A method, apparatus, electronic device and storage medium for establishing a monitoring model
By acquiring information on hazardous chemical operation accidents and constructing a monitoring model using multi-source heterogeneous data and Bayesian networks, the accuracy problem of hazardous chemical operation safety monitoring was solved, and efficient management and control of hazardous chemical operations was achieved.
Patent Information
- Application Number
- CN202211419090.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-11-14
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2042-11-14
AI Technical Summary
How to effectively monitor and control hazardous chemical operations in order to reduce the occurrence of safety accidents and improve the safety and reliability of hazardous chemical operations.
By acquiring information on hazardous chemical operation accidents, using multi-source heterogeneous data to determine monitoring and early warning indicators and model parameters, constructing a Bayesian network, generating a target model, and combining a TF-IDF model and an expert database, extracting feature keywords and generating a set of feature keywords, and then conducting abnormal behavior detection.
It improved the accuracy of the hazardous chemical operation monitoring model, enhanced the efficiency of the hazardous chemical operation control process, and reduced management costs.
Smart Images

Figure CN116128689B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of monitoring model building technology, and in particular to a monitoring model building method, a monitoring model building device, an electronic device, and a computer-readable storage medium. Background Technology
[0002] Hazardous chemical operations, or simply hazardous operations, combine the characteristics of both the chemical industry and the construction industry. Safety accidents in the hazardous chemical industry result in extremely large losses. Therefore, only by continuously exploring and researching scientific methods for the management and control of hazardous chemical operations and effectively improving the level of management and control can we fundamentally achieve safe production.
[0003] Therefore, how to manage hazardous chemical operations in order to reduce and prevent unsafe acts and unsafe conditions of objects, and reduce the possibility of production safety accidents, is a technical problem that those skilled in the art need to overcome. Summary of the Invention
[0004] The present invention provides a method, apparatus, electronic device, and computer-readable storage medium for establishing a monitoring model to solve the problem of how to establish a monitoring model for hazardous chemical operations.
[0005] This invention discloses a method for establishing a monitoring model, which is applied to an intelligent management and control system for hazardous chemical special operations. The method may include: Acquire information on hazardous chemical operation accidents; the hazardous chemical operation accident information includes multi-source heterogeneous data; Based on the hazardous chemical operation accident information, monitoring and early warning indicators are determined for the hazardous chemical operation accident information; The model parameters are determined using the aforementioned multi-source heterogeneous data; A target model is generated based on the model parameters and the monitoring and early warning indicators.
[0006] Optionally, the intelligent management and control system for hazardous chemical special operations has a corresponding TF-IDF model and a corresponding expert database. The step of determining the monitoring and early warning indicators for the hazardous chemical operation accident information based on the hazardous chemical operation accident information may include: The TF-IDF model was used to extract feature keywords for the hazardous chemical operation accident information; A set of feature keywords is generated using the aforementioned feature keywords; Monitoring and early warning indicators are determined from the set of feature keywords using the expert database and the set of feature keywords.
[0007] Optionally, the multi-source heterogeneous data includes location data and video image data, and the method may further include: The location data is used to generate on-duty status data for the workers; The video image data is used to generate behavioral risk factors for the workers.
[0008] Optionally, the monitoring and early warning indicators include risk factor data and first abnormal behavior data, and the step of generating a target model based on the model parameters and the monitoring and early warning indicators may include: Construct a risk factor data set for the aforementioned risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in the Bayesian network; Construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network; The prior probability of an action occurring for the risk factor dataset is calculated based on the expert database using the risk factor nodes and the abnormal behavior nodes. A Bayesian network is trained using the prior probability of the aforementioned behavior, and a conditional probability table for each node in the Bayesian network is determined. Generate a first subset of abnormal behaviors; the first subset of abnormal behaviors is an empty set; A second subset of abnormal behaviors is determined from the monitoring and early warning indicators, the second subset of abnormal behaviors including second abnormal behavior data; An abnormal behavior threshold is generated by performing a threshold evaluation on the second abnormal behavior data, and an abnormal behavior threshold set is constructed based on the abnormal behavior threshold. A risk factor set is generated using the set of abnormal behavior thresholds; the risk factor set is an empty set. The second abnormal behavior data corresponding to the abnormal behavior threshold is determined from the abnormal behavior threshold set in a preset order; A third subset of abnormal behavior data is generated using the second abnormal behavior data and the monitoring and early warning indicators; The target risk factor set is determined by the risk factor set and the third abnormal behavior data subset; The first subset of abnormal behaviors is determined by the first subset of abnormal behaviors and the second subset of abnormal behavior data; When the target risk factor set meets the preset conditions, the first abnormal behavior subset of the target is taken as the final abnormal behavior subset, and the target model is generated using the final abnormal behavior subset.
[0009] Optionally, it may also include: The target model is encapsulated into an application; the application includes an application programming interface for the intelligent management and control system for hazardous chemical special operations.
[0010] This invention also discloses a monitoring model establishment device, which is applied to an intelligent management and control system for hazardous chemical special operations. The device includes: The operation accident information acquisition module is used to acquire hazardous chemical operation accident information; the hazardous chemical operation accident information includes multi-source heterogeneous data; The monitoring and early warning indicator determination module is used to determine monitoring and early warning indicators for the hazardous chemical operation accident information based on the hazardous chemical operation accident information. The model parameter determination module is used to determine model parameters using the multi-source heterogeneous data. The target model generation module is used to generate a target model based on the model parameters and the monitoring and early warning indicators.
[0011] Optionally, the intelligent management and control system for special hazardous chemical operations has a corresponding TF-IDF model, the intelligent management and control system for special hazardous chemical operations has a corresponding expert database, and the monitoring and early warning indicator determination module may include: The feature keyword extraction submodule is used to extract feature keywords for the hazardous chemical operation accident information using the TF-IDF model. A set generation submodule is used to generate a set of feature keywords using the feature keywords; The monitoring and early warning indicator determination submodule is used to determine the monitoring and early warning indicators from the feature keyword set through the expert database and the feature keyword set.
[0012] Optionally, the multi-source heterogeneous data includes positioning data and video image data, and the device may further include: The on-duty status data generation submodule is used to generate on-duty status data for operators based on the location data. The behavioral risk factor generation submodule is used to generate behavioral risk factors for workers based on the video image data.
[0013] Optionally, the monitoring and early warning indicators include risk factor data and first abnormal behavior data, and the target model generation module may include: The risk factor data set construction submodule is used to construct a risk factor data set for the risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in the Bayesian network. An abnormal behavior data set construction submodule is used to construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network. The prior probability calculation submodule is used to calculate the prior probability of an action occurring for the risk factor data set based on the expert database using the risk factor node and the abnormal behavior node. The conditional probability table determination submodule is used to train a Bayesian network using the prior probability of the occurrence of the behavior, and to determine the conditional probability table of each node in the Bayesian network. The first abnormal behavior subset generation submodule is used to generate a first abnormal behavior subset; the first abnormal behavior subset is an empty set. The second abnormal behavior subset generation submodule is used to determine a second abnormal behavior subset from the monitoring and early warning indicators, wherein the second abnormal behavior subset includes second abnormal behavior data; The abnormal behavior threshold set construction submodule is used to generate abnormal behavior thresholds by performing threshold evaluation on the second abnormal behavior data, and to construct an abnormal behavior threshold set based on the abnormal behavior thresholds. The risk factor set generation submodule is used to generate a risk factor set from the abnormal behavior threshold set; the risk factor set is an empty set. The second abnormal behavior data determination submodule is used to determine the second abnormal behavior data corresponding to the abnormal behavior threshold from the abnormal behavior threshold set in a preset order; The third abnormal behavior data subset generation submodule is used to generate a third abnormal behavior data subset using the second abnormal behavior data and the monitoring and early warning indicators; The target risk factor set determination submodule is used to determine the target risk factor set through the risk factor set and the third abnormal behavior data subset; The first abnormal behavior subset determination submodule is used to determine the first abnormal behavior subset of the target through the first abnormal behavior subset and the second abnormal behavior data; The target model generation submodule is used to generate a target model by taking the first abnormal behavior subset of the target as the final abnormal behavior subset when the target risk factor set meets the preset conditions.
[0014] Optionally, it may also include: An application encapsulation submodule is used to encapsulate the target model into an application; the application includes an application programming interface for the intelligent management and control system for hazardous chemical special operations.
[0015] This invention also discloses an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes a program stored in the memory, it implements the method described in the embodiments of the present invention.
[0016] This invention also discloses a computer-readable storage medium storing instructions that, when executed by one or more processors, cause the processors to perform the methods described in this invention.
[0017] The embodiments of the present invention have the following advantages: This invention, in its embodiments, acquires hazardous chemical operation accident information, including multi-source heterogeneous data; determines monitoring and early warning indicators based on the hazardous chemical operation accident information; determines model parameters using the multi-source heterogeneous data; and generates a target model based on the model parameters and the monitoring and early warning indicators. This improves the accuracy of hazardous chemical operation monitoring model establishment, thereby enhancing the efficiency of hazardous chemical operation control and reducing management costs. Attached Figure Description
[0018] Figure 1 This is a flowchart illustrating the steps of a monitoring model establishment method provided in this embodiment of the invention; Figure 2 This is a structural block diagram of a monitoring model establishment device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a monitoring model establishment method provided in an embodiment of the present invention; Figure 4 This is a hardware structure block diagram of an electronic device provided in various embodiments of the present invention. Detailed Implementation
[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0020] To enable those skilled in the art to better understand the embodiments of the present invention, the technical terms involved in the embodiments of the invention are described below.
[0021] k-fold cross validation: A cross-validation method that divides the training set into K parts, uses K-1 parts for training, and uses the remaining part for testing. Finally, the average test error is taken as the generalization error.
[0022] TF-IDF model (term frequency-inverse document frequency): a commonly used weighting technique for information retrieval and data mining. TF stands for Term Frequency, and IDF stands for Inverse Document Frequency.
[0023] The embodiments of the present invention provide a monitoring model establishment method, apparatus, electronic device, and computer-readable storage medium, which combine hazardous chemical operation accident information to establish a monitoring model, thereby improving the accuracy of monitoring illegal and abnormal operations, and thus enhancing the safety and reliability of hazardous chemical operation processes.
[0024] Reference Figure 1 The diagram illustrates a flowchart of a monitoring model establishment method provided in an embodiment of the present invention, which may specifically include the following steps: Step 101: Obtain hazardous chemical operation accident information; the hazardous chemical operation accident information includes multi-source heterogeneous data; Step 102: Determine monitoring and early warning indicators for the hazardous chemical operation accident information based on the hazardous chemical operation accident information; Step 103: Determine model parameters using the multi-source heterogeneous data; Step 104: Generate a target model based on the model parameters and the monitoring and early warning indicators.
[0025] In specific implementations, embodiments of the present invention can be applied to an intelligent management and control system for hazardous chemical special operations. This system can be a dedicated intelligent management and control system for monitoring hazardous chemical special operations. Embodiments of the present invention can acquire hazardous chemical operation accident information, which may include data such as accident cause, accident type, time of occurrence, types of hazardous chemicals involved in the accident, and information on relevant personnel. The multi-source heterogeneous data may include monitoring data, location data, reporting data, and accident data related to hazardous chemical operations.
[0026] Of course, the above examples are merely illustrative, and those skilled in the art can use other types of data as information on hazardous chemical operation accidents. In this regard, the embodiments of the present invention do not impose any limitations.
[0027] In practical applications, targeted technical means can be adopted to acquire different types of data. For example, for location data, information technology can be used to manage the process traces of the entire operation. For application data, real-time personnel positioning technology can be used to obtain data on whether the signatory is on-site, and the on-duty status data of the operator and the supervisor can also be obtained during the operation. For video image data such as surveillance footage, optical flow and foreground target contour methods can be used to extract and describe the low-level feature information of the video image, and risk factors at different levels can be defined and described based on ontology semantic technology and domain knowledge models.
[0028] In specific implementations, embodiments of the present invention can determine monitoring and early warning indicators for hazardous chemical operation accident information based on such accident information. For example, if the hazardous chemical operation accident information includes data on types of violations related to oil depots, such as "carrying a source of ignition into the depot" and "non-warehouse personnel entering the depot," then "carrying a source of ignition into the depot" and "non-warehouse personnel entering the depot" can be used as monitoring and early warning indicators. Of course, the above examples are merely illustrative, and those skilled in the art can use other types of data as monitoring and early warning indicators; therefore, embodiments of the present invention do not impose limitations on this.
[0029] In specific implementations, embodiments of the present invention can use the multi-source heterogeneous data to determine model parameters. For example, the multi-source heterogeneous data may include the number of workers on duty and their working hours at the "oil depot," where the number of workers on duty is "20" and the working hours are "9:00-12:00." In this case, "oil depot," "20 workers," and "9:00-12:00" can be used as model parameters. Of course, the above example is only an illustration, and those skilled in the art can use other types of data as model parameters. In this regard, embodiments of the present invention do not impose any limitations.
[0030] In specific implementations, embodiments of the present invention can generate a target model based on model parameters and monitoring and early warning indicators. For example, when the model parameters are "oil depot", "20 people", and "9:00-12:00", and the monitoring and early warning indicators are "carrying fire sources into the depot" and "non-warehouse personnel entering the warehouse", the target model can be established when "location: oil depot_time: 9:00-12:00_number of workers: 20_early warning conditions: carrying fire sources into the depot_non-warehouse personnel entering the warehouse" is identified. Of course, the above example is only an example, and those skilled in the art can use other data as model parameters and monitoring and early warning indicators to establish a target model. In this regard, embodiments of the present invention do not impose any restrictions.
[0031] This invention, in its embodiments, acquires hazardous chemical operation accident information, including multi-source heterogeneous data; determines monitoring and early warning indicators based on the hazardous chemical operation accident information; determines model parameters using the multi-source heterogeneous data; and generates a target model based on the model parameters and the monitoring and early warning indicators. This improves the accuracy of hazardous chemical operation monitoring model establishment, thereby enhancing the efficiency of hazardous chemical operation control and reducing management costs.
[0032] Based on the above embodiments, modified embodiments of the above embodiments are proposed. It should be noted that, in order to keep the description brief, only the differences from the above embodiments are described in the modified embodiments.
[0033] In an optional embodiment of the present invention, the intelligent management and control system for hazardous chemical special operations has a corresponding TF-IDF model, the intelligent management and control system for hazardous chemical special operations has a corresponding expert database, and the step of determining the monitoring and early warning indicators for the hazardous chemical operation accident information through the hazardous chemical operation accident information includes: The TF-IDF model was used to extract feature keywords for the hazardous chemical operation accident information; A set of feature keywords is generated using the aforementioned feature keywords; Monitoring and early warning indicators are determined from the set of feature keywords using the expert database and the set of feature keywords.
[0034] In practical applications, the importance of a word increases proportionally to its frequency of occurrence in a document, but decreases inversely proportionally to its frequency of occurrence in a corpus. Therefore, if a word has a high frequency of occurrence in one article but rarely appears in other articles, it is considered to have good category differentiation ability and is suitable for classification. In hazardous chemical operations, to improve the accuracy of monitoring, data with greater discriminative power should be selected as distinguishing markers. This invention can use a TF-IDF model to extract feature keywords for hazardous chemical operation accident information, and use these feature keywords to generate a feature keyword set. Through the expert database and the feature keyword set, monitoring and early warning indicators are determined from the feature keyword set.
[0035] For example, if hazardous chemical operation accident information includes "fire accident" and "explosion accident", and the TF-IDF scores of "fire accident" and "explosion accident" are greater than 0.1, then "fire accident" and "explosion accident" can be used as feature keywords, and a feature keyword set "fire accident_explosion accident" can be generated. Of course, the above example is only illustrative, and those skilled in the art can use any TF-IDF value and / or uniform setting rules. In this regard, the embodiments of the present invention do not impose any limitations.
[0036] For example, when the set of feature keywords is "fire accident_explosion accident", and the expert database contains data on "fire accident" and "explosion accident" including "high temperature spontaneous combustion" and "physical impact", then the monitoring and early warning indicator can be determined as "fire accident_explosion accident_high temperature spontaneous combustion_physical impact". Of course, the above example is only an example, and those skilled in the art can use any data as feature keywords. In this regard, the embodiments of the present invention do not limit it.
[0037] For example, information related to accidents in special operational processes was collected and organized, and statistical analysis was conducted on accidents related to special operational processes in the hazardous chemicals industry. Based on the TF-IDF model, data mining and analysis were performed to extract feature keywords, resulting in a set of feature keywords. .
[0038] This invention employs a TF-IDF model to extract feature keywords for hazardous chemical operation accident information; generates a feature keyword set using the feature keywords; and determines monitoring and early warning indicators from the feature keyword set using the expert database and the feature keyword set, thereby making the feature keywords more discriminative and improving the accuracy of monitoring illegal and abnormal operations.
[0039] In practical applications, cross-validation is usually used to select the most suitable data object from different data objects. Among them, k-fold cross-validation is a validation method that divides the data into k parts, uses k-1 parts as training data and 1 part as validation data in turn to conduct experiments, and finally obtains the average ability of each data object and selects the best one from them.
[0040] In an optional embodiment of the present invention, the multi-source heterogeneous data includes positioning data and video image data, and the method further includes: The location data is used to generate on-duty status data for the workers; The video image data is used to generate behavioral risk factors for the workers.
[0041] In specific implementations, embodiments of the present invention can generate on-duty status data for operators using location data. For example, if the location data of "Oil Depot Operator 1" during the time period of "9:00-12:00" is "Oil Depot", then the on-duty status of "Oil Depot Operator 1" during the time period of "9:00-12:00" can be generated as "On Duty". Embodiments of the present invention can also generate behavioral risk factors for operators using video image data. For example, if the monitoring collects the operation video data of "Oil Depot Operator 1" and obtains that "Oil Depot Operator 1" has the behavior of carrying a fire source into the depot, then the behavioral risk factor "carrying a fire source into the depot" can be generated for "Oil Depot Operator 1".
[0042] This invention generates on-duty status data for workers through the location data and generates behavioral risk factors for workers through the video image data, thereby obtaining targeted worker data. This enables accurate judgment of workers' status and behavior, improving the accuracy and timeliness of monitoring violations and abnormal operations.
[0043] In an optional embodiment of the present invention, the monitoring and early warning indicators include risk factor data and first abnormal behavior data, and the step of generating a target model based on the model parameters and the monitoring and early warning indicators includes: Construct a risk factor data set for the aforementioned risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in the Bayesian network; Construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network; The prior probability of an action occurring for the risk factor dataset is calculated based on the expert database using the risk factor nodes and the abnormal behavior nodes. A Bayesian network is trained using the prior probability of the aforementioned behavior, and a conditional probability table for each node in the Bayesian network is determined. Generate a first subset of abnormal behaviors; the first subset of abnormal behaviors is an empty set; A second subset of abnormal behaviors is determined from the monitoring and early warning indicators, the second subset of abnormal behaviors including second abnormal behavior data; An abnormal behavior threshold is generated by performing a threshold evaluation on the second abnormal behavior data, and an abnormal behavior threshold set is constructed based on the abnormal behavior threshold. A risk factor set is generated using the set of abnormal behavior thresholds; the risk factor set is an empty set. The second abnormal behavior data corresponding to the abnormal behavior threshold is determined from the abnormal behavior threshold set in a preset order; A third subset of abnormal behavior data is generated using the second abnormal behavior data and the monitoring and early warning indicators; The target risk factor set is determined by the risk factor set and the third abnormal behavior data subset; The first subset of abnormal behaviors is determined by the first subset of abnormal behaviors and the second subset of abnormal behavior data; When the target risk factor set meets the preset conditions, the first abnormal behavior subset of the target is taken as the final abnormal behavior subset, and the target model is generated using the final abnormal behavior subset.
[0044] In a specific implementation, embodiments of the present invention can construct a risk factor data set for risk factor data. For example, if the risk factor data includes "fire" and "lost goods", then a risk factor set can be constructed as "fire" and "lost goods", and a risk factor node "fire_lost goods" with a one-to-one correspondence with "fire" and "lost goods" can be constructed in the Bayesian network.
[0045] In a specific implementation, embodiments of the present invention can construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network; for example, if the abnormal behavior data includes "carrying a fire source into the warehouse" and "non-warehouse personnel entering", then the abnormal behavior data set can be constructed as "carrying a fire source into the warehouse" and "non-warehouse personnel entering", and there is an abnormal behavior node "carrying a fire source into the warehouse_non-warehouse personnel entering" in the Bayesian network that corresponds one-to-one with "carrying a fire source into the warehouse" and "non-warehouse personnel entering".
[0046] In a specific implementation, embodiments of the present invention can calculate the prior probability of an action occurring for a risk factor dataset based on an expert database using risk factor nodes and abnormal behavior nodes. A Bayesian network is then trained using these prior probabilities, and a conditional probability table for each node in the Bayesian network is determined. For example, when the risk factor node is "fire_lost goods" and the abnormal behavior node is "carrying a source of ignition into the warehouse_non-warehouse personnel entering," the probabilities of the abnormal behavior nodes "carrying a source of ignition into the warehouse" and "non-warehouse personnel entering" causing "fire" and "lost goods" in the risk factor node can be calculated based on the expert database. The calculated probability of "carrying a source of ignition into the warehouse" causing a fire is... Given a 30% probability of carrying a source of ignition into the warehouse leading to the loss of goods, a 0% probability of non-warehouse personnel entering the warehouse causing a fire, a 5% probability of this leading to the loss of goods, and a 10% probability of this leading to the loss of goods, the prior probabilities of the abnormal behavior "carrying a source of ignition into the warehouse" and "non-warehouse personnel entering the warehouse" for the risk factor of fire are 30% and 5%, respectively. For the risk factor of goods loss, the prior probabilities of "carrying a source of ignition into the warehouse" and "non-warehouse personnel entering the warehouse" are 0% and 10%, respectively. The structure of the diagnostic Bayesian network can be determined through Bayesian network structure learning based on training samples. Background knowledge is pre-edited on the constructed network to improve modeling efficiency. The order of variables is determined by cause first, then effect, establishing a causal network. The Bayesian network is then trained, and calculations are performed on other risk factor nodes and other abnormal behavior nodes. The prior probabilities of each node are used as the conditional probability table for each node in the Bayesian network. Of course, the above examples are merely illustrative, and those skilled in the art can use other methods to calculate prior probability data. In this regard, the embodiments of the present invention do not impose any limitations.
[0047] In a specific implementation, embodiments of the present invention may include a step of generating a target model based on model parameters and monitoring and early warning indicators. For example, when the monitoring and early warning indicator is... At that time, the first subset of abnormal behaviors is generated as follows: , Initially Based on monitoring and early warning indicators, abnormal behavior nodes corresponding to risk factor nodes in the Bayesian network are identified, and the abnormal behavior data corresponding one-to-one with the abnormal behavior nodes are used as the second abnormal behavior data. And generate a subset of second abnormal behaviors based on the second abnormal behavior data. ,right The second abnormal behavior data Perform threshold evaluation to generate abnormal behavior thresholds. ), and will Add to collection Construct a set of abnormal behavior thresholds, and then, based on the set of abnormal behavior thresholds... Generate a set of risk factors and initialize make ; from the abnormal behavior threshold set in descending order Obtaining abnormal behavior thresholds ), and based on Determine the corresponding second abnormal behavior data. , obtain In the current monitoring and early warning indicators The corresponding set of risk factors And satisfy and will As the third subset of abnormal behavior data, if right If it contributes to growth, then ,and , will the new Define it as the target risk factor set, and let And determine the first subset of abnormal behaviors of the target as Then, determine the set of target risk factors at this time. Does it meet the following conditions:
[0048] If the target risk factor set If the conditions are met, then the first subset of abnormal behaviors of the target will be defined as follows: The target model is generated from the final subset of abnormal behaviors.
[0049] If the target risk factor set If the condition is not met, repeat the process from the abnormal behavior threshold set in descending order. Obtaining abnormal behavior thresholds ), and based on Determine the corresponding second abnormal behavior data. , obtain In the current monitoring and early warning indicators The corresponding set of risk factors And satisfy and will As the third subset of abnormal behavior data, if right If it contributes to growth, then ,and , will the new Define it as the target risk factor set, and let And determine the first subset of abnormal behaviors of the target as Then, determine the set of target risk factors at this time. Check if the conditions are met, until they are met.
[0050] This invention embodiment constructs a risk factor data set for the risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in a Bayesian network; constructs an abnormal behavior data set for the first abnormal behavior data; the second element in the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in a Bayesian network; calculates the prior probability of behavior occurrence for the risk factor data set based on the expert database using the risk factor nodes and the abnormal behavior nodes; trains the Bayesian network using the prior probability of behavior occurrence and determines the conditional probability table for each node in the Bayesian network; generates a first abnormal behavior subset; the first abnormal behavior subset is an empty set; determines a second abnormal behavior subset from the monitoring and early warning indicators, the second abnormal behavior subset including second abnormal behavior data; and generates an abnormal behavior threshold by performing threshold evaluation on the second abnormal behavior data. The system calculates the abnormal behavior threshold value and constructs an abnormal behavior threshold set based on the abnormal behavior threshold. A risk factor set is generated from the abnormal behavior threshold set; the risk factor set is an empty set. Second abnormal behavior data corresponding to the abnormal behavior threshold is determined from the abnormal behavior threshold set in a preset order. A third abnormal behavior data subset is generated using the second abnormal behavior data and the monitoring and early warning indicators. A target risk factor set is determined using the risk factor set and the third abnormal behavior data subset. A target first abnormal behavior subset is determined using the first abnormal behavior subset and the second abnormal behavior data. When the target risk factor set meets preset conditions, the target first abnormal behavior subset is used as the final abnormal behavior subset, and the target model is generated using the final abnormal behavior subset. This accurately determines the correspondence between abnormal behavior and risk factors, further improving the accuracy of the hazardous chemical operation monitoring model.
[0051] In an optional embodiment of the present invention, it further includes: In practical applications, embodiments of the present invention can encapsulate the target model into an application program; the application program includes an application programming interface for the intelligent management and control system for hazardous chemical special operations.
[0052] In practical implementation, by encapsulating the target model into an application and providing an application programming interface (API) for the intelligent management and control system for hazardous chemical special operations, it is easier for users to use the system to call the target model based on the API, thereby improving the convenience of monitoring and operating hazardous chemical operations.
[0053] To enable those skilled in the art to better understand the embodiments of the present invention, a complete example is used below to illustrate the embodiments of the present invention.
[0054] refer to Figure 3 , Figure 3 This is a schematic diagram of a monitoring model establishment method provided in this embodiment of the invention. The invention can classify events based on the city's operational IOC center using topic distribution and text classification, and establish a hazardous chemical operation monitoring model. For example, based on collecting and organizing information related to accidents in special operation stages, statistical analysis was performed on accidents related to special operation stages in the hazardous chemical industry. Based on the TF-IDF model, mining and analysis were performed to extract feature keywords, resulting in a feature keyword set. Furthermore, through thorough communication and exchange with specialists and domain experts, and utilizing human factor analysis and classification methods, from... Preliminary screening of risk factors affecting operational safety and key contents This study establishes a monitoring and early warning indicator system for operational safety risks in hazardous chemical industrial parks and enterprises. It extracts leading events from operational activities that reflect the status of operational safety risk factors within the indicator system. Unexpected or abnormal events during these events are used as monitoring and early warning indicators. These events should be identifiable and measurable. Furthermore, it organizes, integrates, and aggregates multi-source heterogeneous data related to hazardous chemical special operations, primarily involving monitoring, location, reporting, and accident data. This data is mapped to a unified conceptual space. Targeted technical methods are employed to acquire different types of data involved in hazardous chemical special operations. For location data, information technology is used to manage the entire operational process. By using real-time personnel positioning technology, data on whether the signatory is on-site is obtained. During operations, the on-duty status data of the operator and the supervisor is recorded. For video image data such as surveillance footage, optical flow and foreground target contour methods are used to extract and describe the low-level feature information of the video image, such as color or grayscale. Based on ontology semantic technology and domain knowledge models, different levels of risk factors are defined and described. Then, cross-validation is used to select the most suitable parameters for the model, making the model's performance and generalization ability stronger. Next, a monitoring model for abnormal behavior in hazardous chemical special operations is constructed based on Bayesian networks. The purpose of Bayesian network learning is to obtain a Bayesian network model that best fits the observed data based on existing prior knowledge and observation data. Network nodes are selected according to the principles of representativeness, independence, and effectiveness, and the node states are defined discretely. The specific construction steps are as follows: Based on... Determine if it contains A set of safety risk factors for individual operations , Each risk factor in the network corresponds to a risk factor node in the Bayesian network, and the nodes containing the risk factor node are determined. A collection of abnormal behaviors of elements , Each behavior in the process corresponds to a behavior node in a Bayesian network. Then, by combining expert prior knowledge with historical data, the set of security risk factors is analyzed. The prior probability of each behavior occurring in Then, based on the training samples, Bayesian network structure learning is performed to determine the structure of the diagnostic Bayesian network. Background knowledge is pre-edited on the constructed network to improve modeling efficiency. The order of variables is determined according to cause first, then effect, to establish a causal network. Simultaneously, Bayesian network parameter learning is performed based on the training samples to determine the conditional probability table (CPT) for each node in the diagnostic Bayesian network. Next, the trained Bayesian network model is used for abnormal behavior detection. First, a subset of abnormal behaviors is set. , Initially Then identify the set of safety risk factors for the current operation. The set of all abnormal behaviors corresponding to the factors in. ,right Abnormal behavior Threshold evaluation ), and will Add to collection Initialize the set of safety risk factors for the operation. From in descending order Extract elements from the list and obtain the corresponding abnormal behavior. and obtain The set of risk factors corresponding to the current set of detected risk factors. ,satisfy ,if right If it contributes to growth, then ,and Repeat the above steps until the following condition is met:
[0055] When the condition is met, let And output the final subset of abnormal behaviors. Furthermore, the model is iterated and optimized based on the detection results. At the same time, based on the implementation of abnormal behavior detection, the detection model and functional modules are further encapsulated into an application system, and an API interface is provided.
[0056] The intelligent control system established in this invention fills the gap in the detection of abnormal behavior in hazardous chemical special operations, providing a detection model based on machine learning and data mining, and meeting the needs of hazardous chemical parks and enterprises for operational safety control. Compared with existing similar technologies, this invention integrates and aggregates multi-source heterogeneous data in hazardous chemical special operations, and adopts targeted methods for processing different types of data; it establishes an operational safety risk monitoring and early warning indicator system for hazardous chemical parks and enterprises to guide and support the detection of abnormal behavior in hazardous chemical special operations; simultaneously, by constructing and training a Bayesian network that simultaneously includes risk factor nodes and abnormal behavior nodes, this invention can detect abnormal behavior in hazardous chemical special operations based on prior knowledge and observation data.
[0057] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.
[0058] Reference Figure 2 The diagram shows a structural block diagram of a monitoring model establishment device provided in an embodiment of the present invention, which may specifically include the following modules: The operation accident information acquisition module 201 is used to acquire hazardous chemical operation accident information; the hazardous chemical operation accident information includes multi-source heterogeneous data. The monitoring and early warning indicator determination module 202 is used to determine monitoring and early warning indicators for the hazardous chemical operation accident information based on the hazardous chemical operation accident information. The model parameter determination module 203 is used to determine model parameters using the multi-source heterogeneous data; The target model generation module 204 is used to generate a target model based on the model parameters and the monitoring and early warning indicators.
[0059] Optionally, the intelligent management and control system for special hazardous chemical operations has a corresponding TF-IDF model, the intelligent management and control system for special hazardous chemical operations has a corresponding expert database, and the monitoring and early warning indicator determination module may include: The feature keyword extraction submodule is used to extract feature keywords for the hazardous chemical operation accident information using the TF-IDF model. A set generation submodule is used to generate a set of feature keywords using the feature keywords; The monitoring and early warning indicator determination submodule is used to determine the monitoring and early warning indicators from the feature keyword set through the expert database and the feature keyword set.
[0060] Optionally, the multi-source heterogeneous data includes positioning data and video image data, and the device may further include: The on-duty status data generation submodule is used to generate on-duty status data for operators based on the location data. The behavioral risk factor generation submodule is used to generate behavioral risk factors for workers based on the video image data.
[0061] Optionally, the monitoring and early warning indicators include risk factor data and first abnormal behavior data, and the target model generation module may include: The risk factor data set construction submodule is used to construct a risk factor data set for the risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in the Bayesian network. An abnormal behavior data set construction submodule is used to construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network. The prior probability calculation submodule is used to calculate the prior probability of an action occurring for the risk factor data set based on the expert database using the risk factor node and the abnormal behavior node. The conditional probability table determination submodule is used to train a Bayesian network using the prior probability of the occurrence of the behavior, and to determine the conditional probability table of each node in the Bayesian network. The first abnormal behavior subset generation submodule is used to generate a first abnormal behavior subset; the first abnormal behavior subset is an empty set. The second abnormal behavior subset generation submodule is used to determine a second abnormal behavior subset from the monitoring and early warning indicators, wherein the second abnormal behavior subset includes second abnormal behavior data; The abnormal behavior threshold set construction submodule is used to generate abnormal behavior thresholds by performing threshold evaluation on the second abnormal behavior data, and to construct an abnormal behavior threshold set based on the abnormal behavior thresholds. The risk factor set generation submodule is used to generate a risk factor set from the abnormal behavior threshold set; the risk factor set is an empty set. The second abnormal behavior data determination submodule is used to determine the second abnormal behavior data corresponding to the abnormal behavior threshold from the abnormal behavior threshold set in a preset order; The third abnormal behavior data subset generation submodule is used to generate a third abnormal behavior data subset using the second abnormal behavior data and the monitoring and early warning indicators; The target risk factor set determination submodule is used to determine the target risk factor set through the risk factor set and the third abnormal behavior data subset; The first abnormal behavior subset determination submodule is used to determine the first abnormal behavior subset of the target through the first abnormal behavior subset and the second abnormal behavior data; The target model generation submodule is used to generate a target model by taking the first abnormal behavior subset of the target as the final abnormal behavior subset when the target risk factor set meets the preset conditions.
[0062] Optionally, it may also include: An application encapsulation submodule is used to encapsulate the target model into an application; the application includes an application programming interface for the intelligent management and control system for hazardous chemical special operations.
[0063] As the apparatus embodiment is basically similar to the method embodiment, it is described in a relatively simple manner. For relevant details, please refer to the description of the method embodiment.
[0064] In addition, this invention also provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described monitoring model establishment method embodiment and achieves the same technical effect. To avoid repetition, it will not be described again here.
[0065] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described monitoring model establishment method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.
[0066] Figure 4 A schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.
[0067] The electronic device 400 includes, but is not limited to, components such as: a radio frequency unit 401, a network module 402, an audio output unit 403, an input unit 404, a sensor 405, a display unit 406, a user input unit 407, an interface unit 408, a memory 409, a processor 410, and a power supply 411. Those skilled in the art will understand that... Figure 4 The electronic device structures shown are not intended to limit the electronic device. An electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements. In embodiments of the present invention, the electronic device includes, but is not limited to, mobile phones, tablet computers, laptops, PDAs, in-vehicle terminals, wearable devices, and pedometers.
[0068] It should be understood that, in this embodiment of the invention, the radio frequency unit 401 can be used for receiving and transmitting signals during information transmission or calls. Specifically, it receives downlink data from the base station and processes it with the processor 410; additionally, it transmits uplink data to the base station. Typically, the radio frequency unit 401 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low-noise amplifier, a duplexer, etc. Furthermore, the radio frequency unit 401 can also communicate with networks and other devices through a wireless communication system.
[0069] The electronic device provides users with wireless broadband internet access through network module 402, such as helping users send and receive emails, browse web pages, and access streaming media.
[0070] The audio output unit 403 can convert audio data received by the radio frequency unit 401 or the network module 402 or stored in the memory 409 into audio signals and output them as sound. Furthermore, the audio output unit 403 can also provide audio output related to specific functions performed by the electronic device 400 (e.g., call signal reception sound, message reception sound, etc.). The audio output unit 403 includes a speaker, a buzzer, and a receiver, etc.
[0071] Input unit 404 is used to receive audio or video signals. Input unit 404 may include a graphics processing unit (GPU) 4041 and a microphone 4042. The GPU 4041 processes image data of still images or videos acquired by an image capture device (such as a camera) in video capture mode or image capture mode. The processed image frames can be displayed on display unit 406. The image frames processed by GPU 4041 can be stored in memory 409 (or other storage media) or transmitted via radio frequency unit 401 or network module 402. Microphone 4042 can receive sound and process such sound into audio data. The processed audio data can be converted into a format that can be transmitted to a mobile communication base station via radio frequency unit 401 in telephone call mode.
[0072] The electronic device 400 also includes at least one sensor 405, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor includes an ambient light sensor and a proximity sensor. The ambient light sensor can adjust the brightness of the display panel 4061 according to the ambient light level, and the proximity sensor can turn off the display panel 4061 and / or backlight when the electronic device 400 is moved to the ear. As a type of motion sensor, an accelerometer sensor can detect the magnitude of acceleration in various directions (generally three axes). When stationary, it can detect the magnitude and direction of gravity and can be used to identify the posture of the electronic device (such as landscape / portrait switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc. The sensor 405 may also include a fingerprint sensor, pressure sensor, iris sensor, molecular sensor, gyroscope, barometer, hygrometer, thermometer, infrared sensor, etc., which will not be described in detail here.
[0073] The display unit 406 is used to display information input by the user or information provided to the user. The display unit 406 may include a display panel 4061, which may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.
[0074] User input unit 407 can be used to receive input numerical or character information, and to generate key signal inputs related to user settings and function control of electronic devices. Specifically, user input unit 407 includes a touch panel 4071 and other input devices 4072. Touch panel 4071, also known as a touch screen, can collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near touch panel 4071). Touch panel 4071 may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch position and the signal generated by the touch operation, and transmits the signal to the touch controller; the touch controller receives touch information from the touch detection device, converts it into touch point coordinates, and sends it to the processor 410, which receives and executes commands from the processor 410. In addition, touch panel 4071 can be implemented using various types such as resistive, capacitive, infrared, and surface acoustic wave. Besides touch panel 4071, user input unit 407 may also include other input devices 4072. Specifically, other input devices 4072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, joysticks, etc., which will not be described in detail here.
[0075] Furthermore, the touch panel 4071 can cover the display panel 4061. When the touch panel 4071 detects a touch operation on or near it, it transmits the information to the processor 410 to determine the type of touch event. Subsequently, the processor 410 provides corresponding visual output on the display panel 4061 based on the type of touch event. Although in Figure 4 In this embodiment, the touch panel 4071 and the display panel 4061 are two independent components to realize the input and output functions of the electronic device. However, in some embodiments, the touch panel 4071 and the display panel 4061 can be integrated to realize the input and output functions of the electronic device. The specific implementation is not limited here.
[0076] Interface unit 408 serves as an interface for connecting external devices to electronic device 400. For example, external devices may include a wired or wireless headphone port, an external power supply (or battery charger) port, a wired or wireless data port, a memory card port, a port for connecting a device with an identification module, an audio input / output (I / O) port, a video I / O port, a headphone port, and so on. Interface unit 408 can be used to receive input from external devices (e.g., data, power, etc.) and transmit the received input to one or more components within electronic device 400, or it can be used to transmit data between electronic device 400 and external devices.
[0077] The memory 409 can be used to store software programs and various data. The memory 409 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function (such as sound playback, image playback, etc.), etc.; the data storage area may store data created based on the use of the mobile phone (such as audio data, phonebook, etc.). Furthermore, the memory 409 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.
[0078] The processor 410 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in the memory 409, and by calling data stored in the memory 409, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. The processor 410 may include one or more processing units; preferably, the processor 410 may integrate an application processor and a modem processor. The application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 410.
[0079] The electronic device 400 may also include a power supply 411 (such as a battery) that supplies power to various components. Preferably, the power supply 411 can be logically connected to the processor 410 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system.
[0080] In addition, the electronic device 400 includes some functional modules not shown, which will not be described in detail here.
[0081] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0082] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.
[0083] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of the present invention.
[0084] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.
[0085] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0086] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.
[0087] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0088] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0089] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.
[0090] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.
Claims
1. A method for establishing a monitoring model, characterized in that, The method is applied to an intelligent control system for hazardous chemical special operations, and the method includes: Acquire information on hazardous chemical operation accidents; the hazardous chemical operation accident information includes multi-source heterogeneous data; Based on the hazardous chemical operation accident information, monitoring and early warning indicators are determined for the hazardous chemical operation accident information; The model parameters are determined using the aforementioned multi-source heterogeneous data; A target model is generated based on the model parameters and the monitoring and early warning indicators; The intelligent management and control system for special hazardous chemical operations has a corresponding expert database. The monitoring and early warning indicators include risk factor data and first abnormal behavior data. The step of generating a target model based on the model parameters and the monitoring and early warning indicators includes: Construct a risk factor data set for the aforementioned risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in the Bayesian network; Construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network. The prior probability of an action occurring for the risk factor dataset is calculated based on the expert database using the risk factor nodes and the abnormal behavior nodes. A Bayesian network is trained using the prior probability of the aforementioned behavior, and a conditional probability table for each node in the Bayesian network is determined. Generate a first subset of abnormal behaviors; the first subset of abnormal behaviors is an empty set; A second subset of abnormal behaviors is determined from the monitoring and early warning indicators, the second subset of abnormal behaviors including second abnormal behavior data; An abnormal behavior threshold is generated by performing a threshold evaluation on the second abnormal behavior data, and an abnormal behavior threshold set is constructed based on the abnormal behavior threshold. A risk factor set is generated using the set of abnormal behavior thresholds; the risk factor set is an empty set. The second abnormal behavior data corresponding to the abnormal behavior threshold is determined from the abnormal behavior threshold set in a preset order; A third subset of abnormal behavior data is generated using the second abnormal behavior data and the monitoring and early warning indicators; The target risk factor set is determined by the risk factor set and the third abnormal behavior data subset; The first subset of abnormal behaviors is determined by the first subset of abnormal behaviors and the second subset of abnormal behavior data; When the target risk factor set meets the preset conditions, the first abnormal behavior subset of the target is taken as the final abnormal behavior subset, and the target model is generated using the final abnormal behavior subset.
2. The method according to claim 1, characterized in that, The intelligent management and control system for hazardous chemical special operations has a corresponding TF-IDF model. The step of determining the monitoring and early warning indicators for the hazardous chemical operation accident information through the hazardous chemical operation accident information includes: The TF-IDF model was used to extract feature keywords for the hazardous chemical operation accident information; A set of feature keywords is generated using the aforementioned feature keywords; Monitoring and early warning indicators are determined from the set of feature keywords using the expert database and the set of feature keywords.
3. The method according to claim 1, characterized in that, The multi-source heterogeneous data includes location data and video image data, and the method further includes: The location data is used to generate on-duty status data for the workers; The video image data is used to generate behavioral risk factors for the workers.
4. The method according to any one of claims 1-3, characterized in that, Also includes: The target model is encapsulated into an application; the application includes an application programming interface for the intelligent management and control system for hazardous chemical special operations.
5. A monitoring model establishment device, characterized in that, The device is used in an intelligent control system for hazardous chemical special operations. The device includes: The operation accident information acquisition module is used to acquire hazardous chemical operation accident information; the hazardous chemical operation accident information includes multi-source heterogeneous data; The monitoring and early warning indicator determination module is used to determine monitoring and early warning indicators for the hazardous chemical operation accident information based on the hazardous chemical operation accident information. The model parameter determination module is used to determine model parameters using the multi-source heterogeneous data. The target model generation module is used to generate a target model based on the model parameters and the monitoring and early warning indicators. The intelligent management and control system for special hazardous chemical operations has a corresponding expert database. The monitoring and early warning indicators include risk factor data and first abnormal behavior data, and the target model generation module includes: The risk factor data set construction submodule is used to construct a risk factor data set for the risk factor data; the first element in the risk factor data set has a one-to-one corresponding risk factor node in the Bayesian network. An abnormal behavior data set construction submodule is used to construct an abnormal behavior data set for the first abnormal behavior data; the second element of the abnormal behavior data set has a one-to-one corresponding abnormal behavior node in the Bayesian network. The prior probability calculation submodule is used to calculate the prior probability of an action occurring for the risk factor data set based on the expert database using the risk factor node and the abnormal behavior node. The conditional probability table determination submodule is used to train a Bayesian network using the prior probability of the behavior occurring, and to determine the conditional probability table of each node in the Bayesian network. The first abnormal behavior subset generation submodule is used to generate a first abnormal behavior subset; the first abnormal behavior subset is an empty set. The second abnormal behavior subset generation submodule is used to determine a second abnormal behavior subset from the monitoring and early warning indicators, wherein the second abnormal behavior subset includes second abnormal behavior data; The abnormal behavior threshold set construction submodule is used to generate abnormal behavior thresholds by performing threshold evaluation on the second abnormal behavior data, and to construct an abnormal behavior threshold set based on the abnormal behavior thresholds. The risk factor set generation submodule is used to generate a risk factor set from the abnormal behavior threshold set; the risk factor set is an empty set. The second abnormal behavior data determination submodule is used to determine the second abnormal behavior data corresponding to the abnormal behavior threshold from the abnormal behavior threshold set in a preset order; The third abnormal behavior data subset generation submodule is used to generate a third abnormal behavior data subset using the second abnormal behavior data and the monitoring and early warning indicators; The target risk factor set determination submodule is used to determine the target risk factor set through the risk factor set and the third abnormal behavior data subset; The first abnormal behavior subset determination submodule is used to determine the first abnormal behavior subset of the target through the first abnormal behavior subset and the second abnormal behavior data; The target model generation submodule is used to generate a target model by taking the first abnormal behavior subset of the target as the final abnormal behavior subset when the target risk factor set meets the preset conditions.
6. The apparatus according to claim 5, characterized in that, The intelligent management and control system for hazardous chemical special operations has a corresponding TF-IDF model, and the monitoring and early warning indicator determination module includes: The feature keyword extraction submodule is used to extract feature keywords for the hazardous chemical operation accident information using the TF-IDF model. A set generation submodule is used to generate a set of feature keywords using the feature keywords; The monitoring and early warning indicator determination submodule is used to determine the monitoring and early warning indicators from the feature keyword set through the expert database and the feature keyword set.
7. The apparatus according to claim 5, characterized in that, The multi-source heterogeneous data includes positioning data and video image data, and the device further includes: The on-duty status data generation submodule is used to generate on-duty status data for operators based on the location data. The behavioral risk factor generation submodule is used to generate behavioral risk factors for workers based on the video image data.
8. The apparatus according to any one of claims 5-7, characterized in that, Also includes: An application encapsulation submodule is used to encapsulate the target model into an application; the application includes an application programming interface for the intelligent management and control system for hazardous chemical special operations.
9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes a program stored in the memory, it implements the method as described in any one of claims 1-4.
10. A computer-readable storage medium having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the method as described in any one of claims 1-4.
Citation Information
Patent Citations
Method and system for establishing risk monitoring model and risk monitoring method and system
CN113821899A
Risk prediction method based on TF-IDF and fuzzy Bayesian network
CN115115159A
