Fuel gas safety management method and device, electronic equipment and readable storage medium

Abstract

The embodiment of the invention provides a gas safety management method and device, electronic equipment and a readable storage medium, and the method comprises the steps: carrying out the information recognition based on at least two types of target data collected for a current gas scene through at least two information extraction models, and obtaining scene information of at least two dimensions; the scene information of at least two dimensions is used for monitoring the scene state of the current gas scene; performing scene recognition based on the scene information of at least two dimensions through a scene recognition model, and determining a target scene type corresponding to the current gas scene; and based on the target scene type, determining a safety management scheme for the current gas scene. Thus, through macroscopic and microscopic dual monitoring of the information extraction model and the scene recognition model, the response speed and early warning capability of gas safety management are improved, the risk of safety accidents is reduced, and the management efficiency of gas safety management is improved by using big data and artificial intelligence technologies.

Description

Technical Field

[0001] The present invention relates to the field of gas safety technology, and in particular to a gas safety management method, device, electronic equipment and readable storage medium. Background Art

[0002] With the rapid development of economy and the acceleration of urbanization, gas, as an important energy source for modern urban life, its safety management has become a key link in ensuring public safety and social stability. The frequent occurrence of gas safety accidents not only threatens the safety of people's lives and property, but also poses a potential threat to the development of cities. Therefore, the continuous development and improvement of gas safety management technology has become the primary task of the gas industry and urban management departments.

[0003] In the related technologies, the safety management of the traditional gas industry often relies on manual inspections and experience-based judgments to conduct daily gas scene monitoring, and manually determine the corresponding safety management plan based on the monitoring results. However, this safety management method has the problems of low efficiency and untimely response. Summary of the invention

[0004] In order to overcome the problems existing in the related art, the present invention provides a gas safety management method, device, electronic device and readable storage medium.

[0005] In a first aspect, the present invention provides a gas safety management method, the method comprising:

[0006] By using at least two information extraction models, information recognition is performed based on at least two types of target data collected for the current gas scene to obtain scene information of at least two dimensions; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene;

[0007] Performing scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene;

[0008] Based on the target scenario type, a safety management plan for the current gas scenario is determined.

[0009] Optionally, the method further comprises:

[0010] Acquire historical recognition results of the scene recognition model; the historical recognition results include multiple scene types recognized by the scene recognition model;

[0011] If any first scene type in the historical recognition results satisfies the first optimization condition, the scene recognition model is optimized based on the target factor; the first optimization condition includes that a preset number of scene information corresponding to the first scene type all contain the target factor, and the target factor is a factor other than the preset scene factor used to identify the first scene type.

[0012] Optionally, the optimizing the scene recognition model based on the target factor includes:

[0013] The target factor is added to the preset scene factors corresponding to the first scene type.

[0014] Optionally, the at least two information extraction models include a speech recognition model, a target recognition model, a text recognition model and a data analysis model, and the at least two types of target data include target real-time data, text feedback information and speech feedback information collected by different target devices.

[0015] Optionally, the method further comprises:

[0016] The at least two information extraction models are optimized based on a type recognition evaluation corresponding to the target scene type.

[0017] Optionally, performing scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene includes:

[0018] Acquire a preset scene type and a preset scene factor corresponding to the preset scene type;

[0019] Matching the scene information of the at least two dimensions with the preset scene factors corresponding to each of the preset scene types through the scene recognition model;

[0020] In a case where the scene information of at least two dimensions includes all preset scene factors corresponding to any preset scene type, the preset scene type is determined as the target scene type.

[0021] In a second aspect, the present invention provides a gas safety management device, the device comprising:

[0022] A first recognition module is used to perform information recognition based on at least two types of target data collected for the current gas scene through at least two information extraction models to obtain scene information of at least two dimensions; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene;

[0023] A second recognition module is used to perform scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene;

[0024] The first determination module is used to determine a safety management solution for the current gas scene based on the target scene type.

[0025] Optionally, the device further comprises:

[0026] A first acquisition module is used to acquire historical recognition results of the scene recognition model; the historical recognition results include multiple scene types recognized by the scene recognition model;

[0027] The first optimization module is used to optimize the scene recognition model based on a target factor if any first scene type in the historical recognition results satisfies a first optimization condition; the first optimization condition includes that a preset number of scene information corresponding to the first scene type all contain the target factor, and the target factor is a factor other than the preset scene factor used to identify the first scene type.

[0028] Optionally, the first optimization module includes:

[0029] The first adding module is used to add the target factor to the preset scene factor corresponding to the first scene type.

[0030] Optionally, the at least two information extraction models include a speech recognition model, a target recognition model, a text recognition model and a data analysis model, and the at least two types of target data include target real-time data, text feedback information and speech feedback information collected by different target devices.

[0031] Optionally, the device further comprises:

[0032] The second optimization module is used to optimize the at least two information extraction models based on the type recognition evaluation corresponding to the target scene type.

[0033] Optionally, the second identification module includes:

[0034] A second acquisition module, used to acquire a preset scene type and a preset scene factor corresponding to the preset scene type;

[0035] A first matching module, configured to match the scene information of the at least two dimensions with preset scene factors corresponding to each preset scene type through the scene recognition model;

[0036] The second determination module is configured to determine the preset scene type as the target scene type when the scene information of the at least two dimensions includes all preset scene factors corresponding to any preset scene type.

[0037] In a third aspect, the present invention provides an electronic device comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the gas safety management method described in any one of the first aspects above is implemented.

[0038] In a fourth aspect, the present invention provides a readable storage medium, which, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the steps in the gas safety management method in any one of the embodiments of the first aspect above.

[0039] In the embodiment of the present invention, at least two information extraction models are used to perform information recognition based on at least two types of target data collected for the current gas scene, and scene information of at least two dimensions is obtained; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene; the scene recognition model is used to perform scene recognition based on the scene information of at least two dimensions, and the target scene type corresponding to the current gas scene is determined; based on the target scene type, the safety management plan for the current gas scene is determined. In this way, through the collaborative cooperation of the information extraction model and the scene recognition model, the scene recognition model identifies the target scene type corresponding to the current gas scene to perform macro-level safety analysis and trend prediction, and each information extraction model is responsible for processing local data of different dimensions, performing real-time monitoring and specific risk identification at the micro level, realizing macro- and micro-level dual monitoring, and improving the comprehensiveness and accuracy of gas safety management. At the same time, through the intelligent collaboration of the information extraction model and the scene recognition model, the response speed and early warning capability of gas safety management are improved, the risk of safety accidents is reduced, and the use of big data and artificial intelligence technology reduces the dependence on manual inspections, and improves the management efficiency of gas safety management. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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 paying creative work.

[0041] Figure 1 is a flow chart of steps of a gas safety management method provided by an embodiment of the present invention;

[0042] Figure 2 is a structural diagram of a gas safety management device provided by an embodiment of the present invention;

[0043] Figure 3 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] Figure 1 is a flow chart of the steps of a gas safety management method provided by an embodiment of the present invention, such as Figure 1 As shown, the method may include:

[0046] Step 101: Using at least two information extraction models, information recognition is performed based on at least two types of target data collected for a current gas scene to obtain scene information of at least two dimensions; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene.

[0047] In an embodiment of the present invention, at least two types of target data obtained for the current gas scene are obtained. Among them, the current gas scene may include scenes such as gas production, transportation, distribution, use and monitoring management, such as outdoor pipe network monitoring scenes, plant station gas transportation scenes, indoor gas use scenes, etc. At least two types of target data may include target real-time data, text feedback information and voice feedback information collected by different target devices. Multiple target devices are pre-set in the gas scene that needs to be monitored and safely managed, and the target devices may include data acquisition devices such as sensors and cameras. By installing target devices that realize different monitoring functions at different locations, parameter data in the gas scene is collected in real time as target real-time data. At the same time, for the gas scene, text feedback information and voice feedback information for the gas scene can be obtained, wherein the text feedback information may include text content fed back by business personnel in real time according to the gas scene and text feedback content based on other channels for the gas scene. Voice feedback information may include voice content fed back by business personnel in real time according to the gas scene and voice feedback content based on other channels for the gas scene.

[0048] By at least two information extraction models, the target data is identified respectively to obtain scene information of at least two dimensions. An information extraction model is used to identify a target data and obtain scene information of one dimension. Among them, the information extraction model may include a speech recognition model, a target recognition model, a text recognition model and a data analysis model, and each model may be pre-trained based on historical data. Correspondingly, at least two dimensions may include a speech dimension, an image dimension, a text dimension and a data dimension. Exemplarily, the speech feedback data can be identified based on the speech recognition model to obtain scene information of the speech dimension, for example, the scene information may include speech keywords, semantic understanding content, etc. The target data (such as pictures or videos, etc.) collected by the camera in real time is identified based on the target recognition model to obtain scene information of the image dimension, for example, the scene information may include the category, outline, position, etc. of the target. The text feedback data is identified based on the text recognition model to obtain scene information of the text dimension, for example, the scene information may include text keywords, text understanding content, etc. Based on the data analysis model, the target data (such as pressure, temperature, flow and other parameters) collected by the sensor in real time are identified to obtain the scene information of the data dimension. For example, the scene information may include average values, peak and valley values, outliers, etc.

[0049] At least two dimensions of scene information can be used to monitor the scene status of the current gas scene. Specifically, risk identification can be performed based on at least two dimensions of scene information to determine the risks and hidden dangers in the current gas scene, and the scene status of the current gas scene can be monitored in real time and risks can be identified. Exemplarily, scene information based on data dimensions can determine whether the current gas scene has risks and hidden dangers such as unstable pressure; scene information based on text dimensions and voice dimensions can determine whether the current gas scene has risks and hidden dangers such as pipeline damage; scene information based on image dimensions can determine whether the current gas scene has risks and hidden dangers such as external risks. It can be understood that the method of risk identification based on scene information can be pre-defined according to needs, and the embodiments of the present invention do not limit this.

[0050] Step 102: Perform scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene.

[0051] In an embodiment of the present invention, a scene recognition model is used to perform comprehensive scene recognition based on scene information of at least two dimensions output by at least two information extraction models to determine the target scene type corresponding to the current gas scene. The scene recognition model can be pre-trained based on multiple groups of historical scene features corresponding to multiple groups of scene types, and the scene recognition model can be used to identify the scene type of the gas scene based on the multi-dimensional scene information corresponding to the gas scene. Among them, the scene type can include normal gas scenes such as indoor gas use scenes, plant gas monitoring scenes, etc., and can also include abnormal gas scenes such as gas leakage scenes, pipeline construction scenes, etc.

[0052] The scene information of at least two dimensions is input into the scene recognition model, and the scene recognition model matches the scene information of at least two dimensions with the recognition pattern learned during the training process, thereby identifying the target scene type. It is understandable that the recognition methods of the scene recognition models obtained based on different training methods and different model training may be different, and the embodiments of the present invention do not limit this.

[0053] Each information extraction model can be used to realize information recognition at the perception level, and the scene recognition model is used to understand and analyze information and then identify specific scene types. In other words, the scene recognition model controls the brain to judge the scene type, while specific single tasks (such as information acquisition in different dimensions) are realized by each information extraction model.

[0054] Optionally, step 102 may include the following steps:

[0055] Step 201: Obtain a preset scene type and a preset scene factor corresponding to the preset scene type.

[0056] In an embodiment of the present invention, preset scene types and preset scene factors corresponding to the preset scene types are predefined. The preset scene factors may be key feature factors or unique factors for identifying the preset scene types. The scene recognition model may identify and output a unique preset scene type based on the defined preset scene factors. The preset scene types and the preset scene factors corresponding to each preset scene type are obtained.

[0057] Step 202: Match the scene information of the at least two dimensions with the preset scene factors corresponding to each of the preset scene types through the scene recognition model.

[0058] In the embodiment of the present invention, scene information of at least two dimensions is input into the scene recognition model, and the scene recognition model matches the scene information with preset scene factors corresponding to each preset scene type one by one.

[0059] Step 203: When the scene information of at least two dimensions includes all preset scene factors corresponding to any preset scene type, determine the preset scene type as the target scene type.

[0060] In an embodiment of the present invention, for any preset scene type, in the process of matching scene information with its corresponding preset scene factors, if the scene information of at least two dimensions contains all the preset scene factors corresponding to the preset scene type, that is, the match is successful, then the preset scene type is determined as the target scene type. Exemplarily, assuming that the preset scene factors corresponding to preset scene type A include a, b, and c, when the scene information of at least two dimensions includes a, b, and c, the preset scene type A is determined as the target scene type.

[0061] In a possible implementation, the scene information may include preset scene factors corresponding to at least two preset scene types. In this case, the preset scene type including the largest number of preset scene factors may be used as the target scene type. Assuming that the preset scene factors corresponding to preset scene type A include a, b, and c, the preset scene factors corresponding to preset scene type B include a, b, c, and d, and the preset scene factors corresponding to preset scene type C include a, b, c, d, and f, when the scene information of at least two dimensions includes a, b, c, d, and f, the scene recognition model will determine that the scene information can be successfully matched with preset scene types A, B, and C after matching, and the preset scene type C including the largest number of preset scene factors will be determined as the target scene type.

[0062] In the embodiment of the present invention, the scene recognition model can accurately identify the target scene type corresponding to the current scene by matching scene information with preset scene factors corresponding to each preset scene type, thereby improving the efficiency of scene type recognition.

[0063] Step 103: Based on the target scenario type, determine a safety management solution for the current gas scenario.

[0064] In an embodiment of the present invention, a corresponding safety management scheme is set in advance for each preset scene type. One preset scene type corresponds to one safety management scheme, and different preset scene types can correspond to the same safety management scheme, that is, one safety management scheme can establish a matching relationship with different preset scene types. After determining the target scene type, the safety management scheme for the current gas scene is determined according to the correspondence between the preset scene type and the safety management scheme. Among them, the safety management scheme may include a daily maintenance plan for normal gas scenes and an emergency response plan for abnormal gas scenes. Based on the safety management scheme, the current gas scene can be maintained or managed or problems can be handled.

[0065] In summary, in the embodiment of the present invention, at least two information extraction models are used to perform information recognition based on at least two types of target data collected for the current gas scene, and scene information of at least two dimensions is obtained; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene; the scene recognition model is used to perform scene recognition based on the scene information of at least two dimensions to determine the target scene type corresponding to the current gas scene; based on the target scene type, the safety management plan for the current gas scene is determined. In this way, through the collaborative cooperation of the information extraction model and the scene recognition model, the scene recognition model identifies the target scene type corresponding to the current gas scene to perform macro-level safety analysis and trend prediction, and each information extraction model is responsible for processing local data of different dimensions, performing real-time monitoring and specific risk identification at the micro level, realizing macro- and micro-level dual monitoring, and improving the comprehensiveness and accuracy of gas safety management. At the same time, through the intelligent collaboration of the information extraction model and the scene recognition model, the response speed and early warning capability of gas safety management are improved, the risk of safety accidents is reduced, and the use of big data and artificial intelligence technology reduces the dependence on manual inspections and improves the management efficiency of gas safety management.

[0066] Optionally, the embodiment of the present invention may further include the following steps:

[0067] Step 301: Obtain historical recognition results of the scene recognition model; the historical recognition results include multiple scene types recognized by the scene recognition model.

[0068] In an embodiment of the present invention, after the scene recognition model outputs the target scene type based on the scene information, the scene information of at least two dimensions corresponding to the current recognition process and the target scene type are recorded. The historical recognition results of the scene recognition model are obtained, wherein the historical recognition results may include multiple scene types recognized by the scene recognition model.

[0069] Step 302: If any first scene type in the historical recognition results satisfies a first optimization condition, the scene recognition model is optimized based on a target factor; the first optimization condition includes that a preset number of scene information corresponding to the first scene type all contain a target factor, and the target factor is a factor other than a preset scene factor used to identify the first scene type.

[0070] In an embodiment of the present invention, scene information of at least two dimensions corresponding to each scene type is obtained, and scene information of at least two dimensions corresponding to the same scene type is counted. If there is a preset number of scene information corresponding to a first scene type that contains the same target factor, the target factor is a factor other than the preset scene factor corresponding to the first scene type. In other words, if the same target factor is included multiple times in each scene information corresponding to an identified scene type, it means that the target factor can be used as a scene factor for identifying the scene type. Based on the target factor, the scene recognition model can be optimized so that the scene recognition model uses the target factor as a preset scene factor for the first scene type as the output recognition result in the subsequent recognition process. Among them, the preset number can be set according to demand, and the embodiment of the present invention does not limit this.

[0071] Optionally, step 302 may include the following steps:

[0072] Step 401: Add the target factor to the preset scene factors of the first scene type.

[0073] In an embodiment of the present invention, the target factor is added to the preset scene factors corresponding to the first scene type. Specifically, the target factor can be input into the scene recognition model as training data, so that the scene recognition model learns to use the target factor as a necessary factor for identifying the first scene type.

[0074] For example, assuming that the preset scene factors corresponding to the preset scene type A include a, b, and c, when the output result of the scene recognition model is statistics of the scene information corresponding to the preset scene type A, there are a preset number (for example, 100) of scene information that all contain the target factor d, then the target factor d can be used as the preset scene factor corresponding to the preset scene type A.

[0075] In the embodiment of the present invention, the recognition capability of the scene recognition model is reversely optimized based on the historical recognition results of the scene recognition model, thereby further improving the recognition accuracy of the scene recognition model and reducing the recognition error of the scene recognition model.

[0076] Optionally, the embodiment of the present invention may further include the following steps:

[0077] Step 501: Optimize the at least two information extraction models based on the type recognition evaluation corresponding to the target scene type.

[0078] In an embodiment of the present invention, after the scene recognition model identifies the target scene type, a type recognition evaluation for the target scene type is collected, and the type recognition evaluation is used to evaluate the degree of match and accuracy between the target scene type and the actual scene type of the current gas scene. The type recognition evaluation can be determined based on feedback from business personnel or based on a comparison between the scene type corresponding to the current gas scene determined by other means and the target scene type. Among them, the type recognition evaluation can be divided into evaluation results of different levels, for example, the type recognition evaluation can include excellent, good, and poor. For different type recognition evaluations, at least two information extraction models can be optimized to different degrees. Specifically, the methods of optimizing at least two information extraction models to different degrees may include: increasing data diversity by generating new training samples; collecting and using data content of specific scenarios for retraining for specific scenario types; tuning model parameters, such as optimizing each information extraction model by adjusting hyperparameters such as learning rate, batch size, number of iterations, and the number of layers and neurons in the model architecture; using models pre-trained on large-scale data sets to apply their knowledge to specific tasks through transfer learning, thereby reducing data requirements and improving the performance of each information extraction model; and continuously adjusting model parameters and structures through multiple rounds of training and feedback to improve the extraction capability and accuracy of the model. It is understandable that the method of optimizing each information extraction model can be selected according to demand, and the embodiments of the present invention do not limit this.

[0079] In an embodiment of the present invention, by optimizing the information extraction model through type recognition evaluation of the target scene type output by the scene recognition model, it can be ensured that the macro analysis results of the scene recognition model can guide the model performance of the information extraction model, forming a coordinated optimization of the scene recognition model and the information extraction model, so that the gas safety management system composed of the scene recognition model and the information extraction model can be more intelligent and reliable, and realize the rapid identification, evaluation and early warning of gas safety risks.

[0080] Figure 2 is a schematic diagram of the structure of a gas safety management device provided by an embodiment of the present invention, such as Figure 2 As shown, the device may specifically include:

[0081] The first recognition module 601 is used to perform information recognition based on at least two types of target data collected for the current gas scene through at least two information extraction models to obtain scene information of at least two dimensions; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene;

[0082] A second recognition module 602 is used to perform scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene;

[0083] The first determination module 603 is used to determine a safety management solution for the current gas scene based on the target scene type.

[0084] Optionally, the device further comprises:

[0085] A first acquisition module is used to acquire historical recognition results of the scene recognition model; the historical recognition results include multiple scene types recognized by the scene recognition model;

[0086] The first optimization module is used to optimize the scene recognition model based on a target factor if any first scene type in the historical recognition results satisfies a first optimization condition; the first optimization condition includes that a preset number of scene information corresponding to the first scene type all contain the target factor, and the target factor is a factor other than the preset scene factor used to identify the first scene type.

[0087] Optionally, the first optimization module includes:

[0088] The first adding module is used to add the target factor to the preset scene factor corresponding to the first scene type.

[0089] Optionally, the at least two information extraction models include a speech recognition model, a target recognition model, a text recognition model and a data analysis model, and the at least two types of target data include target real-time data, text feedback information and speech feedback information collected by different target devices.

[0090] Optionally, the device further comprises:

[0091] The second optimization module is used to optimize the at least two information extraction models based on the type recognition evaluation corresponding to the target scene type.

[0092] Optionally, the second identification module 602 includes:

[0093] A second acquisition module, used to acquire a preset scene type and a preset scene factor corresponding to the preset scene type;

[0094] A first matching module, configured to match the scene information of the at least two dimensions with preset scene factors corresponding to each preset scene type through the scene recognition model;

[0095] The second determination module is configured to determine the preset scene type as the target scene type when the scene information of the at least two dimensions includes all preset scene factors corresponding to any preset scene type.

[0096] The present invention also provides an electronic device, see Figure 3 , including: a processor 701, a memory 702, and a computer program 7021 stored in the memory and executable on the processor, wherein the processor implements the gas safety management method of the aforementioned embodiment when executing the program.

[0097] The present invention also provides a readable storage medium, when the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the gas safety management method of the aforementioned embodiment.

[0098] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0099] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific languages ​​is for disclosing the best mode of the present invention.

[0100] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0101] Similarly, it should be understood that in order to streamline the present invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the following intention: that the claimed invention requires more features than the features explicitly recited in each claim. More specifically, as reflected in the claims below, inventive aspects lie in less than all the features of the individual embodiments disclosed above. Therefore, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, with each claim itself serving as a separate embodiment of the present invention.

[0102] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0103] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the sorting device according to the present invention. The present invention may also be implemented as a device or apparatus program for executing part or all of the methods described herein. Such a program for implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0104] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0105] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0106] It should be noted that all actions of acquiring signals, information or data in this application are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0107] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0108] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A gas safety management method, characterized in that: The method comprises: By using at least two information extraction models, information recognition is performed based on at least two types of target data collected for the current gas scene to obtain scene information of at least two dimensions; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene; Performing scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene; Based on the target scenario type, a safety management plan for the current gas scenario is determined.

2. The method according to claim 1, characterized in that The method further comprises: Acquire historical recognition results of the scene recognition model; the historical recognition results include multiple scene types recognized by the scene recognition model; If any first scene type in the historical recognition results satisfies the first optimization condition, the scene recognition model is optimized based on the target factor; the first optimization condition includes that a preset number of scene information corresponding to the first scene type all contain the target factor, and the target factor is a factor other than the preset scene factor used to identify the first scene type.

3. The method according to claim 2, characterized in that The optimizing the scene recognition model based on the target factor includes: The target factor is added to the preset scene factors corresponding to the first scene type.

4. The method according to claim 1, characterized in that: The at least two information extraction models include a speech recognition model, a target recognition model, a text recognition model and a data analysis model, and the at least two types of target data include target real-time data, text feedback information and speech feedback information collected by different target devices.

5. The method according to claim 1, characterized in that: The method further comprises: The at least two information extraction models are optimized based on a type recognition evaluation corresponding to the target scene type.

6. The method according to claim 1, characterized in that The performing scene recognition based on the scene information of the at least two dimensions by using the scene recognition model to determine the target scene type corresponding to the current gas scene includes: Acquire a preset scene type and a preset scene factor corresponding to the preset scene type; Matching the scene information of the at least two dimensions with the preset scene factors corresponding to each of the preset scene types through the scene recognition model; In a case where the scene information of at least two dimensions includes all preset scene factors corresponding to any preset scene type, the preset scene type is determined as the target scene type.

7. A gas safety management device, characterized in that: The device comprises: A first recognition module is used to perform information recognition based on at least two types of target data collected for the current gas scene through at least two information extraction models to obtain scene information of at least two dimensions; the scene information of at least two dimensions is used to monitor the scene state of the current gas scene; A second recognition module is used to perform scene recognition based on the scene information of the at least two dimensions through a scene recognition model to determine a target scene type corresponding to the current gas scene; The first determination module is used to determine a safety management solution for the current gas scene based on the target scene type.

8. The method according to claim 7, characterized in that The device also includes: A first acquisition module is used to acquire historical recognition results of the scene recognition model; the historical recognition results include multiple scene types recognized by the scene recognition model; The first optimization module is used to optimize the scene recognition model based on a target factor if any first scene type in the historical recognition results satisfies a first optimization condition; the first optimization condition includes that a preset number of scene information corresponding to the first scene type all contain the target factor, and the target factor is a factor other than the preset scene factor used to identify the first scene type.

9. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the gas safety management method as described in any one of claims 1 to 6 when executing the program.

10. A readable storage medium, characterized in that: When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the gas safety management method described in any one of claims 1-6.

Images