Production safety early warning method and system
By classifying and zoning the oil and petrochemical production process, constructing a safety accident early warning model, collecting multi-dimensional indicators, and utilizing blockchain technology, the backwardness of oil and petrochemical safety management has been solved, and the accuracy of risk early warning and safety improvement have been achieved.
Patent Information
- Application Number
- CN202410648190.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-23
- Publication Date
- 2025-11-25
AI Technical Summary
Existing safety management methods in the oil and petrochemical sector are outdated and cannot meet the safety supervision needs under the new circumstances. Production processes and storage pose significant risks and are prone to large-scale safety accidents.
The system categorizes safety incidents, constructs corresponding early warning models, sets up servers within sub-regions to collect multi-dimensional early warning indicators, outputs risk types and probabilities through the models, and enhances data security by combining blockchain technology.
It has enabled more precise risk warnings, allowing for the prediction and prevention of safety accidents in advance, and improving the safety and management efficiency of the oil and petrochemical production process.
Smart Images
Figure CN121010194A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of safety production, and particularly relates to a production safety early warning method and system. BACKGROUND
[0002] The petroleum and chemical industry is an important basic industry and a pillar industry in China, but at the same time, major accidents frequently occur in the petroleum and chemical field, specifically, at present, compared with the improvement of production performance in the petroleum and chemical field, the safety management method for the petroleum and chemical industry still stays in a relatively traditional stage; the current safety management system and management ability of the petroleum and chemical industry are not suitable for the high growth rate and large volume of the industry development, and the traditional management mode and means have been difficult to meet the demand of safety supervision of the petroleum and chemical industry under the new situation; the production process and storage of the petroleum and chemical industry both have great risks, and are prone to cause large-scale safety accidents due to supervision failure. SUMMARY
[0003] The application aims to provide a production safety early warning method to solve the problems of the existing early warning method in actual application in the background technology.
[0004] To achieve the above-mentioned purpose, the application provides the following technical scheme: a production safety early warning method, comprising:
[0005] classifying the safety accident types and constructing safety accident early warning models corresponding to different safety accident types;
[0006] dividing the prediction interval into a plurality of sub-regional intervals, and erecting at least one safety accident early warning model corresponding to the safety accident type of the sub-regional interval on the server in the different sub-regional intervals;
[0007] collecting early warning indexes of safety accidents in different sub-regional intervals, wherein the early warning indexes include human indexes, machine indexes, process indexes, capability indexes and environmental indexes;
[0008] outputting risk types and risk probabilities through the safety accident early warning model in the sub-regional interval based on the collected early warning indexes in the sub-regional interval.
[0009] Preferably, the early warning method further comprises:
[0010] analyzing different safety accident types to obtain influence factors of different safety accident types, wherein the influence factors constitute a subset of the early warning indexes;
[0011] collecting influence factor data of different safety accident types to obtain training samples;
[0012] training the safety accident early warning model based on the training samples.
[0013] Preferably, the early warning method further comprises:
[0014] uploading the early warning indicators of safety accidents in different sub-regional intervals to the blockchain.
[0015] Preferably, the early warning indicators of safety accidents in different sub-regional intervals are uploaded to the blockchain at the same time as the safety accident early warning model is input or are uploaded to the blockchain and then transmitted to the safety accident early warning model by the blockchain.
[0016] Preferably, the human data includes quality data of training personnel, the process indicators include process parameter data, and the environmental indicators include weather parameter data.
[0017] Preferably, the machine indicators include risk rates of static equipment, risk rates of dynamic equipment, risk rates of equipment operation changes, equipment interlocking commissioning parameter data, automatic instrument commissioning parameter data, and special operation parameter data.
[0018] Preferably, the capability indicators include emergency capability parameter data, accident hidden danger investigation parameter data, and fire-fighting facility parameter data.
[0019] Preferably, the safety accident types include object strike accidents, mechanical injury accidents, electric shock accidents, fire accidents, and high-fall accidents.
[0020] Preferably, the early warning method further comprises:
[0021] setting a risk probability threshold value;
[0022] comparing the risk probability output by the safety accident early warning model with the risk probability threshold value;
[0023] when the risk probability output by the safety accident early warning model is greater than the risk probability threshold value, generating early warning information.
[0024] Preferably, the early warning information includes risk types, risk probabilities, and coping schemes.
[0025] Another aspect of the present application discloses a production safety early warning system, comprising:
[0026] a classification module configured to classify safety accident types and build safety accident early warning models corresponding to different safety accident types;
[0027] an erecting module configured to divide a prediction interval into a plurality of sub-regional intervals and erect at least one safety accident early warning model corresponding to a safety accident type in a sub-regional interval on a server in the sub-regional interval;
[0028] The collection module is configured to collect early warning indexes of safety accidents in different sub-regional intervals, and the early warning indexes include human indexes, machine indexes, process indexes, capacity indexes and environment indexes.
[0029] The prediction module is configured to output risk types and risk probabilities through the safety accident early warning model in the sub-regional interval based on the collected early warning indexes in the sub-regional interval.
[0030] Compared with the prior art, the present application has the following beneficial effects:
[0031] The present application classifies the safety accident types, divides the prediction interval, arranges the safety accident early warning model corresponding to the actual situation in each sub-regional interval, combines the multi-dimensional early warning indexes, determines the risk assessment data of the enterprise according to the multi-parameter calculation results, and further determines the risk level and risk type of the enterprise, so that the precision of risk early warning can be realized. BRIEF DESCRIPTION OF DRAWINGS
[0032] Figure 1 The present application is a method flowchart. DETAILED DESCRIPTION
[0033] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.
[0034] A production safety early warning method, comprising the following steps:
[0035] S100: classifying the safety accident types and constructing safety accident early warning models corresponding to different safety accident types;
[0036] S200: dividing the prediction interval to obtain a plurality of sub-regional intervals, and setting up at least one safety accident early warning model corresponding to the safety accident type of the sub-regional interval on the server in the different sub-regional intervals;
[0037] S300: collecting early warning indexes of safety accidents in different sub-regional intervals, and the early warning indexes include human indexes, machine indexes, process indexes, capacity indexes and environment indexes;
[0038] S400: outputting risk types and risk probabilities through the safety accident early warning model in the sub-regional interval based on the collected early warning indexes in the sub-regional interval.
[0039] In some embodiments, in step S100, the classification of the safety accident type can include an object strike accident, a mechanical injury accident, an electric shock accident, a fire accident, a high fall accident, etc., and the safety accident early warning model generated for different accident types is different. Specifically, the risk type output by the safety accident early warning model corresponding to different safety accident types is different. For example, when the safety accident type is a fire accident, the risk type output by the safety accident early warning model is the size and range of the fire. When the safety accident type is a mechanical injury accident, the risk type output by the safety accident early warning model is the type of mechanical equipment and the type of injury (such as splashing or hitting). In some embodiments, the above safety accident early warning model can be obtained by machine learning.
[0040] In some embodiments, in step S200, by dividing the prediction interval (such as a production operation interval), the safety accident early warning model corresponding to the sub-area interval can be better deployed based on the actual situation on site. For example, in a concentrated area of equipment (such as a concentrated area of large operation equipment or production), the probability of occurrence of a mechanical injury accident in this area is much higher than that of other types of safety accidents. That is, it is most necessary to arrange a safety accident prediction model matching the mechanical injury accident type in this area. Further, when dividing the prediction interval into sub-area intervals, the existing house structure can be used for division, for example, it can be divided into production sub-area and office sub-area. In some embodiments, when there are multiple types of safety accidents in a single sub-area interval, a plurality of safety accident early warning models of different types can be set up on the server to ensure accurate prediction of safety accidents, and the plurality of safety early warning models can be independent of each other, which can realize parallel early warning.
[0041] In some embodiments, in step S300, based on the comprehensive and multi-dimensional early warning indicators sorted out, the early warning indicators for different safety accident types are classified and processed, and an early warning indicator system for different safety accident types is constructed. Specifically, the above early warning indicators are as follows:
[0042] The human indicators can include quality data of trained personnel (such as training rate of key post personnel, certificate holding rate of personnel operation, safety training proportion, and full-time safety personnel proportion, etc.).
[0043] The machine indicators can include risk rates of static devices, risk rates of dynamic devices, risk rates of device operation variations, interlocking commissioning parameter data (such as interlocking commissioning rates and interlocking commissioning durations), automatic instrument commissioning parameter data (such as automatic rates of automatic devices, automatic efficiency, and automatic durations), and special operation parameter data (such as types, quantities, operation efficiency, and success rates of special operations (such as special fire operations and special restricted operations)). The risk rates of dynamic devices can be the average risk rates of dynamic devices in different positions or use states, and the risk rates of device operation variations can be the average risk rates of different levels of operation variations.
[0044] The process indicators can include multi-dimensional process parameter data, such as temperature data, pressure data, liquid level data, flammable degree data, and toxic content data.
[0045] The capability indicators include emergency capability parameter data (such as the number of full-time emergency personnel, the number of emergency plans stored, the modification data of emergency plans, the approval rate of emergency parameters, the efficiency of emergency parameters, and the planned completion rate of emergency drills), accident hidden danger investigation parameter data (such as the setting of accident hidden danger investigation positions, the investigation cycle data of accident hidden dangers, the grade data of accident hidden dangers, and the duration of accident hidden dangers), and fire-fighting facility parameter data (such as the damage rate of fire-fighting equipment, the use rate of fire-fighting equipment, the use duration of fire-fighting equipment, and the maximum life of fire-fighting equipment).
[0046] The environmental indicators can include weather parameter data, such as temperature data, rainfall data, snowfall data, wind grade, lightning grade, and the like.
[0047] Based on the above multi-dimensional early warning indicators, the safety accident pre-warning model can comprehensively analyze various risk types that may exist in an enterprise from five aspects of “person, machine, process, capability reserve, and environment”, starting from each risk factor type of personnel configuration, personnel training, hidden danger and risk management, system and plan, emergency capability, special operation risk, key link risk, daily operation, surrounding environment risk, natural disaster risk, device facility risk, leakage, fire and explosion, and poisoning risk, determining risk assessment data of the enterprise according to multi-parameter calculation results, and further determining the risk level and risk type of the enterprise, so as to realize the precision of risk early warning.
[0048] In some embodiments, the above early warning method further includes:
[0049] analyzing different safety accident types to obtain influence factors of different safety accident types, the influence factors constituting a subset of the early warning indicators;
[0050] collecting influence factor data of different safety accident types to obtain training samples;
[0051] Train the safety accident early warning model based on the training samples.
[0052] Based on the above steps, a trained safety accident early warning model can be obtained. Specifically, different early warning indicators can be used for different types of safety accidents. For example, for mechanical injury accidents, the roles of machine indicators and process indicators in the accident prediction process are relatively large. Accordingly, by analyzing and extracting the influencing factors related to different types of safety accidents in the early warning indicators, and collecting the influencing factor data of different types of safety accidents as training samples, the safety accident early warning model is trained until the safety accident early warning model converges, thereby completing the training process of the safety accident early warning model, and using the trained safety accident early warning model as a subsequent prediction model.
[0053] In some embodiments, the above early warning method further comprises:
[0054] Uploading the collected early warning indicators of safety accidents in different sub-regional intervals to a blockchain.
[0055] Specifically, in the above steps, in some embodiments, the operation of uploading the early warning indicators to the blockchain and inputting the early warning indicators into the safety accident early warning model for risk prediction can be performed synchronously, and in other embodiments, the operation of uploading the early warning indicators to the blockchain can be performed before the prediction process of inputting the early warning indicators into the safety accident early warning model for risk prediction. At this time, the early warning indicators input into the safety accident early warning model come from the blockchain, i.e., the safety accident early warning model obtains early warning indicator data from the blockchain, thereby realizing the prediction of the occurrence probability of safety accidents, and enabling early warning and early decision-making of safety accidents based on the predicted occurrence probability. At the same time, uploading the collected early warning indicator data to the blockchain can effectively improve the tamper resistance of the early warning indicator data, which is beneficial to improving the security of the early warning indicator data, and thereby facilitating effective tracing of related early warning indicator data or related safety abnormal accidents.
[0056] In some embodiments, the above early warning method further comprises:
[0057] Setting a risk probability threshold;
[0058] Comparing the risk probability output by the safety accident early warning model with the risk probability threshold;
[0059] When the risk probability output by the safety accident early warning model is greater than the risk probability threshold, an early warning information is generated.
[0060] Specifically, the risk probability threshold can be set correspondingly based on the different types of safety accidents. In some embodiments, the early warning information can include the risk type, the risk probability, and the coping scheme, to facilitate targeted prevention, correction, and rescue operations by workers.
[0061] Another aspect of the present application discloses a production safety early warning system, comprising:
[0062] a classification module configured to classify the safety accident types and build safety accident early warning models corresponding to different safety accident types;
[0063] an erecting module configured to divide the prediction interval into a plurality of sub-regional intervals and erect at least one safety accident early warning model corresponding to the safety accident type of the sub-regional interval on the server in the different sub-regional interval;
[0064] a collection module configured to collect early warning indicators of safety accidents in the different sub-regional intervals, the early warning indicators including human indicators, machine indicators, process indicators, capacity indicators and environmental indicators;
[0065] a prediction module configured to output risk types and risk probabilities through the safety accident early warning model in the sub-regional interval based on the collected early warning indicators in the sub-regional interval.
[0066] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, different embodiments or examples described in the present application and the features of different embodiments or examples can be combined by those skilled in the art without contradiction, if possible.
[0067] Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A production safety early warning method, characterized in that, include: Classify the types of safety accidents and construct early warning models for different types of safety accidents; The prediction interval is divided into multiple sub-regions, and at least one safety accident early warning model corresponding to the safety accident type of the sub-region is set up on the server in each sub-region. Early warning indicators for safety accidents are collected within different sub-regional intervals. These early warning indicators include human-related indicators, machine-related indicators, process-related indicators, capability-related indicators, and environmental indicators. Based on the early warning indicators collected within the sub-region intervals, the risk type and risk probability are output through the safety accident early warning model within the sub-region intervals.
2. The production safety early warning method according to claim 1, characterized in that: The early warning method also includes: Different types of safety accidents are analyzed to obtain the influencing factors of different types of safety accidents, and the influencing factors constitute a subset of the early warning indicators; Collect data on influencing factors of different types of safety accidents to obtain training samples; The safety accident early warning model is trained based on the training samples.
3. A production safety early warning method according to claim 1 or 2, characterized in that: The early warning method also includes: The early warning indicators of safety accidents collected in different sub-regions are uploaded to the blockchain.
4. The production safety early warning method according to claim 3, characterized in that: The early warning indicators of safety accidents collected in different sub-regions are uploaded to the blockchain and simultaneously input into the safety accident early warning model, or are uploaded to the blockchain and then transmitted to the safety accident early warning model by the blockchain.
5. The production safety early warning method according to claim 1, characterized in that: The human data includes the quality data of the trainees, the process indicators include process parameter data, and the environmental indicators include weather parameter data.
6. The production safety early warning method according to claim 1, characterized in that: The machine indicators include the risk rate of static equipment, the risk rate of dynamic equipment, the risk rate of equipment operation changes, equipment interlocking operation parameter data, automatic control instrument operation parameter data, and special operation parameter data.
7. The production safety early warning method according to claim 1, characterized in that: The capability indicators include emergency response capability parameters, accident hazard investigation parameters, and fire protection facility parameters.
8. The production safety early warning method according to claim 1, characterized in that: The types of safety accidents include accidents caused by falling objects, mechanical injuries, electric shocks, fires, and falls from heights.
9. A production safety early warning method according to claim 1, characterized in that: The early warning method also includes: Set a risk probability threshold; The risk probability and risk probability threshold output by the safety accident early warning model are compared; An early warning message is issued when the risk probability output by the safety accident early warning model is greater than the risk probability threshold.
10. A production safety early warning method according to claim 9, characterized in that: The early warning information includes the risk type, risk probability, and response plan.
11. A production safety early warning system, characterized in that: include: The classification module is configured to classify safety accident types and build safety accident early warning models corresponding to different safety accident types. The module is configured to partition the prediction interval to obtain multiple sub-region intervals, and to set up at least one safety accident early warning model corresponding to the safety accident type of the sub-region interval on the server in different sub-region intervals; The data acquisition module is configured to collect early warning indicators for safety accidents in different sub-regions. The early warning indicators include human factors, machine factors, process factors, capability factors, and environmental factors. The prediction module is configured to output risk type and risk probability based on the early warning indicators collected within the sub-region interval through the safety accident early warning model within the sub-region interval.