). If so, the smart gas indoor safety management sub-platform generates a gas concentration alarm information and sends it to the smart gas service platform through the smart gas data center, and then sends it to the smart gas user platform 110.

[0051] In some embodiments, the supervisory user sub-platform of the smart gas user platform can issue a monitoring information acquisition instruction to the smart gas service platform, and the smart gas service platform further issues the instruction to the smart gas data center of the smart gas safety management platform. The smart gas data center uploads the stored detection information to the supervisory user sub-platform through the smart gas service platform.

[0052] In some embodiments, the smart gas data center obtains monitoring information from the smart gas object platform through the smart gas sensor network platform and sends it to the smart gas indoor safety management sub-platform for processing.

[0053] User description information refers to information entered by the user to describe the gas system status. For example, user description information may include a customized alarm message uploaded by the user.

[0054] User-uploaded custom alarm information refers to information entered by the user to describe the current abnormality in the gas system. For example, custom alarm information may include failure to ignite, insufficient firepower, or inability to adjust firepower.

[0055] In some embodiments, the customized alarm information can be input by the user through the user terminal based on text or filling in a form, and uploaded by the user to the smart gas user platform 110, and then sent by the smart gas user platform 110 to the smart gas service platform 120 or the smart gas safety management platform 130.

[0056] In some embodiments, users can upload user description information to the smart gas service sub-platform of the smart gas service platform through the gas user sub-platform of the smart gas user platform. The smart gas service sub-platform can upload the user description information to the smart gas data center, and the smart gas data center will send the user description information to the smart gas indoor safety management sub-platform for further processing.

[0057] Step 220: Determine the cause of the gas leakage based on the monitoring information and the user description information.

[0058] In some embodiments, the smart gas safety management platform 130 can determine the cause of a gas leak based on monitoring information and user-defined information. In some embodiments, the cause of a gas leak can include a loose connection (e.g., a hose) between the gas meter, stove, or water heater, or a malfunctioning component in the gas meter. For example, if the gas system detects an abnormal pressure at the gas meter connection, the smart gas safety management platform 130 can determine that the cause of the gas leak is a loose connection at the gas meter interface.

[0059] In some embodiments, the smart gas safety management platform 130 can input the detection information of the gas system and the user description information into the gas leakage cause prediction model to obtain the gas leakage cause, wherein the gas leakage cause prediction model can be trained based on the historical data of the gas system. For more information about the gas leakage cause prediction model, please refer to the Figure 3 and Figure 4 See the relevant instructions in .

[0060] Step 230: Determine a target solution based on the cause of the gas leak.

[0061] A target solution can refer to a solution for resolving a gas system failure. For example, if the smart gas safety management platform 130 determines that a gas leak is caused by a loose gas meter connection, the smart gas safety management platform 130 may select "Replace the gas meter hose" as the target solution. For another example, if the smart gas safety management platform 130 determines that a gas leak is caused by a faulty component in the gas meter, the smart gas safety management platform 130 may select "Replace the gas meter" or "Repair the gas meter" as the target solution. For another example, if the smart gas safety management platform 130 determines that a gas leak is caused by corrosion in the gas pipeline, the smart gas safety management platform 130 may select "Replace the gas pipeline" as the target solution.

[0062] In some embodiments, the smart gas indoor safety management sub-platform of the smart gas safety management platform 130 can determine the target solution based on various methods, such as historical data statistics or model prediction. In some embodiments, the smart gas safety management platform 130 can collect historical data on solutions to various gas leak causes and select the most frequently used solution or the one with the highest resolution rate as the target solution.

[0063] In some embodiments, the smart gas indoor safety management sub-platform can upload the cause of gas leakage and the corresponding target solution to the smart gas data center, and the smart gas data center will send the cause of gas leakage and the corresponding target solution to the smart gas service platform. The smart gas service platform will send the cause of gas leakage and the corresponding target solution to the smart gas user platform for users to view.

[0064] In some embodiments, the smart gas safety management platform 130 can construct a historical feature vector based on historical data, and construct a gas leakage cause vector based on the monitoring information of the gas system and user description information, and determine the target solution based on the similarity between the historical feature vector and the gas leakage cause vector (for example, the highest similarity means the highest effective solution rate).

[0065] The smart city gas terminal linkage disposal method provided in some embodiments of this specification can automatically determine the cause of gas leakage and provide corresponding target solutions based on the monitoring information of the gas system and user description information, avoiding the low efficiency and difficulty of manual investigation and improving the efficiency of solving gas leakage problems.

[0066] It should be noted that the above description of process 200 is for illustration and purpose only and does not limit the scope of application of this specification. Those skilled in the art may make various modifications and variations to process 200 under the guidance of this specification. However, such modifications and variations are still within the scope of this specification.

[0067] Figure 3 Schematic diagram of a gas leakage cause prediction model according to some embodiments of this specification. Figure 3 As shown, the method 300 for determining the cause of gas leakage using a gas leakage cause prediction model includes the following contents.

[0068] In some embodiments, the smart gas safety management platform 130 can determine the cause of a gas leak using a gas leak cause prediction model based on gas system monitoring information and user description information. The gas leak cause model can be a machine learning model.

[0069] The input of the gas leakage cause prediction model 340 may include monitoring information 310 and user description information 320, and the output may include gas leakage cause 350. For more information on the monitoring information and user description information of the gas system, please refer to Figure 2 Step 210 and its corresponding description, more relevant explanations on the causes of gas leakage can be found in Figure 2 Step 220 and its corresponding description.

[0070] In some embodiments, the input of the gas leakage cause prediction model may further include a historical feature vector 330 .

[0071] The historical feature vector 330 may be a vector of data generated based on historical data that can reflect the historical characteristics of the gas system. A historical feature vector can take various forms. For example, a historical feature vector may be (H1, H2, H3), where H1 represents historical alarm information, H2 represents historical gas terminal monitoring data, and H3 represents historical disposal information.

[0072] Historical data may refer to data reflecting the characteristics of the gas system within a historical time period, and may include historical monitoring information 331-1 and historical user description information 331-2 of the gas system. The historical monitoring information and historical user description information of the gas system may refer to the monitoring information and user description information of the gas system within a historical time period, respectively. The length of the historical time period may be manually preset, for example, one month. For details on the monitoring information and user description information of the gas system, please refer to Figure 2 and its corresponding description.

[0073] In some embodiments, historical data may include historical monitoring information and historical disposal information of the gas system, wherein the historical monitoring information may include historical alarm information of the gas system. Historical alarm information may refer to the alarm information of the gas system within a historical time period. For relevant instructions on alarm information, please refer to Figure 2 and its corresponding description.

[0074] Historical processing information can refer to records of operations performed by users on gas system components, such as repairs or replacements, within a historical period. For example, historical processing information could include "Battery replaced a month ago," "Stove pipe replaced a week ago," or "Ignition device repaired two weeks ago."

[0075] In some embodiments of this specification, determining a historical feature vector through historical data can enable the obtained historical feature vector to better reflect the historical characteristics of the gas system, thereby more accurately determining the cause of the current gas leak in combination with the historical characteristics of the gas system.

[0076] In some embodiments, the historical feature vector 330 can be obtained based on historical data through the embedding layer 331. The embedding layer 331 can be a machine learning model whose input can include historical monitoring information 331-1 of the gas system and historical user description information 331-2, and whose output can be the corresponding historical feature vector 330.

[0077] In some embodiments, the embedding layer 331 can be trained using multiple labeled training samples. For example, multiple labeled training samples can be input into the initial embedding layer, a loss function can be constructed using the labels and the results of the initial embedding layer, and the parameters of the initial embedding layer can be iteratively updated based on the loss function. When the loss function of the initial embedding layer meets a preset condition, model training is completed, resulting in a trained embedding layer. The preset condition can include convergence of the loss function, a number of iterations reaching a threshold, and the like.

[0078] In some embodiments, the gas leak cause prediction model 340 can be trained using multiple labeled training samples. For example, multiple labeled training samples 341-1 can be input into the initial gas leak cause prediction model 341. A loss function is constructed using the labels and the output of the initial gas leak cause prediction model 341. The parameters of the initial gas leak cause prediction model 341 are updated based on the iterations of the loss function. When the loss function satisfies preset conditions, training is complete, resulting in the trained gas leak cause prediction model 340.

[0079] In some embodiments, the training samples may come from historical monitoring information and historical user description information of the gas system, and the labels may be determined by manual input.

[0080] In some embodiments of this specification, by introducing historical feature vectors as input to a gas leakage cause prediction model, the accuracy of the gas leakage cause output by the model can be effectively improved.

[0081] Figure 4 This is a schematic diagram of the structure of the gas leakage cause prediction model shown in some embodiments of this specification. Figure 4As shown, another method 400 for determining a gas leakage cause by using a gas leakage cause prediction model includes the following contents.

[0082] In some embodiments, the gas leakage cause prediction model may include a pre-processing layer 441 and a prediction layer 442 .

[0083] The input of the pre-processing layer 441 may include monitoring information 410 of the gas system and user description information 420 , and the output may include a system feature vector 450 and a user feature vector 460 .

[0084] The pre-processing layer 441 may include a first processing layer 441-1 and a second processing layer 441-2, wherein the first processing layer 441-1 and the second processing layer 441-2 may be machine learning models.

[0085] The first processing layer 441 - 1 may be used to process the monitoring information 410 of the gas system and obtain a system feature vector 450 . The input of the first processing layer 441 - 1 may be the monitoring information 410 of the gas system and the output may be the system feature vector 450 .

[0086] System characteristic vector 450 may be vector data generated based on gas system monitoring information 410 and may reflect gas system characteristics. For example, a system characteristic vector may be (S1, S2, S3), where S1 is the acquisition time of the monitoring information, S2 is the alarm information, and S3 is the gas terminal monitoring data.

[0087] The second processing layer 441 - 2 may be configured to process the user description information 420 and obtain a user feature vector 460 . The input of the second processing layer 441 - 2 may be the user description information 420 , and the output may be the user feature vector 460 .

[0088] The user feature vector 460 may be vector data generated based on the user description information 420 and may reflect gas system characteristics. For example, a user feature vector may be (U1, U2), where U1 represents the time when the user description information was obtained and U2 represents the user description information.

[0089] The input of the prediction layer 442 may include the historical feature vector 430 , the system feature vector 450 , and the user feature vector 460 , and the output may be the gas leak cause 480 .

[0090] In some embodiments, the prediction layer 442 may include a first prediction layer 442-1 and a second prediction layer 442-2, wherein the first prediction layer 442-1 and the second prediction layer 442-2 may be machine learning models.

[0091] The first prediction layer 442-1 can be used to process the historical feature vector 430, the system feature vector 450 and the user feature vector 460 to obtain the gas leakage cause vector 470. Its input may include the historical feature vector 430, the system feature vector 450 and the user feature vector 460, and its output may be the gas leakage cause vector 470.

[0092] For details about the historical feature vector 430, see Figure 3 and its related descriptions.

[0093] Gas leak cause vector 470 may be generated based on the contents of historical feature vector 430, system feature vector 450, and user feature vector 460, and may represent vector data that can reflect the characteristics of the gas leak cause. Gas leak cause vectors can have various forms. For example, a gas leak cause vector may be (R1, R2, R3, R4, R5), where R1 is the time of the gas leak, R2 is the gas concentration detected at the time of the leak, R3 is the alarm information, R4 is the user description, and R5 is the cause of the gas leak.

[0094] The second prediction layer 442 - 2 may be used to process the gas leakage cause vector 470 to obtain the gas leakage cause 480 . The input of the second prediction layer 442 - 2 may be the gas leakage cause vector 470 , and the output may be the gas leakage cause 480 .

[0095] In some embodiments, the input to the second prediction layer 442 - 2 may also include a false alarm probability.

[0096] A false alarm may refer to a situation where the gas system's monitoring information 410 and / or the user description information 420 do not match the actual situation. For example, the gas system has sufficient firepower, but due to reasons such as a malfunction of the monitoring equipment, the corresponding generated gas system monitoring information may contain "small gas firepower". This situation may be determined as a false alarm. The false alarm probability may refer to the possibility of a false alarm, and its size may be represented in the form of a percentage. For example, a false alarm probability of 20% means that the gas system's monitoring information 410 and / or the user description information 420 has a 20% chance of not matching the actual situation. The method for determining the false alarm probability can be found in Figure 5 and its corresponding description.

[0097] In some embodiments, the first processing layer may include a supplementary data module. In response to a false alarm probability exceeding a threshold, the smart gas safety management platform supplements the system monitoring data through the supplementary data module and determines a new system feature vector based on the supplemented system monitoring data. Specifically, before determining the corresponding historical feature vector 330 based on the gas system's historical monitoring information 331-1 and historical user description information 331-2 through the embedding layer 331, the smart gas safety management platform may supplement the gas system's historical monitoring information 331-1 and / or historical user description information 331-2 in response to a false alarm probability exceeding a probability threshold. Data supplementation may involve expanding the types or quantity of data. Expanding the data types may involve adding data types included in the gas system's historical monitoring information 331-1 and historical user description information 331-2. For example, if the original gas system's historical monitoring information 331-1 and historical user description information 331-2 contain numerical data (e.g., gas concentration), the supplementary data type may be image-based data of gas power, for example. The probability threshold may be manually preset.

[0098] The type and quantity of data required for the data supplementation process can be determined based on preset rules. For example, a table can be pre-set to correspond between false alarm probabilities and the types and quantities of data required for supplementation. The types and quantities of data required for supplementation can then be determined based on the correspondence in the table. The types and quantities of data required for the data supplementation process can also be determined based on a data supplementation model. The data supplementation model can be a machine learning model whose inputs may include false alarm probabilities, historical gas system monitoring information 331-1, and historical user profile information 331-2. Its output may be the types and quantities of data required for supplementation.

[0099] In some cases, to conserve energy, not all data requires real-time monitoring. For example, important data such as gas concentration requires real-time monitoring, while other data (e.g., on-site images of the gas system) does not require real-time monitoring or can be monitored over longer time periods. In some embodiments of this specification, through data supplementation, when the false alarm probability exceeds a threshold, data not monitored in real time is obtained as supplementary data. This can correct the data input to the gas leak cause prediction model and prevent the model's output from significantly deviating from reality.

[0100] In some embodiments of this specification, by introducing the false alarm probability, it is possible to avoid the situation where the gas leakage cause is obtained incorrectly due to incorrect data input into the gas leakage cause prediction model.

[0101] In some embodiments of the present specification, the prediction layer of the gas leakage cause prediction model is further divided into two layers, and the monitoring information of the gas system and the user description information are processed respectively to obtain the corresponding system feature vectors and user feature vectors. This can standardize the form of the data and facilitate subsequent processing, thereby helping to improve the prediction accuracy of the gas leakage cause prediction model.

[0102] In some embodiments of the present specification, the gas leakage cause prediction model is divided into multiple layers. The monitoring information and user description information of the gas system are first processed by the preprocessing layer to obtain the system feature vector and the user feature vector. Then, the system feature vector, the user feature vector and the historical feature vector are processed by the prediction layer to obtain the gas leakage cause, which is beneficial to the prediction accuracy of the gas leakage cause prediction model.

[0103] In some embodiments, the output of the preprocessing layer 441 of the gas leakage cause prediction model can be used as the input of the prediction layer 442 , and the preprocessing layer 441 and the prediction layer 442 can be obtained through joint training.

[0104] In some embodiments, the sample data for joint training includes monitoring information of the sample gas system, sample user description information, and sample historical feature vectors, with labels being the sample system feature vector, the sample user feature vector, and the sample gas leakage cause. The joint training process may include: inputting the monitoring information of the sample gas system and the sample user description information into the preprocessing layer 441 to obtain the system feature vector and user feature vector output by the preprocessing layer 441; inputting the system feature vector and user feature vector output by the preprocessing layer 441 as training sample data and the sample historical feature vector into the prediction layer 442 to obtain the gas leakage cause output by the prediction layer 442. A loss function is constructed based on the sample gas leakage cause and the gas leakage cause output by the prediction layer 442, and the parameters of the preprocessing layer 441 and the prediction layer 442 are updated simultaneously. Through the parameter update, the trained preprocessing layer 441 and the prediction layer 442 are obtained.

[0105] In some embodiments of this specification, gas leakage causes are predicted using a gas leakage cause prediction model, which can not only accurately predict the cause of the gas leakage but also save manpower and time costs.

[0106] It should be noted that in some embodiments of this specification, objects with the same terminology but different numbers may represent the same content. For example, monitoring information 310 and monitoring information 410, user description information 320 and user description information 420, historical feature vectors 330 and historical feature vectors 430, and gas leak cause prediction models 340 and 440, etc.

[0107] Figure 5FIG. 1 is an exemplary flow chart of a method for determining a target solution according to some embodiments of this specification. Figure 5 As shown, the process 500 includes the following steps. In some embodiments, the process 500 can be executed by the smart gas safety management platform 130.

[0108] Step 510: Determine candidate solutions based on the cause of the gas leak.

[0109] Candidate solutions refer to one or more solutions corresponding to the cause of the gas leak. In some embodiments, candidate solutions may include one or more of replacing the hose, replacing the gas meter component, or replacing pipeline components. For example, if the cause of the gas leak is hose wear or aging, candidate solutions may include replacing the hose or replacing it with a metal hose.

[0110] In some embodiments, the smart gas safety management platform 130 can determine candidate solutions based on a variety of methods. For example, the smart gas safety management platform 130 can use a model or statistical method to sort multiple solutions corresponding to the causes of gas leakage (for example, sorting based on the historical success rate of the solutions, etc.), and automatically determine candidate solutions (for example, selecting the top N solutions in the sorting results).

[0111] In some embodiments, the smart gas safety management platform 130 may determine a candidate solution based on a feature vector by comparing the feature vector to a reference vector in a database, where the database includes multiple reference vectors, each of which has a corresponding candidate solution.

[0112] A feature vector is a vector constructed based on information about gas terminals in a gas system. There are various ways to construct feature vectors based on relevant information about a gas system. For example, a feature vector p is constructed based on the information (x, y, m, n) of a gas terminal in a gas system. The terminal information (x, y, m, n) can represent the monitoring information of the gas terminal as x, the user description information as y, the false alarm probability as m, and the gas leak cause as n. The gas leak cause can be determined based on the output of the first prediction layer in the gas leak cause prediction model.

[0113] The reference vector is constructed based on historical gas system monitoring information, historical user profiles, and historical gas leak causes. The candidate solution corresponding to the reference vector is the solution to the corresponding gas leak cause. The vector to be matched is constructed based on current monitoring information, current user profiles, and information related to the current gas leak cause. The construction methods of the reference vector and the vector to be matched are similar to the feature vectors described above.

[0114] In some embodiments, the smart gas safety management platform 130 can calculate the distance between the reference vector and the vector to be matched, and determine candidate solutions to the gas leak cause corresponding to the vector to be matched. For example, a reference vector whose distance from the vector to be matched meets a preset condition is used as the target vector, and the solution to the gas leak cause corresponding to the target vector is used as the candidate solution to the gas leak cause corresponding to the vector to be matched. The preset condition can be set according to the situation. For example, the preset condition can include the minimum vector distance or the vector distance being less than a distance threshold.

[0115] In some embodiments, the false alarm probability can be determined based on a false alarm probability determination model. In some embodiments, the false alarm probability determination model can be a machine learning model, for example, a convolutional neural network model. The inputs to the false alarm probability determination model can include system feature vectors, user feature vectors, historical feature vectors, and user reliability (determined based on historical false alarm frequencies, user alarm frequencies, etc.), and the output can include the false alarm probability.

[0116] In some embodiments, the false alarm probability determination model can be obtained through training. For example, training samples are input into the initial false alarm probability determination model, and a loss function is established based on the labels and the output of the initial false alarm probability determination model. The parameters of the initial false alarm probability determination model are updated, and model training is completed when the loss function of the initial false alarm probability determination model meets preset conditions. The preset conditions may include convergence of the loss function, a number of iterations reaching a threshold, etc.

[0117] In some embodiments, the training samples can be multiple historical gas leak causes and corresponding solutions. The training samples can be obtained based on historical data. The labels of the training samples can be monitoring information of gas system false alarms or descriptions of user false alarms. The labels can be manually annotated.

[0118] The embodiments of this specification determine candidate solutions based on vector similarity and false alarm probability, fully consider the reliability of user-reported information, and improve the effectiveness of candidate solutions.

[0119] Step 520: Determine the effective solution rate of the candidate solutions, and determine the target solution from the candidate solutions based on the effective solution rate.

[0120] The effective solution rate refers to the probability that the candidate solution effectively solves the cause of the gas leak. For example, the effective solution rate can be 70%, 80%, etc.

[0121] In some embodiments, the effective solution rate can be determined based on historical data. For example, the smart gas safety management platform 130 can determine the effective solution rate of the candidate solution based on the cause of the gas leak and the corresponding candidate solution in the historical data, as well as whether the solution solves the problem.

[0122] In some embodiments, the effective resolution rate may be determined based on an effective resolution rate determination model.

[0123] The effective resolution rate determination model may be a machine learning model, for example, a convolutional neural network model. The inputs to the effective resolution rate determination model may include the cause of the gas leak and candidate solutions, and the output may include the effective resolution rate.

[0124] In some embodiments, the effective solution rate determination model can be obtained through training based on labeled training samples. For example, the training samples are input into the initial effective solution rate determination model, and a loss function is established based on the labels and the output of the initial effective solution rate determination model. The parameters of the initial effective solution rate determination model are updated. When the loss function of the initial effective solution rate determination model meets a preset condition, the model training is completed, where the preset condition can be that the loss function converges, the number of iterations reaches a threshold, etc.

[0125] In some embodiments, the training samples can be multiple historical gas leak causes and corresponding candidate solutions. The training samples can be obtained based on historical data. The labels of the training samples can be the probability that the candidate solution effectively solves the gas leak cause. The labels can be manually annotated based on the results of the solution's gas leak treatment.

[0126] The embodiments of this specification determine the effective solution rate of candidate solutions based on an effective solution rate determination model, so that the determined effective solution rate is more accurate, thereby better determining the target solution from the candidate solutions.

[0127] The embodiments of this specification determine the effective solution rate of candidate solutions through an effective solution rate determination model, and determine the target solution from the candidate solutions based on the effective solution rate, so that the obtained target solution is more effective, which is conducive to users obtaining a better usage experience.

[0128] In some embodiments, a smart gas terminal linkage disposal device is provided, including a processor, which is used to execute the Internet of Things-based smart gas terminal linkage disposal method in any embodiment of this specification.

[0129] In some embodiments, a computer-readable storage medium is provided, which stores computer instructions. When a computer reads the computer instructions, the computer executes the IoT-based smart gas terminal linkage disposal method in any embodiment of this specification.

[0130] While the basic concepts have been described above, it will be apparent to those skilled in the art that the detailed disclosure is merely illustrative and does not limit this specification. Although not explicitly stated herein, various modifications, improvements, and revisions to this specification may be made by those skilled in the art. Such modifications, improvements, and revisions are suggested in this specification and remain within the spirit and scope of the exemplary embodiments of this specification.

[0131] This specification also uses specific terms to describe the embodiments of this specification. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "one embodiment," "an embodiment," or "an alternative embodiment" two or more times in different locations in this specification do not necessarily refer to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of this specification may be appropriately combined.

[0132] In addition, unless expressly stated in the claims, the order of the processing elements and sequences, the use of alphanumeric characters, or the use of other names described in this specification are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some of the invention embodiments currently considered useful through various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the spirit and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0133] Similarly, it should be noted that, in order to simplify the presentation of this specification and thus facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0134] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A gas leakage intelligent early warning method for smart gas, characterized in that: The method is executed by the smart gas indoor safety management sub-platform and includes: Acquire monitoring information and user description information, wherein the monitoring information includes alarm information and gas terminal monitoring data, and the user description information includes customized alarm information uploaded by the user; Determining the cause of the gas leak based on the monitoring information and the user description information; determining candidate solutions based on the cause of the gas leak; Based on the effective solution rate determination model, the effective solution rate of the candidate solution is determined for the cause of the gas leakage and the candidate solution, and the target solution is determined from the candidate solution according to the effective solution rate, wherein the effective solution rate determination model is a machine learning model.

2. The method according to claim 1, wherein Determining candidate solutions based on the cause of the gas leakage includes: Determine the distance between the feature vector and the reference vector in the database, and determine the reference candidate solution corresponding to the reference vector whose distance meets the preset condition as the candidate solution, wherein, The database includes a plurality of reference vectors and candidate reference solutions corresponding to each reference vector; The feature vector is constructed based on relevant information of the gas terminal in the gas system, and the relevant information includes the monitoring information, the user description information, the cause of the gas leakage and the false alarm probability, and the false alarm probability refers to the possibility of a false alarm in the monitoring information.

3. The method according to claim 1, wherein The determining the cause of the gas leakage based on the monitoring information and the user description information includes: The gas leakage cause is determined by a gas leakage cause prediction model, wherein the input of the gas leakage cause prediction model includes the monitoring information, the user description information and the historical feature vector, and the output includes the gas leakage cause, wherein: The historical feature vector is generated based on historical data, which includes historical alarm information and historical disposal information of the gas system. The historical disposal information refers to record information of users performing maintenance or replacement operations on the gas system components within a historical time period.

4. The method according to claim 3, wherein The historical feature vector is obtained by processing the historical data through an embedding layer.

5. The method according to claim 3, wherein The gas leakage cause prediction model includes a pre-processing layer and a prediction layer, wherein the pre-processing layer includes a first processing layer and a second processing layer; The first processing layer is used to process the monitoring information to obtain a system feature vector; The second processing layer is used to process the user description information to obtain a user feature vector; The prediction layer is used to process the system feature vector, the user feature vector and the historical feature vector to obtain the cause of the gas leakage.

6. The method according to claim 5, wherein The prediction layer includes a first prediction layer and a second prediction layer, wherein: The first prediction layer is used to process the historical feature vector, the system feature vector and the user feature vector to obtain a gas leakage cause vector; The second prediction layer processes the gas leakage cause vector to obtain the gas leakage cause.

7. The method according to claim 1, wherein The monitoring information is obtained by the smart gas data center from the smart gas object platform through the smart gas sensor network platform, and the user description information is obtained by the smart gas data center from the smart gas user platform through the smart gas service platform.

8. The method according to claim 7, wherein Also includes: The smart gas indoor safety management sub-platform uploads the gas leakage cause and the target solution to the smart gas data center; The smart gas data center sends the gas leakage cause and the target solution to the smart gas service platform; The smart gas service platform sends the cause of the gas leakage and the target solution to the smart gas user platform.

9. A gas leak intelligent early warning Internet of Things system for smart gas, characterized in that: The IoT system includes a smart gas user platform, a smart gas service platform, a smart gas safety management platform, a smart gas sensor network platform, and a smart gas object platform. The smart gas safety management platform includes a smart gas indoor safety management sub-platform and a smart gas data center. The smart gas indoor safety management sub-platform is configured to perform the following operations: Acquire monitoring information and user description information, wherein the monitoring information includes alarm information and gas terminal monitoring data, and the user description information includes customized alarm information uploaded by the user; Determining the cause of the gas leak based on the monitoring information and the user description information; determining candidate solutions based on the cause of the gas leak; Based on the effective solution rate determination model, the effective solution rate of the candidate solution is determined for the cause of the gas leakage and the candidate solution, and the target solution is determined from the candidate solution according to the effective solution rate, wherein the effective solution rate determination model is a machine learning model.

10. A gas leakage intelligent early warning device for smart gas, comprising a processor, wherein the processor is used to execute the method according to any one of claims 1 to 8.

Images