Gas safety value-keeping intelligent management method based on mass alarm data

Abstract

This invention relates to an intelligent management method for gas safety monitoring based on massive alarm data, belonging to the field of gas safety monitoring and management technology. This method addresses the problems of low information processing efficiency, non-standardized handling procedures, imperfect supervision mechanisms, difficulties in cross-departmental collaboration, and insufficient data analysis capabilities in traditional gas safety management. It employs artificial intelligence (AI), robotic process automation (RPA), and big data analytics to achieve closed-loop management of the entire process, from data reception, standardized processing, intelligent analysis, AI-based dispatching, handling tracking, to intelligent supervision. This method supports efficient processing of massive alarm data, standardizes handling procedures, improves supervision mechanisms, and enables cross-departmental collaboration and big data-driven predictive early warning, significantly improving the efficiency and level of gas safety management and reducing safety risks.

Description

Technical Field

[0001] This invention relates to the field of gas safety monitoring and management technology, and to a gas safety monitoring and intelligent management method based on massive alarm data. Specifically, it is a gas safety monitoring and intelligent management method that utilizes artificial intelligence (AI), robotic process automation (RPA), and big data analysis technology to achieve closed-loop management of massive alarm data reception, processing, and supervision. Background Technology

[0002] With the acceleration of urbanization and the expansion of gas usage, gas safety issues are becoming increasingly prominent. Traditional gas safety management mainly relies on manual on-site monitoring and inspections, which has the following problems:

[0003] 1. Low information processing efficiency: Gas leak alarms, pressure sensors and other smart IoT sensing devices generate massive amounts of alarm data, which are difficult to handle manually and can easily lead to omissions or delays in alarm information.

[0004] 2. Non-standardized handling procedures: There is a lack of standardized alarm handling procedures, and the efficiency and quality of handling depend on the experience and sense of responsibility of the operators.

[0005] 3. Inadequate supervision mechanism: There is a lack of effective means to supervise alarms and hidden dangers that are not dealt with in a timely manner, which can easily lead to the escalation of safety problems.

[0006] 4. Difficulty in cross-departmental collaboration: Gas safety management involves multiple departments such as early warning platforms, emergency systems, and gas companies, and traditional management methods are difficult to achieve efficient collaboration.

[0007] 5. Insufficient data analysis capabilities: Unable to effectively analyze massive amounts of historical data, making it difficult to predict potential security risks and achieve early warning.

[0008] In existing technologies, some gas companies have adopted simple automation systems, but these systems are often limited in function and cannot achieve closed-loop management of the entire process from data reception and processing to supervision. They also lack AI and big data analytics capabilities. Therefore, there is an urgent need for an intelligent management method for gas safety monitoring that can integrate AI, RPA, and big data analytics technologies. Summary of the Invention

[0009] This invention aims to address the problems existing in current gas safety management, such as low information processing efficiency, non-standard handling procedures, imperfect supervision mechanisms, difficulties in cross-departmental collaboration, and insufficient data analysis capabilities. It provides an intelligent solution for gas safety duty that can achieve closed-loop management of massive alarm data reception, processing, and supervision.

[0010] The technical solution adopted by this invention to achieve the above objectives is: a gas safety monitoring and intelligent management method based on massive alarm data, comprising the following steps:

[0011] Data reception and integration: Receives real-time alarm information from intelligent monitoring devices, gas safety hazard information, and verification information from gas licenses and intelligent monitoring devices, forming multi-source data;

[0012] Data processing and analysis: Preprocess multi-source data, obtain risk scores through risk prediction models, determine the severity and geographical extent of alarm events; and automatically identify overdue warning events and safety hazards.

[0013] AI-powered intelligent dispatching and handling: Based on the severity and geographical scope of alarm events, the system determines handling priorities, automatically matches handling nodes, generates handling work orders, and pushes them to the handling nodes;

[0014] Intelligent supervision and closed-loop management: Automatically trigger secondary supervision for handling nodes that fail to respond within the time limit; when the number of supervisions reaches the threshold, send early warning information; automatically verify the handling results to form a closed-loop record.

[0015] The receiving of real-time alarm information from intelligent monitoring equipment, gas safety hazard information, and verification information of gas licenses and intelligent monitoring equipment includes:

[0016] Alarm information reception: Receives real-time alarm information from intelligent monitoring devices via the Internet of Things; the intelligent monitoring devices include gas leak alarms and environmental sensors.

[0017] Hazard Information Reception: Receives safety hazard information obtained through inspections and intelligent monitoring equipment;

[0018] License Information Monitoring: Real-time monitoring of the validity period and verification information of gas licenses;

[0019] Equipment Inspection and Monitoring: Monitor the inspection date of intelligent monitoring equipment.

[0020] The preprocessing of multi-source data to unify the data format and storage structure includes, but is not limited to, any of the following: unique device identifier, device type, historical alarm frequency, event occurrence time, event type, data source, duration, sensor data, and location information.

[0021] The risk score is obtained through a risk prediction model to determine the severity and geographical extent of alarm events; and warning events and safety hazards that have not been addressed within the specified time are automatically identified, including the following steps:

[0022] (1) Pre-construct risk prediction model: Historical data is mined through big data analysis, including the time pattern and duration of historical events, the area where they occur, and considering factors such as equipment type, historical alarm frequency, and environmental sensor data. The model parameters are determined by the alarm level, type and status change trend, and a risk prediction model is established. The historical data includes: the time of occurrence, duration, area range, equipment type, number of alarms, environmental sensor data, alarm level and alarm type of historical events.

[0023] The main parameters of the risk prediction model are as follows:

[0024] Risk prediction result = Alarm level weight + Equipment type weight + Historical alarm frequency weight + Environmental factor weight; The weights of each parameter are obtained through analysis of historical data;

[0025] (2) Risk prediction is performed on the pre-processed multi-source data by establishing a good risk prediction model. Several parameters in the multi-dimensional environmental sensor data including alarm level, equipment type, historical alarm frequency, area range, pressure, temperature and humidity are input. The parameter weights corresponding to each input parameter are matched respectively. The risk prediction results are output through the risk prediction model to represent the score.

[0026] Events are classified into different levels based on the magnitude of risk prediction results to characterize the severity of alarm events, including different levels of early warning events and potential hazards;

[0027] (3) The time difference between the creation time of the warning event and the current time is automatically compared with the handling time limit threshold to identify warning events that have not been handled within the time limit;

[0028] The difference between the time of hazard registration and the current time is automatically compared with the set time limit for graded handling to identify safety hazards that have not been handled within the time limit;

[0029] Real-time monitoring of gas license validity periods and equipment inspection dates; identification of licenses that have not been reviewed and approved and equipment that have not been inspected.

[0030] The process of determining handling priorities based on the severity and geographical scope of alarm events, automatically matching handling nodes, generating handling work orders, and pushing them to the handling nodes includes the following steps:

[0031] Determine response priorities based on the severity and geographical scope of early warning events and potential hazards:

[0032] The overall score is calculated as follows: Severity weight + Regional scope weight + Historical alarm frequency weight. The severity weight is used to input the risk prediction results, while the regional scope weight and historical alarm frequency weight are obtained through the analysis of historical data.

[0033] The comprehensive scoring results are normalized; priority is assigned based on the normalized scores to determine the response level.

[0034] According to the established responsibility division rules, the handling priority is automatically matched with the corresponding handling node to generate a handling work order; the handling work order includes a unique work order identifier, alarm level, alarm type, alarm value, longitude, latitude, geographical location description, handling priority, and work order creation time;

[0035] RPA automatically pushes work orders to the processing nodes and tracks the processing progress.

[0036] The intelligent supervision and closed-loop management system automatically triggers secondary supervision for handling nodes that fail to respond within the time limit. When the number of supervisions reaches a threshold, an early warning message is sent. The system automatically verifies the handling results to form a closed-loop record, including the following steps:

[0037] For handling nodes that fail to respond within the time limit, a second supervision will be automatically triggered;

[0038] When the number of supervisions reaches the threshold, an early warning message is sent to the superior terminal.

[0039] Receive and automatically verify the results of the completed processing;

[0040] The results of the completed process will be synchronized to the regulatory backend to form a closed-loop record.

[0041] A smart management method for gas safety monitoring based on massive alarm data also includes cross-departmental collaboration and information sharing, as detailed below:

[0042] The entire process of data reception and integration, data processing and analysis, AI-powered intelligent dispatch and handling, intelligent supervision and closed-loop management is visualized and tracked, and synchronized to multiple higher-level terminals.

[0043] The entire process progress information includes: alarm events, handling nodes, handling results, supervision and the number of supervisions, and verification of handling results.

[0044] A gas safety monitoring and intelligent management method based on massive alarm data includes:

[0045] The data receiving and integration unit is used to receive real-time alarm information from intelligent monitoring equipment, gas safety hazard information, and verification information of gas licenses and intelligent monitoring equipment, forming multi-source data;

[0046] The data processing and analysis unit is used to preprocess multi-source data, obtain risk scores through risk prediction models, determine the severity and geographical extent of alarm events, and automatically identify overdue warning events and safety hazards.

[0047] The AI-powered intelligent dispatch and handling unit is used to determine the handling priority based on the severity and geographical scope of alarm events, automatically match handling nodes, generate handling work orders, and push them to the handling nodes.

[0048] The intelligent supervision and closed-loop management unit is used to automatically trigger secondary supervision for handling nodes that have not responded within the time limit. When the number of supervisions reaches a threshold, an early warning message is sent. The handling results are automatically verified to form a closed-loop record.

[0049] A gas safety monitoring and intelligent management device based on massive alarm data includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement the gas safety monitoring and intelligent management method based on massive alarm data when the computer program is executed.

[0050] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned intelligent management method for gas safety monitoring based on massive alarm data.

[0051] The present invention has the following beneficial effects and advantages:

[0052] 1. Improve data processing efficiency: Through AI and RPA technologies, the system can automatically receive, process, and dispatch massive amounts of alarm data, greatly improving data processing efficiency and reducing manual intervention.

[0053] 2. Standardize handling procedures: Establish standardized handling procedures to ensure that alarms and potential hazards are handled in a timely and standardized manner, thereby improving the quality and consistency of handling.

[0054] 3. Improve the supervision mechanism: Implement intelligent supervision function to automatically trigger supervision for events that have not been handled within the time limit, ensure that the handling work is completed on time, and reduce safety risks.

[0055] 4. Enable cross-departmental collaboration: Through the AI-powered multi-departmental collaborative system, cross-departmental information sharing and collaborative processing are achieved, improving overall management efficiency.

[0056] 5. Enhance forecasting and early warning capabilities: Utilize big data analytics to mine historical data, establish risk prediction models, achieve early warning, and reduce the occurrence of safety accidents.

[0057] 6. Reduce management costs: Reduce the workload of manual on-duty personnel and on-site inspections, thereby reducing management costs and improving management efficiency. Attached Figure Description

[0058] Figure 1 System framework diagram of the present invention;

[0059] Figure 2 The alarm information processing flowchart of this invention;

[0060] Figure 3 The safety hazard handling flowchart of the present invention. Detailed Implementation

[0061] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0062] This invention provides a smart management method for gas safety duty based on AI, RPA, and big data analysis for closed-loop management of massive alarm data reception, processing, and supervision, applicable to the following application scenarios:

[0063] Gas company safety management: Used for safety monitoring and management of gas companies to improve the efficiency and level of safety management.

[0064] Regulatory Unit: Used for supervision and management of gas safety by regulatory units such as early warning platforms and emergency systems, to improve regulatory efficiency and effectiveness.

[0065] Gas safety management for large complexes: Used for gas safety management in large places such as commercial complexes and industrial parks to ensure the safety of people and property.

[0066] Gas safety management in residential communities: Used for gas safety management in residential communities to improve the safety of residents' lives.

[0067] This invention integrates AI, RPA, and big data analytics technologies to achieve intelligent, automated, and information-based gas safety management, providing an efficient and reliable solution for gas safety. This invention includes the following steps:

[0068] Data reception and integration: Receives real-time alarm information from smart IoT sensing devices such as gas leak detectors and pressure sensors, as well as gas safety hazard information gathered from multiple channels; monitors the inspection dates of gas-related licenses and equipment; supports multiple communication protocols such as MQTT and HTTP; reads gas equipment data using a loop-based paging method; and converts the raw data into a unified standard format before batch storage.

[0069] Data processing and analysis: Standardize the various types of data received, use big data analytics to build risk prediction models, automatically identify overdue warning events and safety hazards; group and summarize the data by date, and calculate key indicators such as total number of alarms, number of alarms handled, number of overdue alarms, and handling rate.

[0070] AI-powered intelligent dispatching and handling: AI technology is used to analyze and determine handling priorities, automatically match responsible parties according to responsibility division rules, generate standardized handling work orders and push them out; through RPA technology, handling work orders are automatically pushed to the mobile devices of relevant personnel, and data on the handling process is collected in real time to track the handling progress.

[0071] Intelligent supervision and closed-loop management: Automatically trigger secondary supervision for handling nodes that fail to respond within the time limit. When the number of supervisions reaches a threshold, send an early warning message to the superior management unit, automatically verify the handling effect and form a closed-loop record; Based on the preset handling time limit threshold, automatically compare the creation time of the alarm event with the current system time difference to identify early warning events that have not been handled within the time limit.

[0072] Cross-departmental collaboration and information sharing: An AI-powered agent system enables multi-departmental collaboration, synchronizing handling results to each department's system to form a traceable closed-loop record across departments. This AI-powered agent system enables collaboration among multiple departments, including early warning platforms, emergency systems, and gas companies, achieving end-to-end automation from cross-entity early warning distribution, intelligent dispatch of tiered instructions, visualized tracking of the entire process progress, to multi-terminal synchronization of handling results.

[0073] A smart management method for gas safety monitoring based on massive alarm data includes the following steps:

[0074] 1. Data reception and integration

[0075] The system employs a loop-based pagination method to read gas equipment data, ensuring efficient processing of large-scale data. Raw equipment data is converted into a unified standard format, including the equipment's manufacturer, alarm type, alarm code, equipment status, department, and geographic location information (latitude, longitude, and address). The standardized data is then batch-stored into the system database to improve data processing efficiency. The system receives and processes various alarm information, including real-time alarm data from gas leak detectors and pressure sensors. Alarm information is managed uniformly, including saving new alarm information and updating the status of existing alarm information. Multiple communication protocols, such as MQTT and HTTP, are supported to ensure compatibility with different types of smart IoT sensing devices. Real-time monitoring and management of various gas-related licenses and special equipment inspection dates are implemented. Specifically, the following steps are included:

[0076] Alarm information reception: Receives alarm information transmitted in real time from smart IoT sensing devices such as gas leak alarms and pressure sensors through the IoT platform, and supports multi-protocol compatibility to access data from various terminals;

[0077] Hazard Information Reception: Receives gas safety hazard information gathered through multiple channels, including daily inspection reports, intelligent monitoring equipment, and automatic system data collection;

[0078] License and permit information monitoring: Monitor the review and approval dates of various gas-related licenses and permits, such as gas operation licenses and filling licenses;

[0079] Equipment inspection and monitoring: Monitoring the inspection dates of special equipment and other safety devices.

[0080] 2. Data Processing and Analysis

[0081] Query alarm data generated by user alarm devices and gas pipeline monitoring equipment separately. Merge the two types of alarm data into a unified dataset to prepare for subsequent analysis. Determine the analysis time range based on the query conditions; this can be the most recent 30 days from a specified date or user-defined start and end dates. Group and summarize the data by date to ensure that each date has corresponding statistical data. Calculate the following core indicators for each date: Total number of alarms: the total number of alarms generated that day. Number of alarms handled: the number of alarms that have been handled that day. Number of alarms not handled within the preset handling time limit: the number of alarms that have not been handled within the preset handling time limit. Handling rate: the percentage of handled alarms out of the total number of alarms, accurate to four decimal places. Set default values ​​(all indicators are 0) for dates with no data to ensure data integrity and consistency. Sort the analysis results by date to ensure the correct chronological order of the output results. Generate standardized analysis results to provide data support for the system's chart display function. Automatically identify events that have not been handled within the preset handling time limit threshold and trigger warnings. Utilize big data technology to deeply mine historical data, establish a risk prediction model, and realize the pre-warning function.

[0082] The implementation steps are as follows:

[0083] Standardize the received data of all types, and unify the data format and storage structure;

[0084] Pre-build a risk prediction model: Utilize big data analytics to mine historical data, including the timing and duration of historical events, the regions where they occur, and consider factors such as equipment type and historical alarm frequency. Integrate multi-dimensional sensor data of pressure, temperature, and humidity, and determine model parameters such as alarm level, type, and status change trends to establish a risk prediction model.

[0085] The main parameters of the risk prediction model are as follows:

[0086] Risk prediction result (score) = Alarm level weight + Equipment type weight + Historical frequency weight + Environmental factor weight.

[0087] The model parameters of the risk prediction model can also be increased based on historical data, such as: regional range weight, alarm type weight, event occurrence time weight, event duration weight, etc. Each weight is set through historical data mining and analysis. In this embodiment, the alarm level weights are: low (5), medium (15), high (30), and severe (40). The device type weights are: alarm (25), sensor (20), and others (10). The historical frequency weights are: high frequency alarm (20), medium frequency alarm (10), and low frequency alarm (5). The environmental factor weights are: pressure (15), temperature (10), and humidity (5). Each weight is adjusted according to the actual training situation, and the total score is 100 points.

[0088] Then, the established risk prediction model is used to predict the risk of subsequent data. Several input parameters are input from multi-dimensional sensor data, including alarm level, type, equipment type, historical alarm frequency, time and duration of event occurrence, pressure, temperature and humidity. By querying and matching the set weights of each parameter, the risk prediction model outputs the risk prediction result (score).

[0089] Events are classified into different levels based on the risk prediction results (scores) to characterize the severity of alarm events, including different levels of early warning events and potential hazards.

[0090] Based on a preset processing time limit threshold, the system automatically compares the creation time of an alarm event with the current system time difference to identify warning events that have not been processed within the time limit.

[0091] Based on preset tiered handling time limits, the system automatically compares the time difference between the hazard registration time and the current time to identify safety hazards that have not been handled within the time limit.

[0092] Real-time monitoring of license validity periods and equipment inspection dates; identification of licenses and equipment that have not been reviewed or inspected and have expired.

[0093] 3. AI-powered intelligent order dispatch and processing

[0094] This module is responsible for using AI technology to achieve intelligent dispatching and handling. It uses AI technology to analyze factors such as the severity and scope of impact of alarms and potential hazards to determine handling priorities;

[0095] The overall score is calculated using a combination of severity weighting, impact range weighting, and historical frequency weighting, with the results then normalized.

[0096] Among them, the severity weight is used to input the risk prediction result, while the regional scope weight and the historical alarm frequency weight are obtained through the analysis of historical data.

[0097] In this embodiment, the priority is divided as follows: 90-100 points: Immediate response, highest priority; 70-89 points: Response as soon as possible, high priority; 50-69 points: Routine processing, medium priority; 30-49 points: Monitoring and observation, low priority; 0-29 points: Recording and filing, lowest priority.

[0098] Based on the responsibility division rules, the system automatically matches the corresponding responsible entity and generates a standardized handling work order, which includes a unique work order identifier, alarm level, alarm type, alarm value, longitude, latitude, geographical location description, handling priority, and work order creation time.

[0099] RPA technology is used to automatically push work orders to the mobile devices of relevant personnel.

[0100] Collect data in real time during the disposal process and track the progress of disposal.

[0101] 4. Intelligent supervision and closed-loop management

[0102] This module is responsible for implementing intelligent supervision and closed-loop management, including:

[0103] For handling nodes that fail to respond within the time limit, a second supervision is automatically triggered; when the number of supervisions reaches the threshold, an early warning message is sent to the superior management unit; the handling results are received and the handling effect is automatically verified; the handling results are synchronized to the supervision backend to form a closed-loop record.

[0104] 5. Cross-departmental collaboration and information sharing

[0105] This module is responsible for enabling cross-departmental collaboration and information sharing. It realizes an AI-powered agent system that facilitates collaboration among multiple departments, including the early warning platform, emergency system, and gas companies; it automates the entire process from cross-entity early warning distribution, intelligent dispatch of tiered instructions, and full-process progress visualization tracking to multi-terminal synchronization of handling results; it generates standardized handling work orders containing event details, handling time limits, and responsible contact persons, and pushes them to the mobile work platforms of each entity; and it synchronizes the completed handling results to the early warning platform's monitoring backend, the emergency system's command system, and the gas company's archives, forming a cross-departmental traceable closed-loop record.

[0106] like Figure 1 As shown, the system architecture of this invention includes the following core components:

[0107] Data acquisition layer: responsible for collecting data from various smart IoT sensing devices, inspection systems, certificate management systems and other channels.

[0108] Data processing layer: Responsible for the standardized processing, storage, and analysis of data.

[0109] AI Intelligence Layer: Responsible for implementing AI algorithms such as alarm classification, intelligent dispatching, and risk prediction.

[0110] RPA Automation Layer: Responsible for executing automated processes, such as work order push and supervision reminders.

[0111] Application layer: This includes enterprise platforms, regulatory platforms, mobile applications, etc., providing user interfaces.

[0112] Data sharing layer: Enables cross-departmental data sharing and collaboration.

[0113] The data acquisition layer is responsible for receiving information from various data sources, including: receiving real-time alarm data from devices such as gas leak detectors and pressure sensors via an IoT platform, supporting multiple protocols such as MQTT and HTTP; receiving daily inspection reports and safety hazard information detected by intelligent monitoring equipment; monitoring the validity period and review status of gas-related licenses in real time; and monitoring the inspection dates of special equipment and safety devices.

[0114] The data processing layer is responsible for the standardization and analysis of data, including: converting data from different sources and in different formats into a unified standard format; utilizing big data technology to mine historical data and build risk prediction models; and automatically identifying events that have exceeded the time limit for handling based on preset handling time limits.

[0115] The AI ​​intelligence layer is responsible for using AI technology to achieve intelligent work order dispatch and handling, including: classifying alarms and hazards based on AI algorithms to determine handling priorities; automatically matching the corresponding responsible parties according to responsibility division rules; and automatically pushing work orders to relevant personnel's mobile devices through RPA technology.

[0116] The RPA automation layer is responsible for intelligent supervision and closed-loop management, including: automatically triggering secondary supervision for handling nodes that fail to respond within the time limit; sending early warning information to the superior management unit when the number of supervisions reaches a threshold; receiving the results of completed handling and automatically verifying the handling effect; and synchronizing the results of completed handling to the regulatory backend to form a closed-loop record.

[0117] The data sharing layer is responsible for enabling cross-departmental collaboration and information sharing, including: an AI-powered agent system that facilitates collaboration among multiple departments such as the early warning platform, emergency response system, and gas companies. The results of completed actions are synchronized to the systems of each department, forming a traceable, closed-loop record across departments.

[0118] like Figure 2 As shown, the alarm information processing flow is as follows:

[0119] User alarm: The gas user's alarm has been triggered.

[0120] Enterprise platform: Gas companies receive alarm information and perform initial processing.

[0121] Monitoring and early warning robot seat: It includes three modules: alarm receiving, receiving, and supervision, and is responsible for receiving, processing and supervising alarm information.

[0122] Alarm receiving module: Receives alarm information reported by user alarms and enterprise platforms.

[0123] Receiving module: Receives alarm information that has been processed / eliminated reported by the enterprise platform.

[0124] Supervision module: Automatically supervises and handles alarm information that has not been processed or cleared.

[0125] The Urban Construction Bureau's early warning platform receives information reported by robot operators and conducts supervision and management.

[0126] Reporting Alarms That Have Been Handled / Cleared: The robot operator will report the alarm information that has been handled to the Urban Construction Bureau.

[0127] Reporting unresolved / unresolved supervisory information: The robot operator will report unresolved alarm information and supervisory status to the Urban Construction Bureau.

[0128] Manual supervision: The early warning platform can manually supervise unprocessed alarm information.

[0129] like Figure 3 As shown, the safety hazard handling process is as follows:

[0130] Safety hazard reporting: Report safety hazards through daily inspections, intelligent monitoring and other channels.

[0131] Enterprise platform: Gas companies receive hazard information and conduct preliminary processing.

[0132] Monitoring and early warning robot seat: It includes three modules: alarm receiving, receiving, and supervision, and is responsible for receiving, processing and supervising the handling of hidden danger information.

[0133] Alarm receiving module: Receives information on reported safety hazards.

[0134] Receiving module: Receives information on rectified / eliminated potential hazards reported by the enterprise platform.

[0135] Supervision module: Automatically supervises and handles information on unresolved potential hazards.

[0136] The Urban Construction Bureau's early warning platform receives information reported by robot operators and conducts supervision and management.

[0137] Reporting information on safety hazards that have been addressed / eliminated: The robot operator will report information on hazards that have been addressed to the Urban Construction Bureau.

[0138] Reporting unresolved / uneliminated hazard information: The robot operator will report the information on unresolved hazards and the progress of supervision to the Urban Construction Bureau.

[0139] Manual supervision: The early warning platform can provide manual supervision for unaddressed hazard information.

[0140] This invention utilizes artificial intelligence (AI) technology to achieve intelligent analysis, intelligent task dispatch, and intelligent supervision, improving decision-making efficiency and accuracy. Through Robotic Process Automation (RPA), it automates processes such as work order delivery and supervision reminders, reducing manual intervention. Big data technology is used to mine historical data and establish risk prediction models for proactive warnings. It achieves closed-loop management of the entire process from data reception to completion, ensuring that every event is handled promptly and in a standardized manner. It enables a multi-departmental collaborative AI-powered agent system, improving cross-departmental collaboration efficiency. The system adopts a modular design with good scalability, allowing for functional expansion and customization according to actual needs.

Claims

1. A gas safety monitoring and intelligent management method based on massive alarm data, characterized in that, Includes the following steps: Data reception and integration: Receives real-time alarm information from intelligent monitoring devices, gas safety hazard information, and verification information from gas licenses and intelligent monitoring devices, forming multi-source data; Data processing and analysis: Preprocess multi-source data, obtain risk scores through risk prediction models, determine the severity and geographical extent of alarm events; and automatically identify overdue warning events and safety hazards. AI-powered intelligent dispatching and handling: Based on the severity and geographical scope of alarm events, the system determines handling priorities, automatically matches handling nodes, generates handling work orders, and pushes them to the handling nodes; Intelligent supervision and closed-loop management: Automatically trigger secondary supervision for handling nodes that fail to respond within the time limit; when the number of supervisions reaches the threshold, send early warning information; automatically verify the handling results to form a closed-loop record.

2. The intelligent management method for gas safety monitoring based on massive alarm data according to claim 1, characterized in that, The receiving of real-time alarm information from intelligent monitoring equipment, gas safety hazard information, and verification information of gas licenses and intelligent monitoring equipment includes: Alarm information reception: Receives real-time alarm information from intelligent monitoring devices via the Internet of Things; the intelligent monitoring devices include gas leak alarms and environmental sensors. Hazard Information Reception: Receives safety hazard information obtained through inspections and intelligent monitoring equipment; License Information Monitoring: Real-time monitoring of the validity period and verification information of gas licenses; Equipment Inspection and Monitoring: Monitor the inspection date of intelligent monitoring equipment.

3. The intelligent management method for gas safety monitoring based on massive alarm data according to claim 1, characterized in that, The preprocessing of multi-source data to unify the data format and storage structure includes, but is not limited to, any of the following: unique device identifier, device type, historical alarm frequency, event occurrence time, event type, data source, duration, sensor data, and location information.

4. The intelligent management method for gas safety monitoring based on massive alarm data according to claim 1, characterized in that, The risk score is obtained through a risk prediction model to determine the severity and geographical extent of alarm events; and warning events and safety hazards that have not been addressed within the specified time are automatically identified, including the following steps: (1) Pre-construct risk prediction model: Historical data is mined through big data analysis, including the time pattern and duration of historical events, the area where they occur, and considering factors such as equipment type, historical alarm frequency, environmental sensor data, alarm level, type and status change trend to determine model parameters and establish a risk prediction model; The historical data includes: the time of occurrence, duration, area range, equipment type, number of alarms, environmental sensor data, alarm level and alarm type of historical events; The main parameters of the risk prediction model are as follows: Risk prediction result = Alarm level weight + Equipment type weight + Historical alarm frequency weight + Environmental factor weight; The weights of each parameter are obtained through analysis of historical data; (2) Risk prediction is performed on the pre-processed multi-source data by establishing a good risk prediction model. Several parameters in the multi-dimensional environmental sensor data including alarm level, equipment type, historical alarm frequency, area range, pressure, temperature and humidity are input. The parameter weights corresponding to each input parameter are matched respectively. The risk prediction results are output through the risk prediction model to represent the score. Events are classified into different levels based on the magnitude of risk prediction results to characterize the severity of alarm events, including different levels of early warning events and potential hazards; (3) The time difference between the creation time of the warning event and the current time is automatically compared with the handling time limit threshold to identify warning events that have not been handled within the time limit; The difference between the time the hazard was registered and the current time is automatically compared with the set time limit for graded handling to identify safety hazards that have not been handled within the time limit; Real-time monitoring of gas license validity periods and equipment inspection dates; identification of licenses that have not been reviewed and approved and equipment that have not been inspected.

5. The intelligent management method for gas safety monitoring based on massive alarm data according to claim 1, characterized in that, The process of determining handling priorities based on the severity and geographical scope of alarm events, automatically matching handling nodes, generating handling work orders, and pushing them to the handling nodes includes the following steps: Determine response priorities based on the severity and geographical scope of early warning events and potential hazards: The overall score is calculated as follows: Severity weight + Regional scope weight + Historical alarm frequency weight. The severity weight is used to input the risk prediction results, while the regional scope weight and historical alarm frequency weight are obtained through the analysis of historical data. The comprehensive scoring results are normalized; priority is assigned based on the normalized scores to determine the response level. According to the established responsibility division rules, the handling priority is automatically matched with the corresponding handling node to generate a handling work order; the handling work order includes a unique work order identifier, alarm level, alarm type, alarm value, longitude, latitude, geographical location description, handling priority, and work order creation time; RPA automatically pushes work orders to the processing nodes and tracks the processing progress.

6. The intelligent management method for gas safety monitoring based on massive alarm data according to claim 1, characterized in that, The intelligent supervision and closed-loop management system automatically triggers secondary supervision for handling nodes that fail to respond within the time limit. When the number of supervisions reaches a threshold, an early warning message is sent. The system automatically verifies the handling results to form a closed-loop record, including the following steps: For handling nodes that fail to respond within the time limit, a second supervision will be automatically triggered; When the number of supervisions reaches the threshold, an early warning message is sent to the superior terminal. Receive and automatically verify the results of the completed processing; The results of the completed process will be synchronized to the regulatory backend to form a closed-loop record.

7. The intelligent management method for gas safety monitoring based on massive alarm data according to claim 1, characterized in that, It also includes cross-departmental collaboration and information sharing, as detailed below: The entire process of data reception and integration, data processing and analysis, AI-powered intelligent dispatch and handling, intelligent supervision and closed-loop management is visualized and tracked, and synchronized to multiple higher-level terminals. The entire process progress information includes: alarm events, handling nodes, handling results, supervision and the number of supervisions, and verification of handling results.

8. A gas safety monitoring and intelligent management method based on massive alarm data, characterized in that, include: The data receiving and integration unit is used to receive real-time alarm information from intelligent monitoring equipment, gas safety hazard information, and verification information of gas licenses and intelligent monitoring equipment, forming multi-source data; The data processing and analysis unit is used to preprocess multi-source data, obtain risk scores through risk prediction models, and determine the severity and geographical extent of alarm events. It also automatically identifies unhandled warning events and safety hazards that have not been addressed within the specified time. The AI-powered intelligent dispatch and handling unit is used to determine the handling priority based on the severity and geographical scope of alarm events, automatically match handling nodes, generate handling work orders, and push them to the handling nodes. The intelligent supervision and closed-loop management unit is used to automatically trigger secondary supervision for handling nodes that have not responded within the time limit. When the number of supervisions reaches a threshold, an early warning message is sent. The handling results are automatically verified to form a closed-loop record.

9. A gas safety monitoring and intelligent management device based on massive alarm data, characterized in that, It includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement, when executing the computer program, a gas safety monitoring and intelligent management method based on massive alarm data as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements a gas safety monitoring and intelligent management method based on massive alarm data as described in any one of claims 1-7.

Images