CN118171916A - Accident cause analysis method based on complex network theory - Google Patents
Accident cause analysis method based on complex network theory Download PDFInfo
- Publication number
- CN118171916A CN118171916A CN202410585855.9A CN202410585855A CN118171916A CN 118171916 A CN118171916 A CN 118171916A CN 202410585855 A CN202410585855 A CN 202410585855A CN 118171916 A CN118171916 A CN 118171916A
- Authority
- CN
- China
- Prior art keywords
- node
- network
- cause
- accident
- nodes
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 97
- 238000000034 method Methods 0.000 claims abstract description 68
- 238000005065 mining Methods 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims abstract description 14
- 238000009826 distribution Methods 0.000 claims description 21
- 230000001364 causal effect Effects 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 abstract description 54
- 230000002265 prevention Effects 0.000 abstract description 4
- 238000011160 research Methods 0.000 description 15
- 238000013461 design Methods 0.000 description 11
- 230000003993 interaction Effects 0.000 description 9
- 238000004422 calculation algorithm Methods 0.000 description 7
- 238000004451 qualitative analysis Methods 0.000 description 7
- 238000004445 quantitative analysis Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 239000000463 material Substances 0.000 description 6
- 230000009897 systematic effect Effects 0.000 description 6
- 238000007418 data mining Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 4
- 230000006698 induction Effects 0.000 description 4
- 238000012502 risk assessment Methods 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 230000002776 aggregation Effects 0.000 description 3
- 238000004220 aggregation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000001419 dependent effect Effects 0.000 description 3
- 238000005315 distribution function Methods 0.000 description 3
- 238000011835 investigation Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 238000012800 visualization Methods 0.000 description 3
- 229910000831 Steel Inorganic materials 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000009430 construction management Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013138 pruning Methods 0.000 description 2
- 238000011158 quantitative evaluation Methods 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 239000010959 steel Substances 0.000 description 2
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000012098 association analyses Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 238000009435 building construction Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000001351 cycling effect Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000013213 extrapolation Methods 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000013077 scoring method Methods 0.000 description 1
- 238000010206 sensitivity analysis Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to an accident cause analysis method based on complex network theory, comprising the following steps: acquiring accident cause factors, associating the accident cause factors according to the accident occurrence process to form an event chain, and taking the accident cause factors in the event chain as nodes to carry out directed weighting treatment to form a cause network; carrying out characteristic analysis on the cause network to obtain a characteristic analysis result; the characteristic analysis includes: node characteristic analysis, node community characteristic analysis and network overall characteristic analysis; carrying out association rule mining on the cause network to obtain association strength; and forming an accident potential list based on the characteristic analysis result and the association strength. The invention can comprehensively and effectively analyze the safety cause of various engineering construction, eliminates the existing hidden danger of omission, can analyze the importance of the accident cause, and can effectively judge the emphasis point during accident prevention.
Description
Technical Field
The invention relates to the technical field of accident analysis, in particular to an accident cause analysis method based on a complex network theory.
Background
The construction hidden danger judgment is to analyze the safety risk of the construction site, and aims to find out factors and rules possibly causing risk occurrence, thereby reducing the probability of risk occurrence and controlling the hazard brought by the risk occurrence. The safety risk analysis is always carried out in each link of complex system design, operation and maintenance, and the main research contents comprise: identifying possible risk factors before the risk event does not occur; how to effectively control the spread of risks and reduce the consequences caused by the risks when the risks occur; the defect in the system is timely made up after the accident occurs, and the safety of the system is improved by searching the law of the accident. The safety risk analysis is an important means for ensuring the normal and safe operation of each element in the complex system, and has been widely applied to railway transportation management, road transportation management, large-scale building construction management, aerospace system management and the like. Based on different basic theories and research backgrounds, scholars propose various analysis methods for safety risks, and the analysis methods can be classified into qualitative analysis methods and quantitative analysis methods according to formalized differences of the obtained results. The qualitative analysis method is a traditional hidden danger analysis and judgment mainstream method, and mainly researches the possible safety risks in the system through judging by related professionals, objectively regular or summarized experiences in the happened events and other methods, and achieves the aim of risk prevention and control by revealing the incident occurrence rule. The qualitative analysis method mainly comprises the following steps: a systematic accident model method, a safety check list method, an expert evaluation method, a fault analysis method and the like. The qualitative analysis method can effectively classify complicated information by means of deduction, reasoning, summarization, induction and the like, so that accident induction factors and risk propagation processes are discovered. The systematic accident model provides a powerful tool for describing and describing complex systems, and plays an important role in modeling the systems and determining the relationship between factors. However, large construction scenes often imply various complex factors, and under the combined action of various accident causes, it is difficult to analyze and generalize the interaction relationship of the factors and the context mechanism of the accident by experience. Meanwhile, the hidden danger types are classified and distributed in a grading way through an effective statistical calculation method, so that limited resources for hidden danger management can be utilized to the greatest extent. Therefore, when a modern bridge construction major safety production accident hidden trouble judging and early warning system is constructed, the method is not applicable only by means of the traditional qualitative analysis method.
Unlike qualitative analysis, quantitative analysis is a method for analyzing the number characteristics, number relationships, and number change rules of elements in a system, and aims to obtain accurate factor importance and quantitative results of the system safety state. The quantitative analysis method can obtain an accurate value of the importance degree of the factors, further identify more important factors from a plurality of factors, and evaluate the risk level of the whole system and the subsystem at the same time, so that the safety state of the system can be accurately judged. Although the two methods differ, quantitative analysis needs to be performed on the basis of qualitative analysis. The analysis purpose and the required factors are determined by analyzing the whole system, the data sources for finishing the required quantization indexes are collected, an analysis model is built based on the theory of quantitative analysis, and then the quantization research is carried out. The quantitative analysis method mainly comprises the following steps: fault tree analysis method, expert scoring method, extrapolation method, sensitivity analysis method, decision tree analysis method, bayesian network analysis method, complex network and related theory analysis method, data mining, etc.
Existing research considers complex networks as an effective tool for quantitatively describing complex systems and is applied to security risk analysis. Complex networks are widely used as a powerful analytical tool for studying various types of complex systems. The complex network is actually a graph model with a complex topological structure and is composed of nodes with huge data volume and complicated relations among the nodes. The method is applied to the railway construction field, such as railway networks, railway transportation reliability, railway transportation accidents and the like. Since bridge construction systems are also a huge system, they contain numerous personnel, structures, equipment, environments, and a great deal of interactions between them. The factors causing the bridge construction accident are numerous, and the relationship between the factors is complicated, so that the complex network can be used for analyzing the hidden trouble of the bridge construction accident. However, the complex network model of railway accidents established by the prior studies are mainly of two types: an unlicensed network or an undirected network. The causal relationship or the strength of the relationship between the two network unaccounted factors is better for accurately acquiring the valuable information in the accident-causing network by constructing the directional weighting network.
Data mining is generated with the development of computer technology and information technology, and as a large amount of data is generated, how to find valuable information by analyzing the data has important significance in bringing profit to enterprises and improving the production and life of people. The data mining method can be classified into classification, estimation, prediction, association analysis, cluster analysis and the like according to the analysis purpose and the difference of data characteristics. Data mining is also widely used in security management in industries such as nuclear energy, transportation, personal injury and death, and the like. Correlation analysis is an important method for analyzing the strength of relationships between items or objects, and has a great deal of application in security risk analysis. The complex bridge construction system comprises a large number of elements such as workers, machine equipment, structural members, construction processes, site management, external environments and the like. The factors causing accidents are numerous and quite complex, and the association rules obtained by mining from the database are numerous. In the existing method for analyzing accidents by using association rules, most of research objects are structured accident data. Moreover, the analysis factors are limited to weather, time, place and other factors, and the analysis of the complex factors such as people, machines, rings, pipes and the like is lacked. However, the factors are closely related to the occurrence of accidents in actual operation management, and the incidence relation among the factors is found to be more beneficial to improving the safety of the system. Therefore, association rule analysis for this case remains to be studied.
Disclosure of Invention
The invention aims to provide an accident cause analysis method based on complex network theory, which is mainly based on the mode of manually making and executing a construction management scheme for the current engineering construction management level, and has lower reliability, the method can comprehensively and effectively analyze various engineering construction safety causes, eliminate the existing hidden danger, analyze the importance of the accident cause and effectively judge the important point when the accident is prevented.
In order to achieve the above object, the present invention provides the following solutions:
an accident cause analysis method based on complex network theory comprises the following steps:
Acquiring accident cause factors, associating the accident cause factors according to the accident occurrence process to form an event chain, and taking the accident cause factors in the event chain as nodes to perform directed weighting treatment to form a cause network;
Performing characteristic analysis on the cause network to obtain a characteristic analysis result; the characteristic analysis includes: node characteristic analysis, node community characteristic analysis and network overall characteristic analysis; carrying out association rule mining on the cause network to obtain association strength;
and forming an accident potential list based on the characteristic analysis result and the association strength.
Optionally, performing the directional weighting processing with the accident cause factor in the event chain as the node includes:
and taking the accident cause factors in the event chain as nodes, if the accident cause factors in the event chain have a causal relationship, connecting the nodes of the accident cause factors and the nodes through the directed edges, determining the occurrence times of the accident cause factors, namely the weight of the directed edges, based on the causal relationship, and setting the weight on the corresponding directed edges.
Optionally, performing node characteristic analysis on the cause network includes:
the node degree analysis method for the cause network comprises the following steps:
Wherein, For the degree of departure of a node, i.e. the number of edges from the start to the end of the node,/>For the present node,/>For all other nodes except this node,/>To be node/>As a starting point, node/>Side being end point,/>For the ingress of a node, i.e. the number of edges from the end point to the start point of the node,/>To be node/>As a starting point, node/>Edges that are end points;
Performing a degree distribution analysis on the cause network includes: through shape like Analyzing the degree distribution of the causal network;
the method for analyzing the node strength of the cause network comprises the following steps:
Wherein, The intensity of the exit of this node, i.e. the sum of the weights of the edges of this node pointing to other nodes,/>, is calculatedThe ingress strength of this node, i.e. the sum of the weights of the edges pointing to this node by other nodes,/>For the present node,/>For all other nodes except this node,/>To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, a nodeWeighting the edges that are the end points;
The node intensity distribution analysis of the cause network comprises the following steps: through shape like Is used for analyzing the node intensity distribution of the cause network.
Optionally, performing node community characteristic analysis on the cause network includes:
the method for clustering the cause network comprises the following steps:
Wherein, For node/>Cluster coefficient of/>、/>And/>Three adjacent nodes of every twoFor node/>Node strength of/>For node/>Degree of (v)/(v)To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Side being end point,/>To be node/>As a starting point, node/>Side being end point,/>To be node/>As a starting point, node/>Edges that are end points;
Performing a cause network for a cause network comprises:
Wherein, Is the betweenness of the node,/>For connecting nodes/>, in a networkAnd/>All shortest paths through nodes/>Number of,/>For connecting nodes/>, in a networkAnd/>Number of all shortest paths,/>Representing the entire network.
Optionally, the analyzing the network overall characteristic of the cause network includes:
The method for analyzing the network diameter of the cause network comprises the following steps:
Wherein, For the network diameter,/>To connect these two nodes in the network, the number of edges on the shortest path that is traversed,/>For nodes,/>Is a node;
the method for analyzing the network average path length of the cause network comprises the following steps:
Wherein, For the average path length of the network,/>To connect these two nodes in the network, the number of edges on the shortest path that is traversed,/>Is the total number of nodes in the network.
Optionally, performing association rule mining on the causal network includes:
Taking two nodes connected in the cause network as project sets, and acquiring a result project set according to the connection times;
according to the occurrence times of the same item sets, obtaining a support degree, setting a support degree threshold, deleting the same item sets in the result item sets with the support degree smaller than the support degree threshold, and obtaining the latest result item set;
And carrying out association rule mining on the latest result item set to obtain association strength, namely supporting degree, confidence degree and lifting degree.
Optionally, the method for obtaining the support degree is as follows:
Wherein, For connection/>To/>Support of/>For connection/>To/>Number of occurrences,/>Is a leading node,/>Is a successor node.
Optionally, the method for obtaining the confidence coefficient is as follows:
Wherein, For connection/>To/>Confidence of/>For connection/>To/>Frequency of occurrence and node/>Ratio of occurrence times of/(v)For all containing nodes/>Item set of (i.e., node/>)Is the number of occurrences of (a).
Optionally, the method for obtaining the lifting degree is as follows:
Wherein, For connection/>To/>Degree of elevation of/>For all containing nodes/>Item set of (i.e., node/>)Is the number of occurrences of (a).
The beneficial effects of the invention are as follows:
Systematic and comprehensive: the traditional research on accident hidden danger or the research on safety management usually depends on expert experience or certain known factors to analyze, the possibility and influence of accident occurrence can be known only from local or specific angles, the evolution process of bridge construction accidents can be abstracted into a complex network, the development rule and key influence factors of the accidents are researched from the aspects of integrity and systemicity, the possibility and influence of the accident occurrence can be known more comprehensively, the cause factors of related construction accidents are summarized more comprehensively and systematically through the analysis of cause network and association rule mining, the evolution venation of the accident is cleared, and the importance of different cause factors is well differentiated.
Objectivity and scientificity: the traditional research on accident hidden danger or the research on safety management is often dependent on subjective judgment and experience summary, and certain subjectivity and uncertainty exist, but the invention extracts the topological structure and characteristic parameters of a network by excavating and analyzing massive data, analyzes the accident evolution rule by adopting a plurality of quantitative indexes such as degree and degree distribution, connectivity, shortest path, mediates, clustering coefficients, association rule related indexes and the like, and quantitatively analyzes the accident evolution rule, so that the quantitative analysis result is more systematic and accurate, the evolution rule and key influence factors of the accident are researched from objective and scientific angles, the interference of subjective factors can be reduced, and the reliability and the scientificity of the research are improved.
Flexibility and visualization: the invention can present the network structure and the key nodes in a graphic mode, and is convenient for observation and analysis. Meanwhile, the method based on the complex network has flexibility, can be adjusted and improved according to different research requirements and data characteristics, and improves the applicability and flexibility of research.
Prospective and predictive: traditional researches on accident potential or safety management often can only be summarized and analyzed after accidents have occurred, and the prospective and predictive performance are lacked. According to the invention, the development trend and potential hidden trouble of bridge construction accidents can be predicted by researching the evolution law of the network, the development evolution of the accidents can be deduced, the accidents and causative factors can be distinguished more accurately, a plurality of new scenes which are not easy to perceive by the traditional manual method are generated, and scientific basis is provided for the prevention and control of the accidents.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of an accident cause analysis method based on a complex network theory according to an embodiment of the invention;
FIG. 2 is a diagram illustrating an exemplary determination of nodes and directed weighted edges according to an embodiment of the present invention;
Fig. 3 is a schematic diagram of a high altitude overhead operation cause network according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
The invention discloses an accident cause analysis method based on a complex network theory, which comprises the following steps: acquiring accident cause factors, associating the accident cause factors according to the accident occurrence process to form an event chain, and taking the accident cause factors in the event chain as nodes to carry out directed weighting treatment to form a cause network; carrying out characteristic analysis on the cause network to obtain a characteristic analysis result; the characteristic analysis includes: node characteristic analysis, node community characteristic analysis and network overall characteristic analysis; carrying out association rule mining on the cause network to obtain association strength; based on the characteristic analysis result and the association strength, forming an accident potential list, specifically:
The process comprises five steps of A to E, wherein step A firstly carries out characteristic investigation and characteristic analysis of typical accidents in the construction period, and comprises (1) collecting accident data, (2) sorting and analyzing the data; step B, carrying out engineering construction accident hidden trouble identification and attribute calibration, including (3) classifying accident causes and (4) extracting the accident causes; step C, modeling and analyzing the directed weighted accident cause network, including (5) forming an event chain, (6) forming a cause network, (7) node characteristic analysis, (8) community characteristic analysis, and (9) network overall characteristic analysis; step D, carrying out association rule mining based on construction hidden danger of the Apriori algorithm; the method comprises the steps of (10) a connection step, (11) a pruning step, (12) calculation of support, confidence and promotion degree, and (13) definition of a strong association rule; and E, finally forming a hidden danger list and an index system of the major construction accident, wherein the step comprises the steps of (14) forming a hidden danger list and (15) forming a hidden danger index system.
Step A typical accident feature investigation and feature analysis in construction period (1) collecting accident data, wherein the sources of the collected accident case data comprise but are not limited to reports of news media, officially disclosed accident data and descriptions of construction accidents related to related research papers, and the data types comprise texts, videos, audios and the like.
The method related to the (2) data arrangement and analysis in the characteristic investigation and characteristic analysis of typical accidents in the construction period of the step A comprises image recognition, natural language processing, manual screening, standard comparison and the like, and raw data are processed into structured data.
And B, classifying methods used for classifying the accident causes in the engineering construction accident hidden trouble recognition and attribute calibration (3) comprise a systematic accident analysis method, a 2-4 model accident cause classifying method, a 4M cause classifying method and the like. The causes and accidents are classified into human factors, design factors, equipment factors, environmental factors, management factors, accidents and the like.
And B, in engineering construction accident hidden trouble identification and attribute calibration, accident causes are extracted (4), and on the basis of classifying the accident causes (3), the causes are subdivided into observable or measurable behaviors, including but not limited to personnel illegal operation, unqualified machine tool quality, structural problems and the like.
Step C, forming an event chain in the modeling and analysis of the directed weighted accident cause network, and extracting and summarizing the event chain which describes the accident occurrence process and contains all key information by analyzing the accident occurrence process, key cause factors, potential factors, induction factors, possible factors and the like.
The method for performing directed weighting processing by taking the accident cause factors in the event chain as nodes comprises the following steps: and taking the accident cause factors in the event chain as nodes, if the accident cause factors in the event chain have a causal relationship, connecting the nodes of the event chain and the accident cause factors through the directional edges, determining the occurrence times of the accident cause factors, namely the weights of the directional edges based on the causal relationship, and setting the weights on the corresponding directional edges, as shown in fig. 2.
The association rule mining of the causal network comprises the following steps: the method comprises the steps of using two nodes connected in a cause network as item sets, comparing the occurrence times of the same item sets with each other, connecting the same item sets to generate a plurality of candidate item sets, and connecting the candidate item sets according to the order of the occurrence times to obtain a result item set; according to the occurrence times of the same item sets, obtaining the support degree, setting a support degree threshold value, deleting the same item sets in the result item sets with the support degree smaller than the support degree threshold value, and obtaining the latest result item set; carrying out association rule mining on the latest result item set to obtain association strength, namely supporting degree, confidence degree and lifting degree, wherein the association strength comprises the following specific steps:
And C, forming a cause network by (6) in the modeling and analysis of the directed weighted accident cause network, coding an event chain according to a coding format of the directed graph, taking the accident causes as nodes, and taking the mutual causing relationship between the causes and the accidents as edges to realize the visualization of the cause network and the analysis of network characteristics. The structure of the network is represented by an adjacency matrix, one by Directed weighted network of individual nodesMatrix/>Expressed as: /(I)Wherein if node/>Pointing node/>Then/>，/>Representing nodes/>Pointing node/>Strength of (2); otherwise the first set of parameters is selected,，/>。
And C, performing (7) node characteristic analysis in modeling and analysis of the directed weighted accident cause network, wherein the node characteristic analysis comprises node degree and degree distribution and node strength distribution analysis, the degree of a node is reflected by the number of other nodes directly related to the node, the node strength comprehensively considers the number of the nodes directly related to the node and the strength of the relationship between the nodes, the distribution of the degree and the strength is an important index for reflecting the network characteristic, and the network can be divided into a regular network, a random network, a small world network, a non-scale network and the like through the difference of the degree distribution. The degree of egress of a node is the number of edges starting with that node and ending with other nodes, such as: The ingress of a node refers to the number of edges that terminate with the node and start with other nodes, namely: /(I) Wherein/>For the degree of departure of a node, i.e. the number of edges from the start to the end of the node,/>For the present node,/>For all other nodes except this node,/>To be node/>As a starting point, node/>Side being end point,/>For the ingress of a node, i.e. the number of edges from the end point to the start point of the node,/>To be node/>As a starting point, node/>Is the edge of the endpoint.
The total degree of a node is the sum of the outbound degree and the inbound degree of the node, namely: the degree distribution of nodes is shaped as/> Is represented by a distribution function of (a). The implementation method of the degree distribution analysis comprises the following steps: ordering the degrees of all nodes from big to small, taking the degrees of the nodes as independent variables, and taking the number of times of occurrence of specific numerical values of the degrees as dependent variables to form the/>Is a function fit of both, wherein/>And/>Is a constant, fitting the resulting function/>I.e. the distribution function of the node degree.
The outgoing strength of a node is the sum of the weights of the edges pointing from this node to other nodes, namely: The node ingress strength is the sum of the weights of the edges of the other nodes pointing to the node, namely: /(I) Wherein/>The intensity of the exit of this node, i.e. the sum of the weights of the edges of this node pointing to other nodes,/>, is calculatedThe ingress strength of this node, i.e. the sum of the weights of the edges pointing to this node by other nodes,/>For the present node,/>For all other nodes except this node,/>To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Weighting the edges that are the end points;
The total intensity of a node represents the sum of the outgoing and incoming intensities of the node, i.e Distribution of node intensities is shaped as/>Is described. The method for realizing the intensity distribution analysis comprises the following steps: ordering the intensities of all nodes from big to small, taking the intensity of the nodes as independent variables, and taking the number of times of occurrence of specific numerical values of the intensity as dependent variables to form the modelIs a function fit of both, wherein/>And/>Is a constant, fitting the resulting function/>I.e. the distribution function of the node intensity.
And C, directionally weighting community characteristic analysis in modeling and analysis of the accident cause network, wherein the community characteristic analysis comprises cluster coefficient and betweenness analysis, and the characteristic of aggregation exists among nodes in the network. There may also be an association between other nodes associated with one node, reflecting that they are more prone to be clustered together. This characteristic is represented by a cluster coefficient, which reflects the coefficient of the degree of aggregation of nodes in a network. In a directional weighted network, nodes need to be pairedAnd node/>The directed edges between them are processed. If/>Or/>Then/>; The weight of an edge is/>The treatment method comprises the following steps of. Node/>Clustering coefficient/>: The calculation can be performed by the following formula: Wherein/> For node/>Cluster coefficient of/>、/>And/>Three adjacent nodes of every twoFor node/>Node strength of/>For node/>Degree of (v)/(v)To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Side being end point,/>To be node/>As a starting point, node/>Side being end point,/>To be node/>As a starting point, node/>Side being end point,/>Representing node/>Is a degree of (3). Interactions between nodes in the network are mainly through shortest paths. Depending on the situation of the nodes through which the shortest path passes, this leads to a differentiation of the degree of importance of the nodes in the network, which is reflected by the betweenness. On the network, the betweenness/>, of the nodesRefers to the transit node/>, among all shortest pathsThe ratio of the number of paths to the total number of shortest paths can be calculated by the following formula: /(I)Wherein/>Representing a connection node/>, in a networkAndNumber of all shortest paths,/>Representing a connection node/>, in a networkAnd/>Through nodes in all shortest paths of (a)Number of,/>Representing the entire network.
And C, performing (9) network overall characteristic analysis in modeling and analysis of the directed weighted accident cause network, wherein the network overall characteristic analysis comprises a network diameter and an average path length, and the shortest path in the network represents a path with the minimum sum of all side weights in paths from one node to the other node. In the bridge construction accident cause network established by the invention, the larger the value of the edge is, the closer the relationship between representing nodes is or the more frequent the interaction is. In this case, the inverse of the weight of the edge is calculated, and then the Dijkstra algorithm is used to calculate the shortest path. In a network, a nodeAnd another node/>Distance of/>And (3) representing. Wherein, by/>Representing the number of edges on the shortest path that will be traversed when connecting the two nodes in the network. The maximum distance between all node pairs on the network represents the diameter of the network, use/>The representation is: /(I)To reflect the average speed of information propagation or interaction between nodes in a network, the average path length/>, is used to determine the average speed of information propagation or interaction between nodes in the networkMeasured, i.e./>Wherein/>Is the total number of nodes in the network.
And step D, the step (10) of connection step in the association rule mining based on the Apriori algorithm construction hidden danger is to scan the whole database, take every two nodes connected in the cause network as candidate item sets, and take two nodes connected in the cause network as item sets, and obtain a result item set according to the order of the connection times. And D, according to the step (11) of pruning in the association rule mining based on the construction hidden danger of the Apriori algorithm, acquiring the support degree according to the occurrence times of the same item set, setting a support degree threshold, deleting the same item set in the result item set with the support degree smaller than the support degree threshold, acquiring the latest result item set, repeatedly cycling in this way, traversing the database, and ending the algorithm when no new item set is output.
And D, calculating the support, the confidence and the lifting degree based on the (12) in the association rule mining of the construction hidden danger of the Apriori algorithm. The support represents a set of items among all itemsProbability of occurrence: /(I). Confidence means prerequisite/>On the premise of occurrence, another transaction/>Probability of occurrence: . Promotion is the ratio of the pre-term confidence to post-term support of the association rule, i.e., transaction/> Under the conditions of occurrence, simultaneously contain/>Probability and/>Ratio of probabilities of overall occurrence: ; wherein/> For connection/>To/>Support of/>For connection/>To/>Number of occurrences,/>Is a leading node,/>For successor node,/>For connection/>To/>Confidence of/>For connection/>To/>Frequency of occurrence and node/>Ratio of occurrence times of/(v)For all containing nodes/>Item set of (i.e., node/>)Frequency of occurrence of/>For connection/>To/>Degree of elevation of/>For all containing nodes/>Item set of (i.e., node/>)Is the number of occurrences of (a).
And D, defining a strong association rule based on the (13) in the association rule mining of the Apriori algorithm construction hidden danger, and calculating the support degree, the confidence degree and the lifting degree for every two adjacent factors. And screening out a plurality of association rules (for example, front 20) in front of the support degree bit column, defining the association rules as strong association rules, and carrying out preliminary statistical analysis on the front and rear factors appearing in the strong association rules obtained through final screening so as to know key factors influencing accident safety.
And E, forming a hidden danger list in the construction major accident hidden danger list and the index system (14), forming an accident hidden danger list, wherein the hidden danger list is divided into a plurality of major categories according to hidden danger categories, hidden danger matters needing to be concerned are divided into two categories of important matters and general matters, the important matters need to be strictly managed and evaluated from the aspects of design, management and the like, the general matters need to be strengthened and supervised, and the constraint is carried out through construction safety culture. The specific content is expressed in detail after each hidden danger item, so that construction managers can conveniently control, examine and manage the hidden accidents on the construction site.
And E, forming a hidden danger index system by forming a hidden danger list of the construction major accident and (15) in the index system, analyzing the importance degree of parameter indexes in each hidden danger list on the bridge construction safety, prescribing and classifying parameters in the hidden danger list according to hidden danger grades, and providing a major accident hidden danger list parameter index grading standard, wherein the bridge major accident hidden danger index system consists of 11 indexes such as personnel states, construction operations, material quality, component states, structural systems, construction organizations, field management, structural equipment, machines, geological conditions, climate conditions and the like, and can be used for subsequent quantitative evaluation.
The embodiment provides an accident cause analysis flow under a bridge overhead working scenario. Firstly, collecting related data of construction accidents of overhead working at the high position of a bridge, and collecting a plurality of cases of the construction accidents of the bridge related to the overhead working, wherein the case data mainly come from official data and news media data.
And further processing the factors by using a systematic accident analysis method. The whole system is divided into 6 categories, namely: human factor, design factor, equipment factor, environmental factor, management factor, and incident occurrence. The human factor includes psychological, physiological, behavioral and other factors of personnel in the system, and the personnel involved are electrician, electric welder, steel reinforcement, concrete, motor vehicle driver, machine operator, machine maintenance worker, mixer operator, derrick driver, hoist driver, plasterer, woodworker, common worker, gas welder, tower crane driver, earthwork machine driver and the like. The design factors comprise structural design, materials and 3 layers of components, and relate to different materials such as concrete, steel, new materials and the like, and design factors such as various components such as girders, piers, bridge towers and the like, environmental survey, system design and the like. The equipment factors include all the situations of faults or damages in equipment or facilities related to bridge construction, and the related machine equipment comprises: a bracket and related equipment, a template and related equipment, lifting equipment, concrete construction equipment, prestress tensioning equipment and the like. Environmental factors include security risks in terms of natural environment, geological environment, personnel work environment, and the like. Management factors include problems in the management process, which relate to regulations, supervision, technical processes, security education, skill training, equipment maintenance machine repair, risk emergency, security culture, and the like. Incidents include various types of incidents or security risk events. By combining the characteristics of data and the characteristics of the system, the whole system is researched in a classified and layered mode, all elements in the system are analyzed, and preparation is made for constructing a bridge construction accident cause network.
Further, according to the description of the direct factors, the induction factors, the possible factors, the potential factors and the like in the accident data, the collected overhead temporary space related accident reports are analyzed one by one, and finally 49 accident cause factors are extracted, wherein the results are shown in the table 1: the human factor comprises 13 factors, the structural factor comprises 13 factors, the equipment factor comprises 6 factors, the environment factor comprises 12 factors, the management factor comprises 5 factors, and the accident comprises 10 accidents or risk events. All factors are numbered and are shown to the left of the factor.
TABLE 1
And further analyzing the bridge construction accident report and extracting an accident event chain. The first column in the table represents the incident report number. The second column is to extract and generalize from the accident descriptions by analyzing the accident occurrence process, key cause factors, potential factors, inducement factors, possible factors, etc., which contain all the key information that led to the accident occurrence. The third column is the chain of events extracted from the event description. By analyzing the event chain, and associating events with 49 causative factors. Finally, in the fourth column, the event chain is represented with the number of factors. For all incident reports, the analysis was performed as above and the chain of events was obtained as in table 2.
TABLE 2
And further, the event chain is encoded according to the encoding format of the directed graph, the accident causes are used as nodes, the mutual causing relationship between the causes and the accident are used as edges, and the visualization and the network characteristic analysis of the high-altitude overhead operation cause network are realized, as shown in fig. 3.
Further node characteristic analysis is performed. The factors corresponding to the nodes in the front 5 of the total degree bit column are respectively: local component problem (D07), implement emergency (F05), violation of safety construction (H10), in-situ management failure (M04), and bracket reliability failure (F04), the values are respectively: 12, 10, 10, 10 and 7. The 4 factors with the greatest output are respectively: the field management out-of-place (M04), the local component problem (D07), the implement emergency (F05) and the violation safe construction (H10) values are respectively as follows: 10,5,5,5. The 5 factors with the greatest incidence are respectively: local component problem (D07), implement emergency (F05), violation of safety construction (H10), and bracket reliability deficiency (F04), the values are respectively: 7,5,5,5. These larger scale factors play a more critical role in the occurrence of accidents because they are directly causally related to more factors than others. Once an abnormality occurs in the system, these factors are more susceptible to or dangerous to other factors, with the result that unsafe factors in the system increase dramatically, resulting in a serious decrease in the overall safety of the system. The cumulative degree distribution obeying function of the network model is as followsThe power law distribution of (c) illustrates that this network has the characteristics of a scaleless network. The degree values of most nodes in the network are smaller, and the degree values of less nodes are larger. Factors with a large small index value are related to a large cause relationship. If this part of the factors with larger values are controlled and safeguarded, the cause of this is that the network will become very fragile and will break down into several small sub-networks. This will greatly reduce the interplay capability between the fault factors and the propagation scale of the security risk, thereby improving the overall system security.
The factors corresponding to the nodes in front 5 of the intensity bit column are respectively: violating safety construction (H10), support reliability deficiency (F04), local component problem (D07), implement emergency (F05) and crane operation failure (H04) are more important than other accident types, and the total strength values of the nodes are respectively: 29, 27, 16, 16, 13. The factor with the larger total intensity value of the nodes is higher than other factors, so that the factors are controlled in a major way in safety management. The cumulative intensity distribution obeying function of the network model is as followsIs a power law distribution of (c). The results show that most of the nodes in the network have smaller intensity values, and fewer nodes have larger intensity values. By controlling this few nodes, the frequency of influence between causative factors can be effectively reduced.
And further analyzing node community characteristics. Factors with the largest cluster coefficients include: the method comprises the steps of preventing work from being out of place (M05), preventing the reliability of templates and related machines from being insufficient (F01), preventing the reliability of cranes from being insufficient (F02), preventing the reliability of bridge girder erection machine from being insufficient (F03), preventing the design of a bracket from being wrong (D03), preventing the design of a system from being wrong (D04) and preventing a support from being invalid (D09). The factors with larger clustering coefficients have stronger correlation with neighbor factors, and once the factors are abnormal, the neighbor factors are easy to cause risks. Therefore, in order to prevent the occurrence of the chain reaction in the network, it is necessary to control the nodes with relatively high cluster coefficients. Among these factors of relatively large cluster coefficients, the machine and device layers and the structural factor layer occupy relatively large areas. These two types of factors are illustrated to have stronger factor aggregation properties than other layers.
Factors with the largest betweenness include: local component problem (D07), construction failure (H07), material index failure (D06), violation of safety construction (H10), implement emergency (F05) and site emergency problem (E05). These nodes with larger mediates play a key role in the process of interaction between factors in the system, and their existence makes the influence between factors quicker and the propagation efficiency of security risks higher. The nodes are effectively controlled, the occurrence probability of the nodes is reduced, the influence speed among factors is greatly reduced, and the whole system is safer.
And further analyzing the overall characteristics of the network. The average path length in the accident-causing network is an indicator of the strength of the interaction between the factors. Shorter average path lengths represent that the effect between factors can be achieved with fewer factors. In a practical engineering system, abnormal factors affect their neighbors with different probabilities, with the neighbors that interact more closely or more weakly being more susceptible. The average path length of the network is calculated to be 1.844, which illustrates that in the network, a factor can reach any one factor in the incident type layer in less than 2 steps on average. This means that when an anomaly occurs in a system element, on average only about 2 steps are required to cause an accident or serious risk event to occur. Furthermore, the diameter of the network was calculated to be 4. The network has a shorter average path length and a smaller diameter, which indicates that it has the characteristics of a small world network. This reveals the root cause of the difficulty of accident prevention.
And further performing association rule mining. And calculating the support, confidence and promotion degree for every two adjacent factors. The set support threshold was 0.02 and the confidence threshold was 0.15. And screening the association rule of the front 20 of the support degree bit column, defining the association rule as a strong association rule, and selecting the front and rear factors appearing in the final screened strong association rule for preliminary statistical analysis so as to know key factors influencing accident safety. The specific statistical results are shown in table 3. By observing the accident cause types, the occurrence frequency of the factors related to the design factors and the equipment factors is more, and the occurrence frequency is 9 times and 8 times respectively. The frequency of the management factor and the human factor is up to 4 times and 3 times respectively. This suggests that all four types of factors are susceptible to other factors, thus creating a risk.
TABLE 3 Table 3
Further forming a major accident potential list in a high-altitude temporary construction scene, analyzing the importance degree of parameter indexes in each potential hazard list on bridge construction safety aiming at a classification list, prescribing and classifying parameters in the list according to potential hazard grades, and providing a major accident potential list parameter index classification standard, wherein a beam major accident potential index system consists of 11 indexes such as personnel states, construction operations, material quality, component states, structural systems, construction organizations, field management, structural equipment, machines, geological conditions, climate conditions and the like, and can be used for subsequent quantitative evaluation, wherein table 4 is a major accident potential list for highway bridge construction.
TABLE 4 Table 4
The above embodiments are merely illustrative of the preferred embodiments of the present invention, and the scope of the present invention is not limited thereto, but various modifications and improvements made by those skilled in the art to which the present invention pertains are made without departing from the spirit of the present invention, and all modifications and improvements fall within the scope of the present invention as defined in the appended claims.
Claims (9)
1. An accident cause analysis method based on complex network theory is characterized by comprising the following steps:
Acquiring accident cause factors, associating the accident cause factors according to the accident occurrence process to form an event chain, and taking the accident cause factors in the event chain as nodes to perform directed weighting treatment to form a cause network;
Performing characteristic analysis on the cause network to obtain a characteristic analysis result; the characteristic analysis includes: node characteristic analysis, node community characteristic analysis and network overall characteristic analysis; carrying out association rule mining on the cause network to obtain association strength;
and forming an accident potential list based on the characteristic analysis result and the association strength.
2. The method for analyzing accident cause based on the complex network theory according to claim 1, wherein the step of performing the directional weighting processing on the accident cause factors in the event chain as the nodes comprises the steps of:
and taking the accident cause factors in the event chain as nodes, if the accident cause factors in the event chain have a causal relationship, connecting the nodes of the accident cause factors and the nodes through the directed edges, determining the occurrence times of the accident cause factors, namely the weight of the directed edges, based on the causal relationship, and setting the weight on the corresponding directed edges.
3. The method for analyzing the cause of accident based on the theory of the complex network according to claim 1, wherein the analysis of the node characteristics of the cause network comprises:
the node degree analysis method for the cause network comprises the following steps:
Wherein/> For the degree of departure of a node, i.e. the number of edges from the start to the end of the node,/>For the present node,/>For all other nodes except this node,/>To be node/>As a starting point, node/>Side being end point,/>For the ingress of a node, i.e. the number of edges from the end point to the start point of the node,/>To be node/>As a starting point, node/>Edges that are end points;
Performing a degree distribution analysis on the cause network includes: through shape like Analyzing the degree distribution of the causal network;
the method for analyzing the node strength of the cause network comprises the following steps:
Wherein/> The intensity of the exit of this node, i.e. the sum of the weights of the edges of this node pointing to other nodes,/>, is calculatedThe ingress strength of this node, i.e. the sum of the weights of the edges pointing to this node by other nodes,/>For the present node,/>For all other nodes except this node,/>To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Weighting the edges that are the end points;
The node intensity distribution analysis of the cause network comprises the following steps: through shape like Is used for analyzing the node intensity distribution of the cause network.
4. The method of claim 1, wherein performing node community characteristic analysis on the cause network comprises:
the method for clustering the cause network comprises the following steps:
Wherein/> For node/>Cluster coefficient of/>、/>And/>Three adjacent nodes of every twoFor node/>Node strength of/>For node/>Degree of (v)/(v)To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Weights of edges being end points,/>To be node/>As a starting point, node/>Side being end point,/>To be node/>As a starting point, node/>Side being end point,/>To be node/>As a starting point, node/>Edges that are end points;
Performing a cause network for a cause network comprises:
Wherein/> Is the betweenness of the node,/>For connecting nodes/>, in a networkAnd/>All shortest paths through nodes/>Number of,/>For connecting nodes/>, in a networkAnd/>Number of all shortest paths,/>Representing the entire network.
5. The method for analyzing causes of accidents based on the theory of complex networks according to claim 1, wherein the analysis of the overall characteristics of the causes of accidents network comprises:
The method for analyzing the network diameter of the cause network comprises the following steps:
Wherein/> For the network diameter,/>To connect these two nodes in the network, the number of edges on the shortest path that is traversed,/>For nodes,/>Is a node;
the method for analyzing the network average path length of the cause network comprises the following steps:
Wherein/> For the average path length of the network,/>To connect these two nodes in the network, the number of edges on the shortest path that is traversed,/>Is the total number of nodes in the network.
6. The complex network theory-based accident cause analysis method according to claim 1, wherein performing association rule mining on the cause network comprises:
Taking two nodes connected in the cause network as project sets, and acquiring a result project set according to the connection times;
according to the occurrence times of the same item sets, obtaining a support degree, setting a support degree threshold, deleting the same item sets in the result item sets with the support degree smaller than the support degree threshold, and obtaining the latest result item set;
And carrying out association rule mining on the latest result item set to obtain association strength, namely supporting degree, confidence degree and lifting degree.
7. The accident cause analysis method based on the complex network theory according to claim 6, wherein the method for obtaining the support degree is as follows:
Wherein/> For connection/>To/>Support of/>For connection/>To/>Number of occurrences,/>Is a leading node,/>Is a successor node.
8. The accident cause analysis method based on the complex network theory according to claim 6, wherein the method for obtaining the confidence is:
Wherein/> For connection/>To/>Is used to determine the confidence level of the (c) in the (c),For connection/>To/>Frequency of occurrence and node/>Ratio of occurrence times of/(v)For all containing nodes/>Item set of (i.e., node/>)Is the number of occurrences of (a).
9. The accident cause analysis method based on the complex network theory according to claim 8, wherein the method for obtaining the lifting degree is as follows:
Wherein/> For connection/>To/>Degree of elevation of/>For all containing nodes/>Item set of (i.e., node/>)Is the number of occurrences of (a).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410585855.9A CN118171916A (en) | 2024-05-13 | 2024-05-13 | Accident cause analysis method based on complex network theory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410585855.9A CN118171916A (en) | 2024-05-13 | 2024-05-13 | Accident cause analysis method based on complex network theory |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118171916A true CN118171916A (en) | 2024-06-11 |
Family
ID=91350858
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410585855.9A Pending CN118171916A (en) | 2024-05-13 | 2024-05-13 | Accident cause analysis method based on complex network theory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118171916A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113506020A (en) * | 2021-07-26 | 2021-10-15 | 西安科技大学 | House building construction risk evaluation method based on complex network |
| CN117408059A (en) * | 2023-10-26 | 2024-01-16 | 武汉理工大学 | Accident element analysis method for ship collision accident |
-
2024
- 2024-05-13 CN CN202410585855.9A patent/CN118171916A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113506020A (en) * | 2021-07-26 | 2021-10-15 | 西安科技大学 | House building construction risk evaluation method based on complex network |
| CN117408059A (en) * | 2023-10-26 | 2024-01-16 | 武汉理工大学 | Accident element analysis method for ship collision accident |
Non-Patent Citations (2)
| Title |
|---|
| 周明悦等: "基于复杂网络模型的施工安全事故致因研究", 《工程管理学报》, vol. 38, no. 2, 30 April 2024 (2024-04-30), pages 90 - 94 * |
| 黄源等: "《大数据技术与应用》", 31 May 2020, 机械工业出版社, pages: 134 - 135 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107886235A (en) | A kind of Fire risk assessment method for coupling certainty and uncertainty analysis | |
| Shen et al. | Design of building construction safety prediction model based on optimized BP neural network algorithm | |
| CN110807580B (en) | Method for analyzing key safety risk of super high-rise construction machinery based on complex network | |
| CN114841660B (en) | Enterprise intelligent security management and control cloud platform based on field information | |
| CN116934262B (en) | Construction safety supervision system and method based on artificial intelligence | |
| Duan et al. | Cascading vulnerability analysis of unsafe behaviors of construction workers from the perspective of network modeling | |
| CN112508443A (en) | Power industry field operation risk management and control system | |
| Liao et al. | Identifying effective management factors across human errors–A case in elevator installation | |
| CN111507603A (en) | Method and system for evaluating current environmental risk situation of in-production enterprise | |
| Liao et al. | Fusion model for hazard association network development: A case in elevator installation and maintenance | |
| Guo et al. | Comparative analysis of the patterns of unsafe behaviors in accidents between building construction and urban railway construction | |
| CN111626640A (en) | Coal mine safety comprehensive risk evaluation method and system based on neutral reference object | |
| CN111626461A (en) | Safety risk prediction method | |
| Lu et al. | Using cased based reasoning for automated safety risk management in construction industry | |
| Chen et al. | Knowledge graph improved dynamic risk analysis method for behavior-based safety management on a construction site | |
| CN117408519A (en) | Chemical reaction process risk early warning method and device based on deep learning algorithm | |
| CN113506020A (en) | House building construction risk evaluation method based on complex network | |
| CN118171916A (en) | Accident cause analysis method based on complex network theory | |
| CN111723970A (en) | Power operation hidden danger prediction method | |
| Deng et al. | Exploring the metro construction accidents and causations for improving safety management based on data mining and network theory | |
| CN115456482A (en) | XGboost-based chemical enterprise risk classification method | |
| CN111832953A (en) | Emergency drilling script standardized design method for emergency | |
| Chen et al. | A hybrid STAMP-fuzzy DEMATEL-ISM approach for analyzing the factors influencing building collapse accidents in China | |
| CN104503441B (en) | Process fault monitoring method based on improved dynamic visible graph | |
| Liu et al. | Intelligent early warning of on-site safety of infrastructure projects based on dynamic risk factor identification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |