CN113420147A - Special equipment accident reason identification method and system - Google Patents
Special equipment accident reason identification method and system Download PDFInfo
- Publication number
- CN113420147A CN113420147A CN202110691047.7A CN202110691047A CN113420147A CN 113420147 A CN113420147 A CN 113420147A CN 202110691047 A CN202110691047 A CN 202110691047A CN 113420147 A CN113420147 A CN 113420147A
- Authority
- CN
- China
- Prior art keywords
- accident
- case
- reason
- similar
- current
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000000556 factor analysis Methods 0.000 claims abstract description 13
- 239000011159 matrix material Substances 0.000 claims description 59
- 238000012545 processing Methods 0.000 claims description 20
- 230000000295 complement effect Effects 0.000 claims description 19
- 238000012512 characterization method Methods 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000010586 diagram Methods 0.000 claims description 10
- 238000002368 isothermal capactiance transient spectroscopy Methods 0.000 claims description 8
- 230000004927 fusion Effects 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 3
- 238000009411 base construction Methods 0.000 claims description 2
- 230000008520 organization Effects 0.000 description 6
- 206010039203 Road traffic accident Diseases 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000011835 investigation Methods 0.000 description 3
- 235000008553 Allium fistulosum Nutrition 0.000 description 2
- 244000257727 Allium fistulosum Species 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 230000001364 causal effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000002650 habitual effect Effects 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000009659 non-destructive testing Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 231100000331 toxic Toxicity 0.000 description 1
- 230000002588 toxic effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method and a system for identifying accident reasons of special equipment, wherein the method is characterized in that an accident reason library is based on an initial importance degree corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process; analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment; selecting a historical case similar to the current case from the historical cases as a similar accident case according to each accident situation element in the accident situation element system; determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident reason in each similar accident case; and constructing an accident reason tree corresponding to the current case based on a plurality of final accident reasons corresponding to each layer in the current case, and identifying the accident reason of the current case by using the accident reason tree corresponding to the current case, thereby improving the accuracy of identifying the accident reason.
Description
Technical Field
The invention relates to the technical field of accident reason identification, in particular to a method and a system for identifying accident reasons of special equipment.
Background
By the end of 2019, the total amount of special equipment in China reaches 1525.47 ten thousands, which is increased by 9.4% compared with the end of 2018 and 17.08% compared with the end of 2017, and basically presents a steady rising trend. However, since the special equipment is operated at high temperature, high pressure or high speed and usually contains flammable, explosive, toxic media or a large number of people, the special equipment is very easy to cause group death and serious property damage once an accident occurs. Therefore, the accident prevention and control work of special equipment is particularly important. The rapid and accurate identification of the accident reason is the key to effectively prevent and control the accidents of special equipment. At present, the existing accident reason identification research is mainly focused on the field of traffic accidents, but the accidents of special equipment are different from the traffic accidents, and the accident causes, the accident mechanisms and the like of the special equipment and the traffic accidents are greatly different, so that the reason identification method of the traffic accidents cannot be directly used. Meanwhile, limited research on accident causes of special equipment mainly centers on two aspects: 1) the reasons of a large number of accidents are statistically analyzed, but the result universality is strong, and the specific reasons of specific accidents cannot be accurately identified; 2) for a certain accident, the accident investigation technology is used for analyzing reasons, but the accident investigation technology needs professional technologies such as nondestructive testing, the process is long, the requirement is high, and the accident reasons cannot be identified quickly. Further, the above studies are more directed to the posterior aspect, and the cause of an accident cannot be identified in other cases such as the prior art and the prior art. Obviously, it is necessary to develop a deep research for a fast and accurate method for identifying the accident cause of the special equipment.
The comprehensive and systematic cognition on the accident reason is the premise that the accident reason of the special equipment is rapidly and accurately identified. At present, accident cause models such as a Negis model, a 2-4 model and the like exist, but the side emphasis of each model is different, so that the accident cause representation of special equipment is not comprehensive, for example, the Negis model does not embody management factors, the 2-4 model does not consider trigger causes and the like. In addition, students also commonly use fault trees to analyze the cause and effect relationship among accident causes, but the accident cause trees established by different students are very different. In conclusion, the accident cause identification based on the method has the problem of low accuracy.
Disclosure of Invention
The invention aims to provide a method and a system for identifying accident reasons of special equipment so as to improve the accuracy of accident reason identification.
In order to achieve the above object, the present invention provides a method for identifying a cause of an accident of a special equipment, wherein the method comprises:
establishing an accident cause representation model of the special equipment by using the accident cause model and the fault tree;
based on the accident reason characterization model, extracting accident reasons of each historical case to form an accident reason library;
calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process based on the accident reason library; each historical case comprises a plurality of layers, and each layer comprises a plurality of accident causes;
analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment;
selecting a historical case similar to the current case from the historical cases as a similar accident case according to each accident situation element in the accident situation element system; each similar accident case comprises a plurality of layers, and each layer comprises a plurality of similar accident reasons;
determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident reason in each similar accident case;
and constructing an accident reason tree corresponding to the current case according to the quality room structure diagram and the accident reason tree of the similar accident case based on a plurality of final accident reasons corresponding to each layer in the current case, and identifying the accident reason of the current case by using the accident reason tree corresponding to the current case.
Optionally, the calculating, based on the accident cause database, an initial importance degree corresponding to each accident cause in each historical case by using a fuzzy analytic hierarchy process specifically includes:
based on the accident cause library, adopting a fuzzy analytic hierarchy process to establish fuzzy complementary judgment matrixes corresponding to the historical cases;
and calculating the initial importance corresponding to each accident reason in each history case based on the fuzzy complementary judgment matrix corresponding to each history case.
Optionally, the selecting, according to each accident scenario element in the accident scenario element system, a history case similar to the current case from the history cases as a similar accident case specifically includes:
performing word segmentation and stay word removal processing on the descriptive text of each accident situation element in the current case by using an ICTS and a stay word list to obtain a current effective feature word set; performing word segmentation and stay word removal processing on descriptive texts of accident scene elements in the historical case by using an ICTS and a stay word list to obtain a historical effective feature word set;
determining scene element similarity of the current case and each historical case corresponding to each accident scene element by using the current effective characteristic word set and the historical effective characteristic word set;
performing information fusion on the scene element similarity corresponding to each accident scene element under different scenes to obtain the scene similarity of the current case and each historical case under different scenes;
determining a scene similarity threshold SQ;
judging Sim*dWhether it is greater than or equal to SQ; if Sim*dWhen the SQ is greater than or equal to the SQ, the historical case Z is compareddAs a similar accident case of the current case, a historical case ZdTaking the accident reason of each layer as the similar accident reason of each layer of the similar accident case; wherein, Sim*dFor current case and historical case Z under different scenesdD ═ 1, 2.., D represents the total number of history cases.
Optionally, the determining, according to the initial importance degree corresponding to each similar accident reason in each similar accident case, a plurality of final accident reasons corresponding to each layer in the current case specifically includes:
calculating the similarity importance corresponding to each similar accident reason in each similar accident case based on the initial importance corresponding to each similar accident reason in each similar accident case and the scene similarity of the current case and each similar accident case;
combining and de-duplicating a plurality of similar accident reasons on each layer of each similar accident case to obtain a final similar accident reason set corresponding to each layer of the current case;
adding the initial importance degrees corresponding to a plurality of same similar accident reasons on each layer of each similar accident case to obtain a final importance degree vector set corresponding to each layer of the current case;
determining a final importance threshold CQh；
Judgment of
Whether or not it is greater than or equal to CQh(ii) a If it is not
CQ or greaterhIf the final similar accident reason corresponding to the tth final importance of the h layer of the current case is the final accident reason corresponding to the h layer of the current case,
represents the tth final importance of the h layer of the current case, T1, 2h，ThH is the total number of final importance of the h-th layer of the current case, 1, 2.
Optionally, the constructing an accident reason tree corresponding to the current case based on a plurality of final accident reasons corresponding to each layer in the current case according to the quality room structure diagram and the accident reason tree of each similar accident case, and performing accident reason identification on the current case by using the accident reason tree corresponding to the current case specifically includes:
establishing a correlation matrix of each final accident reason in two adjacent layers of the current case according to the quality room structure chart and the accident reason tree of each similar accident case;
establishing an autocorrelation correlation matrix of a plurality of final accident reasons at the h level in the current case, wherein h is 1, 2.
Establishing self or correlation matrixes of a plurality of final accident reasons of the h layer in the current case;
sequencing according to the correlation matrix, the autocorrelation matrix and the autocorrelation or the correlation degree to construct an accident reason tree corresponding to the current case;
and identifying the accident reason of the current case according to the accident reason tree corresponding to the current case.
The invention also provides a special equipment accident reason identification system, which comprises:
the accident cause representation model building module is used for building an accident cause representation model of the special equipment by utilizing the accident cause model and the fault tree;
the accident reason base construction module is used for extracting accident reasons of all historical cases based on the accident reason representation model to form an accident reason base;
the initial importance calculating module is used for calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process based on the accident reason library; each historical case comprises a plurality of layers, and each layer comprises a plurality of accident causes;
the accident scene element system construction module is used for analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment;
the similar accident case determination module is used for selecting a historical case similar to the current case from the historical cases as a similar accident case according to the accident scene elements in the accident scene element system; each similar accident case comprises a plurality of layers, and each layer comprises a plurality of similar accident reasons;
the final accident reason determining module is used for determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident reason in each similar accident case;
and the accident reason identification module is used for constructing an accident reason tree corresponding to the current case according to the quality room structure diagram and the accident reason tree of the similar accident case based on a plurality of final accident reasons corresponding to each layer in the current case, and identifying the accident reason of the current case by using the accident reason tree corresponding to the current case.
Optionally, the initial importance calculating module specifically includes:
the fuzzy complementary judgment matrix determining unit is used for establishing a fuzzy complementary judgment matrix corresponding to each historical case by adopting a fuzzy analytic hierarchy process based on the accident cause library;
and the initial importance calculating unit is used for calculating the initial importance corresponding to each accident reason in each history case based on the fuzzy complementary judgment matrix corresponding to each history case.
Optionally, the similar accident case determination module specifically includes:
the effective characteristic word set determining unit is used for performing word segmentation and stay word removal processing on the descriptive text of each accident situation element in the current case by utilizing the ICTCCLAS and the stay word list to obtain a current effective characteristic word set; performing word segmentation and stay word removal processing on descriptive texts of accident scene elements in the historical case by using an ICTS and a stay word list to obtain a historical effective feature word set;
the scene element similarity calculation unit is used for determining the scene element similarity corresponding to each accident scene element of the current case and each historical case by using the current effective characteristic word set and the historical effective characteristic word set;
the scene similarity calculation unit is used for performing information fusion on the scene element similarity corresponding to each accident scene element under different scenes to obtain the scene similarity of the current case and each historical case under different scenes;
the scene similarity threshold determining unit is used for determining a scene similarity threshold SQ;
a first judgment unit for judging Sim*dWhether it is greater than or equal to SQ; if Sim*dWhen the SQ is greater than or equal to the SQ, the historical case Z is compareddAs a similar accident case of the current case, a historical case ZdTaking the accident reason of each layer as the similar accident reason of each layer of the similar accident case; wherein, Sim*dFor current case and historical case Z under different scenesdD 1, 2.And D represent the total number of the historical cases.
Optionally, the final accident cause determining module specifically includes:
the similarity importance calculating unit is used for calculating the similarity importance corresponding to each similar accident reason in each similar accident case based on the initial importance corresponding to each similar accident reason in each similar accident case and the scene similarity of the current case and each similar accident case;
the merging and duplicate removal unit is used for merging and duplicate removal processing on a plurality of similar accident reasons on each layer of each similar accident case to obtain a final similar accident reason set corresponding to each layer of the current case;
the system comprises an adding processing unit, a calculating unit and a calculating unit, wherein the adding processing unit is used for adding initial importance degrees corresponding to a plurality of same similar accident reasons on each layer of each similar accident case to obtain a final importance degree vector set corresponding to each layer of the current case;
a final importance threshold determination unit for determining a final importance threshold CQh；
A second judgment unit for judging
Whether or not it is greater than or equal to CQh(ii) a If it is not
CQ or greaterhIf the final similar accident reason corresponding to the tth final importance of the h layer of the current case is the final accident reason corresponding to the h layer of the current case,
represents the tth final importance of the h layer of the current case, T1, 2h，ThH is the total number of final importance of the h-th layer of the current case, 1, 2.
Optionally, the accident cause identification module specifically includes:
the correlation matrix establishing unit is used for establishing a correlation matrix of each final accident reason in two adjacent layers of the current case according to the quality room structure chart and the accident reason tree of each similar accident case;
an autocorrelation matrix establishing unit, configured to establish an autocorrelation matrix of a plurality of final accident causes at a h-th level in a current case, where h is 1, 2.
The system comprises an auto or correlation matrix establishing unit, a correlation matrix calculating unit and a correlation matrix calculating unit, wherein the auto or correlation matrix establishing unit is used for establishing auto or correlation matrices of a plurality of final accident reasons of the h layer in the current case;
the accident reason tree establishing unit is used for sequencing according to the correlation matrix, the autocorrelation matrix and the autocorrelation or the correlation degree to establish an accident reason tree corresponding to the current case;
and the accident reason identification unit is used for identifying the accident reason of the current case according to the accident reason tree corresponding to the current case.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention discloses a method for identifying accident causes, which comprises the steps of firstly establishing an accident cause representation model by utilizing a fault tree and each accident cause theory to form an accident cause library of each layer, then analyzing each layer of accident causes under different scenes by utilizing a fuzzy analytic hierarchy process, scene similarity and the like, and finally analyzing the incidence relation among the accident causes of each layer by utilizing a quality room, thereby establishing the accident cause tree of the current case, realizing the identification of the hierarchical relation among the accident causes under different scenes and improving the accuracy of the identification of the accident causes.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a flow chart of an accident cause identification method for special equipment according to the invention;
FIG. 2 is a schematic diagram of an accident cause representation model of the special equipment according to the present invention;
FIG. 3 is a flow chart of the present invention for determining similar accident cases;
FIG. 4 is a flow chart of the present invention for determining the ultimate cause of an accident;
fig. 5 is a structural diagram of an accident cause identification system of special equipment.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a method and a system for identifying accident reasons of special equipment so as to improve the accuracy of accident reason identification.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
The method and the system combine the fault tree and each accident cause model to provide a cause characterization model of the special equipment accident, and the accident cause characterization model is favorable for comprehensively and normatively representing the cause of the special equipment accident. In addition, the accident reason can be quickly and accurately identified by adopting a scene analysis method and combining with the historical case.
As shown in fig. 1, the invention discloses a method for identifying a cause of an accident of a special device, which is characterized by comprising the following steps:
s1: and establishing an accident cause representation model of the special equipment by using the accident cause model and the fault tree.
S2: and extracting the accident reason of each historical case based on the accident reason characterization model to form an accident reason library.
S3: calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process based on the accident reason library; each historical case includes multiple layers, each layer including multiple accident causes.
S4: and analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment.
S5: selecting a historical case similar to the current case from the historical cases as a similar accident case according to each accident situation element in the accident situation element system; each similar accident case includes multiple layers, each layer including multiple similar accident causes.
S6: and determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance corresponding to each similar accident reason in each similar accident case.
S7: and constructing an accident reason tree corresponding to the current case according to the quality room structure diagram and the accident reason tree of the similar accident case based on a plurality of final accident reasons corresponding to each layer in the current case, and identifying the accident reason of the current case by using the accident reason tree corresponding to the current case.
The individual steps are discussed in detail below:
s1: establishing an accident cause representation model of the special equipment by using the accident cause model and the fault tree; the accident cause model comprises a Heinichis model and a 2-4 model.
Before the accident cause is identified, the type, cause and effect relationship and the like of the accident cause need to be relatively comprehensive and systematically known.
And establishing an accident cause characterization model of the special equipment as shown in figure 2. The accident cause characterization model mainly comprises seven layers of accident causes including two-level trigger causes, two-level direct causes, indirect causes, root causes and root causes, and the causal relationship among the causes of each layer is represented by AND/or relationship of a fault tree. The two-stage triggering reasons are respectively the accidental release of energy (including mechanical energy, electric energy and the like) of equipment and media thereof, surrounding objects and the like and the failure or destruction of control or shielding measures which play a role in restraining and limiting the energy. The two direct causes are mainly unsafe behaviors of people (operators, commanders and the like), unsafe states of objects (equipment, surrounding objects and the like) and unsafe conditions of environments (surrounding natural environments such as climate and equipment operation environments such as medium and the like), and the people/objects/environments are divided into two layers of harmful parties and causative parties according to causative relationships. Indirect reasons are habitual behaviors inside the organization, products and service quality of suppliers outside the organization, and the like due to poor safety and physiology. The root cause is that the safety management system inside the organization is deficient due to the lack of safety operation regulations, such as the social environment outside the organization, such as politics, economy, and the like, and human genetic genes. The root cause is the lack of security culture inside the organization, such as the security importance level, the supervision outside the organization, and the influence of other organizations.
S2: and extracting the accident reason of each historical case based on the accident reason characterization model to form an accident reason library.
Aiming at 490 special equipment accident cases (90 boilers, 116 pressure vessels, 26 pressure pipelines, 97 hoisting machines, 43 motor vehicles in the yard, 86 elevators, 4 passenger ropeways and 28 large-scale amusement facilities) in a 2005-2013 special equipment typical accident case set, analyzing the accident cause constitution and the cause and effect relationship of each historical case; secondly, determining the classification and the level of the accident reason of each historical case based on the accident reason characterization model in the figure 1, forming an accident reason table, and establishing an accident reason tree of each historical case; thirdly, according to the names and frequency ratios of various accident reasons, normalizing the accident reason names according to a simple majority principle, namely, a simple and clear reason name with a high frequency ratio is used as a priority item; and finally, updating the accident reason list and the accident reason tree of each historical case according to the normalized accident reason name, and establishing an accident reason library of the special equipment.
S3: and calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process based on the accident reason library, wherein each historical case comprises a plurality of layers, and each layer comprises a plurality of accident reasons.
S3: based on the accident reason library, calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process, which specifically comprises the following steps:
s31: and establishing a fuzzy complementary judgment matrix corresponding to each historical case by adopting a fuzzy analytic hierarchy process based on the accident cause library.
Suppose that
In order to establish a fuzzy complementary judgment matrix based on the accident cause library according to the complementary nine-scale judgment standard in table 1, the specific formula is as follows:
wherein R isdFor history case Zd(D ═ 1, 2.., D) corresponding fuzzy complementary decision matrices,
for history case Zd(D1, 2., D) h (h 1, 2., 7) th layer i (i 1, 2., N)dh) The cause of an accident and the j (j ═ 1, 2.., N)dh) The relative importance of the causes of the individual accidents is compared,
d is the total number of history cases, NdhThe total number of accident causes is included for the h-th layer.
TABLE 1 nine-Scale judgement ruler-chart
S32: calculating the initial importance corresponding to each accident reason in each layer of each history case based on the fuzzy complementary judgment matrix corresponding to each history case, which specifically comprises the following steps:
s321: carrying out consistency conversion on the fuzzy complementary judgment matrixes corresponding to the historical cases to obtain fuzzy consistency matrixes corresponding to the historical cases, wherein the specific formula is as follows:
wherein,
1≤i，j≤Ndh，
for history case ZdThe relative importance of the ith accident cause and the jth accident cause of the h layer is compared, BdRepresenting historical case ZdThe corresponding fuzzy consistency matrix is used to determine,
for history case Zd(D1, 2., D) th i (i 1, 2., N) th layerdh) The cause of an accident is relative to the j (j 1, 2.., N)dh) Fuzzy consistency values of individual accident causes.
S322: respectively carrying out normalization solution on each row in the fuzzy consistency matrix corresponding to each historical case to obtain the initial importance degree corresponding to each accident reason of each layer of each historical case, wherein the specific formula is as follows:
wherein,
representing historical case Zd(D1, 2., D) h (h 1, 2., 7) th layer i (i 1, 2., N)dh) The initial importance level corresponding to each accident cause,
for history case Zd(D ═ 1, 2., D) th i (i ═ 1, 2, …, N) th layerdh) The cause of an accident is relative to the j (j 1, 2.., N)dh) Fuzzy consistency values of individual accident causes.
S4: and analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment. The present invention refers to the root theory and factor analysis method as root-factor analysis method.
Before calculating the similarity of the accident scene, the type, the number, the dimension and the like of the accident scene elements need to be clarified. Therefore, the invention provides a specific process for analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form the accident scene element library of the special equipment, which comprises the following steps:
(1) determining the scope of the scene elements of each special equipment accident: and (3) carrying out concept extraction and category determination on the scene element descriptive texts of the historical cases by utilizing open decoding of a rooting theory aiming at 490 historical cases, thereby forming accident scene element categories of 28 special equipment accidents.
(2) Determining the main category of the accident scene elements according to the category of the scene elements, which comprises the following steps:
firstly, based on accident scene element categories of 28 special equipment accidents, the influence degree of each category on the special equipment accident scene is obtained by using a Likert scale, and then factor analysis is carried out by using a formula (6) to obtain a classification result of each accident scene element category.
Secondly, according to the classification result of each accident situation element category, a main category corresponding to each accident situation element category is summarized by utilizing the main shaft decoding of the root theory.
And finally, properly adjusting the relation between the main category and each accident scenario element category until the meaning set represented by the main category can completely cover the accident scenario element category, thereby obtaining the main category of 8 accident scenario elements.
In the formula, XlIs the main category of the first initial scene element, AoFor the o-th scene element category,
is a factor coefficient.
(3) Determining the core category of the accident scene element according to the main category of the accident scene element, and specifically comprising the following steps:
firstly, according to 8 main categories of accident scene elements, analyzing the influence degree of the main categories on the accident scene of the special equipment by using a Likert scale, and performing factor analysis by using a formula (6) to obtain the classification of each main category.
Secondly, the core categories corresponding to the main categories are summarized by utilizing the selective decoding of the root theory.
Finally, the relationship between the core category and the main category is properly adjusted until the meaning set represented by the core category can completely cover the main category, so that 3 accident scene element core categories are obtained. Finally, a special equipment accident situation element system shown in table 3 is formed, which is mainly divided into three levels of a core category, a main category and an accident situation element category. Meanwhile, the accident situation elements of each historical case are analyzed, and an accident situation element system table of the special equipment is further formed, as shown in table 2.
Table 2 Accident situation element system table
S5: and selecting a historical case similar to the current case from the historical cases as a similar accident case according to each accident situation element in the accident situation element system.
For the prior cause identification, the scene similarity calculation of the accident case is performed mainly based on the prior scene elements in table 2. The in-situ cause recognition is based on the two types of scene elements in advance and in advance in table 2, and the after-cause recognition is based on all the scene elements in table 2, that is, the three types of scene elements in advance, in advance and after.
As shown in fig. 3, the specific step S5 includes:
s51: utilizing ICTS CLAS and stay wordsTable for current case Z*Performing word segmentation and stay word removal processing on the descriptive text of each accident situation element to obtain a current effective characteristic word set T*q(ii) a Historical case Z by using ICTS CLAS and staying vocabularydPerforming word segmentation and word deswelling processing on the descriptive text of each accident situation element in (D1, 2.., D) to obtain a historical effective feature word set TdqWherein
for the E th word in the current valid feature word set*qThe number of the effective characteristic words is one,
for the F-th in the history valid characteristic word setdqAnd D represents the total number of the historical cases. T is*qRepresenting the current case Z*The corresponding valid feature word set, referred to as the current valid feature word set, TdqRepresenting historical case ZdAnd the corresponding effective characteristic word set is called history effective characteristic word set for short. In this embodiment, a Chinese lexical analysis System (ICTCLAS for short) is used.
S52: using the current valid feature word set T*qAnd a set of history valid feature words TdqDetermining a current case Z*And historical case ZdRegarding the similarity of the scene elements corresponding to the accident scene elements, the concrete formula is as follows:
wherein,
representing the current case Z*And historical case ZdElement A concerning accident situationqL2n () is the total length of the feature word, T*qRepresenting the current valid feature set, TdqThe method is characterized by representing a historical valid feature word set, wherein rho represents an adjustment coefficient and can be determined by reference to experience, and generally, rho is 0 4-0 6, q is 1, 2.
S53: carrying out information fusion on the scene element similarity corresponding to each accident scene element under different scenes to obtain the current case Z under different scenes*And historical case ZdScene similarity Sim*dThe concrete formula is as follows:
wherein,
representing the current case Z*And historical case ZdElement A concerning accident situationqThe similarity of scene elements.
S54: determining a threshold value of the scene similarity, wherein the specific formula is as follows:
SQ＝δ×max{Sim*d|d＝1，2，...，D}(9)；
where δ represents the current case Z*And historical case ZdPercentage of maximum similarity between, 0<Delta is less than or equal to 1, the delta value is determined according to historical data and experience, D represents the total number of historical cases, Sim*dRepresenting the current case Z in different scenarios*And historical case ZdSQ denotes a scene similarity threshold.
S55: judging Sim*dWhether the SQ is greater than or equal to the SQ; if Sim*dWhen the SQ is not less than the threshold, the historical case Z is useddAs current case Z*Similar accident case Z*gG is 1, 2, G is less than or equal to D. History case ZdAccident reasons of each layer as similar accident cases Z*gSimilar accidents at each levelFor the reason.
Suppose current case Z*There are G (G.ltoreq.D) similar accident cases, Z*gFor the current case Z*G (1, 2, G) similar accident case, then similar accident case Z*gI (1, 2.., N), N, of the h (1, 2.., 7) th layer of (i ═ 1, 2.., 7)gh) The initial importance corresponding to the similar accident cause is
Current case Z*And similar accident case Z*gHas a scene similarity of Sim*g。
And calculating the final importance of the accident reason by combining the initial importance corresponding to each similar accident reason of each layer of similar accident cases and the scene similarity of the similar accident cases, and determining each layer of accident reasons of the current case by combining the importance threshold. And then, analyzing the accident reason hierarchical relationship of the current case by combining the accident reason hierarchical relationship of the historical case, and establishing an accident reason tree of the current case, thereby realizing the identification of the accident reasons of each level of the current case. As shown in fig. 4, the specific process is as follows:
s6: determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance corresponding to each similar accident reason in each similar accident case, which specifically comprises the following steps:
s61: calculating the similarity importance degree corresponding to each similar accident reason in each similar accident case based on the initial importance degree corresponding to each similar accident reason in each similar accident case and the scene similarity of the current case and each similar accident case, wherein the specific formula is
Wherein,
representing similar accident cases Z*gSimilarity importance, Sim, corresponding to ith similar accident cause at the h-th level*gIs shown asFront case Z*And similar accident case Z*gThe degree of similarity of the scenes of (a),
representing similar accident cases Z*gThe initial importance corresponding to the ith similar accident reason of the h-th layer, G represents the current case Z*The total number of similar accident cases, h 1, 2, 7, i 1, 2, N, is sharedgh，NghIndicating that the h-th layer has the total number of similar accident causes.
For different similar accident cases, the similar accident causes may be repeated, for example, similar accident case Z*1And Z*2The 6 th layer has similar accident reasons of 'lack of emergency plan', therefore, the similar accident reasons of similar accident cases need to be sequentially matched, namely the current case Z*The initial similar accident reason is combined, de-duplicated and added, and the specific process is as follows:
firstly, a plurality of similar accident reasons of the h (h 1, 2, 7) th layer of all similar accident cases need to be summarized; then merging and de-duplicating the same accident reasons, for example, if the 6 th layer of the similar accident cases after summary has 4 'emergency plan defects', the similar accident cases can be merged into 1 'emergency plan defects'; finally, the initial importance of these same similar accident causes is summed, for example, the final importance of the 4 "emergency plan defects" is 0.1, 0.15 and 0.12 in turn, and then the final importance of the 1 "emergency plan defects" after combination is 0.1+0.1+0.15+0.2 to 0.55 by the summing. Thus obtaining the current case Z*The h-th layer of (1) is finally similar to the accident reason and the corresponding final importance.
S62: for similar accident case Z*gA plurality of similar accident reasons on the h-th layer are merged and deduplicated to obtain the current case Z*Final similar accident reason set corresponding to h-th layer
Wherein,
representing the current case Z*Th layer of (1)hThe reason for the similar accident is finally found,
s63: for similar accident case Z*gAdding the initial importance degrees corresponding to a plurality of same similar accident reasons on the h-th layer to obtain the current case Z*Final set of importance vectors corresponding to the h-th layer
Wherein,
representing the current case Z*Th layer of (1)hThe final importance corresponding to each final similar accident reason.
S64: determining a final importance threshold, wherein the specific formula is as follows:
wherein, CQhDenotes the final importance threshold, theta denotes the percentage of the maximum final importance, 0<θ≤1，
Representing the current case Z*The nth final importance of the h-th layer, T1, 2h。
S65: judgment of
Whether or not it is greater than or equal to CQh(ii) a If it is not
CQ or greaterhIf the final similar accident reason corresponding to the tth final importance of the h layer of the current case is the current case Z*The final accident cause corresponding to the h-th layer (T ═ 1, 2., T ·h)。
S7: based on current case Z*A plurality of final accident reasons corresponding to the middle layers are determined according to the quality room structure chart and the similar accident case Z*gCurrent case Z is constructed by accident reason tree*Corresponding accident reason tree, and using current case Z*Corresponding accident reason tree to current case Z*The accident reason identification is carried out, and the method specifically comprises the following steps:
s71: according to the quality room structure chart and the similar accident case Z*gThe current case Z is established*The correlation matrix of the final accident reasons in the h-th layer and the final accident reasons in the h-1 th layer is shown in table 3. Assume in Table 3 that the current case Z*The middle h layer has UhThe h-1 th layer has Uh-1And (4) the final accident reason.
S72: establishing a current case Z*The autocorrelation matrix of the plurality of final accident causes of the h-th layer is specifically shown in table 4. Assume current case Z in tables 4 and 5*The middle h layer has UhAnd (4) the final accident reason.
S73: establishing a current case Z*The h-th layer of the medium is a self or correlation matrix of a plurality of final accident causes, which is specifically shown in table 5.
Specifically, the final accident cause h in table 3 is calculated using the formula (12)uAnd (h-1)vThe degree of correlation between; the final accident cause h in Table 4 is calculated using the formula (13)uAnd hoThe final accident cause h in Table 5 is calculated by the formula (14)uAnd hoThe degree of auto-or correlation between.
Wherein, Sim*gRepresenting the current case Z*And similar accident case Z*gIf in the similar accident case Z*gFinal cause of accident huAnd (h-1)vCorrelation, then ηg1, otherwise ηg0. In case of similar accident Z*gFinal cause of accident huAnd hoPhase and, then xig1, otherwise ξg0. In case of similar accident Z*gFinal cause of accident huAnd hoPhase or, then lambdag1, otherwise λg＝0。
S74: sequencing according to the correlation matrix, the autocorrelation matrix and the autocorrelation degree to construct a current case Z*Corresponding accident reason tree.
S75: according to the current case Z*Corresponding accident reason tree identification current case Z*The cause of the accident.
When the cause of the accident is more, the correlation threshold can be set with reference to equations (9) and (11) to further screen the current case Z*The key accident cause and the correlation thereof.
TABLE 3 Current case Z*Correlation matrix of final accident reasons of h-th layer and h-1 layer
Table 4 current case Z*The h-th layer of (1) and the final accident cause of the accidentCorrelation matrix
TABLE 5 Current case Z*H layer of (3) the autocorrelation or correlation matrix of the ultimate cause of the accident
Compared with the prior art:
1. the accident reason identification under different scenes can be realized quickly and accurately
The existing reason analysis of the special equipment is more specific to the situation after the incident, and the proposed method can quickly reuse the accident reason and the correlation of the corresponding historical case through the situation similarity calculation under the three situations before, in the incident and after the incident, so as to respectively realize the identification of the accident reason of the current case under the three different situations. The method can specifically analyze specific accidents without accident investigation technology, has low requirement on users and is convenient to popularize and apply.
2. Can identify the hierarchical relationship between accident causes
The existing special equipment accident reason research is more oriented to analysis of the same-layer accident reason, and hierarchical accident reason identification similar to a fault tree cannot be formed, particularly in the situations of advance and in the middle of the day. The method comprises the steps of firstly establishing an accident reason representation model by utilizing a fault tree and each accident cause theory to form an accident reason library of each layer, then analyzing each layer of accident reasons under different scenes by utilizing a fuzzy analytic hierarchy process, scene similarity and the like, and finally analyzing the incidence relation among the accident reasons of each layer by utilizing a quality room, thereby establishing the accident reason tree of the current case and realizing the identification of the hierarchical relation among the accident reasons under different scenes.
As shown in fig. 5, the invention discloses a special equipment accident cause identification system, which comprises:
the accident cause characterization model building module 501 is configured to build an accident cause characterization model of the special equipment by using the accident cause model and the fault tree.
And an accident reason library constructing module 502, configured to extract accident reasons of each historical case based on the accident reason characterization model to form an accident reason library.
An initial importance calculation module 503, configured to calculate an initial importance corresponding to each accident reason in each historical case by using a fuzzy analytic hierarchy process based on the accident reason library; each historical case includes multiple layers, each layer including multiple accident causes.
The accident scenario element system construction module 504 is configured to analyze the accident scenario elements of each historical case by using a root-factor analysis method to form an accident scenario element system of the special device.
A similar accident case determining module 505, configured to select, according to each accident situation element in the accident situation element system, a history case similar to the current case from each history case as a similar accident case; each similar accident case includes multiple layers, each layer including multiple similar accident causes.
A final accident cause determining module 506, configured to determine a plurality of final accident causes corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident cause in each similar accident case.
The accident reason identification module 507 is configured to construct an accident reason tree corresponding to the current case according to the quality room structure diagram and the accident reason tree of the similar accident case based on a plurality of final accident reasons corresponding to each layer in the current case, and perform accident reason identification on the current case by using the accident reason tree corresponding to the current case.
As an optional implementation manner, the initial importance calculating module 503 of the present invention specifically includes:
and the fuzzy complementary judgment matrix determining unit is used for establishing a fuzzy complementary judgment matrix corresponding to each historical case by adopting a fuzzy analytic hierarchy process based on the accident cause library.
And the initial importance calculating unit is used for calculating the initial importance corresponding to each accident reason in each history case based on the fuzzy complementary judgment matrix corresponding to each history case.
As an optional implementation manner, the similar accident case determination module 505 specifically includes:
the effective characteristic word set determining unit is used for performing word segmentation and stay word removal processing on the descriptive text of each accident situation element in the current case by utilizing the ICTCCLAS and the stay word list to obtain a current effective characteristic word set; and performing word segmentation and stay word removal processing on the descriptive text of each accident situation element in the historical case by using the ICTCCLAS and the stay word list to obtain a historical effective characteristic word set.
And the scene element similarity calculation unit is used for determining the scene element similarity corresponding to each accident scene element of the current case and each historical case by using the current effective characteristic word set and the historical effective characteristic word set.
And the scene similarity calculation unit is used for performing information fusion on the scene element similarity corresponding to each accident scene element under different scenes to obtain the scene similarity of the current case and each historical case under different scenes.
And the scene similarity threshold determining unit is used for determining a scene similarity threshold SQ.
A first judgment unit for judging Sim*dWhether it is greater than or equal to SQ; if Sim*dWhen the SQ is greater than or equal to the SQ, the historical case Z is compareddAs a similar accident case of the current case, a historical case ZdTaking the accident reason of each layer as the similar accident reason of each layer of the similar accident case; wherein, Sim*dFor current case and historical case Z under different scenesdD ═ 1, 2.., D represents the total number of history cases.
As an optional implementation manner, the final accident cause determining module 506 of the present invention specifically includes:
and the similarity importance calculating unit is used for calculating the similarity importance corresponding to each similar accident reason in each similar accident case based on the initial importance corresponding to each similar accident reason in each similar accident case and the scene similarity of the current case and each similar accident case.
And the merging and duplicate removal unit is used for merging and duplicate removal processing on a plurality of similar accident reasons on each layer of each similar accident case to obtain a final similar accident reason set corresponding to each layer of the current case.
And the addition processing unit is used for adding the initial importance degrees corresponding to the same similar accident reasons on each layer of each similar accident case to obtain a final importance degree vector set corresponding to each layer of the current case.
A final importance threshold determination unit for determining a final importance threshold CQh。
A second judgment unit for judging
Whether or not it is greater than or equal to CQh(ii) a If it is not
CQ or greaterhIf the final similar accident reason corresponding to the tth final importance of the h layer of the current case is the final accident reason corresponding to the h layer of the current case,
represents the tth final importance of the h layer of the current case, T1, 2h，ThH is the total number of final importance of the h-th layer of the current case, 1, 2.
As an optional implementation manner, the accident cause identification module 507 of the present invention specifically includes:
and the correlation matrix establishing unit is used for establishing a correlation matrix of each final accident reason in two adjacent layers of the current case according to the quality room structure chart and the accident reason tree of each similar accident case.
And the autocorrelation matrix establishing unit is used for establishing an autocorrelation matrix of a plurality of final accident reasons at the h level in the current case, wherein h is 1, 2.
And the self or correlation matrix establishing unit is used for establishing self or correlation matrixes of a plurality of final accident reasons of the h layer in the current case.
And the accident reason tree establishing unit is used for sequencing according to the correlation matrix, the autocorrelation matrix and the autocorrelation or the correlation degree to establish an accident reason tree corresponding to the current case.
And the accident reason identification unit is used for identifying the accident reason of the current case according to the accident reason tree corresponding to the current case.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. A method for identifying accident causes of special equipment is characterized by comprising the following steps:
establishing an accident cause representation model of the special equipment by using the accident cause model and the fault tree;
based on the accident reason characterization model, extracting accident reasons of each historical case to form an accident reason library;
calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process based on the accident reason library; each historical case comprises a plurality of layers, and each layer comprises a plurality of accident causes;
analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment;
selecting a historical case similar to the current case from the historical cases as a similar accident case according to each accident situation element in the accident situation element system; each similar accident case comprises a plurality of layers, and each layer comprises a plurality of similar accident reasons;
determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident reason in each similar accident case;
and constructing an accident reason tree corresponding to the current case according to the quality room structure diagram and the accident reason tree of the similar accident case based on a plurality of final accident reasons corresponding to each layer in the current case, and identifying the accident reason of the current case by using the accident reason tree corresponding to the current case.
2. The special equipment accident cause identification method according to claim 1, wherein the calculating of the initial importance degree corresponding to each accident cause in each historical case by using a fuzzy analytic hierarchy process based on the accident cause library specifically comprises:
based on the accident cause library, adopting a fuzzy analytic hierarchy process to establish fuzzy complementary judgment matrixes corresponding to the historical cases;
and calculating the initial importance corresponding to each accident reason in each history case based on the fuzzy complementary judgment matrix corresponding to each history case.
3. The special equipment accident cause identification method according to claim 1, wherein the selecting, from history cases, a history case similar to a current case as a similar accident case according to the accident situation elements in the accident situation element system specifically comprises:
performing word segmentation and stay word removal processing on the descriptive text of each accident situation element in the current case by using an ICTS and a stay word list to obtain a current effective feature word set; performing word segmentation and stay word removal processing on descriptive texts of accident scene elements in the historical case by using an ICTS and a stay word list to obtain a historical effective feature word set;
determining scene element similarity of the current case and each historical case corresponding to each accident scene element by using the current effective characteristic word set and the historical effective characteristic word set;
performing information fusion on the scene element similarity corresponding to each accident scene element under different scenes to obtain the scene similarity of the current case and each historical case under different scenes;
determining a scene similarity threshold SQ;
judging Sim*dWhether it is greater than or equal to SQ; if Sim*dWhen the SQ is greater than or equal to the SQ, the historical case Z is compareddAs a similar accident case of the current case, a historical case ZdTaking the accident reason of each layer as the similar accident reason of each layer of the similar accident case; wherein, Sim*dFor current case and historical case Z under different scenesdD ═ 1, 2.., D represents the total number of history cases.
4. The special equipment accident cause identification method according to claim 1, wherein the determining a plurality of final accident causes corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident cause in each similar accident case specifically comprises:
calculating the similarity importance corresponding to each similar accident reason in each similar accident case based on the initial importance corresponding to each similar accident reason in each similar accident case and the scene similarity of the current case and each similar accident case;
combining and de-duplicating a plurality of similar accident reasons on each layer of each similar accident case to obtain a final similar accident reason set corresponding to each layer of the current case;
adding the initial importance degrees corresponding to a plurality of same similar accident reasons on each layer of each similar accident case to obtain a final importance degree vector set corresponding to each layer of the current case;
determining a final importance threshold CQh；
Judgment of
Whether or not it is greater than or equal to CQh(ii) a If it is not
CQ or greaterhThen the final similar accident reason corresponding to the tth final importance of the h layer of the current case is used as the final accident reason corresponding to the h layer of the current case,
k denotes the tth final importance of the h layer of the current case, T1, 2h，ThH is the total number of final importance of the h-th layer of the current case, 1, 2.
5. The method for identifying the accident reason of the special equipment according to claim 1, wherein the method for identifying the accident reason of the current case based on a plurality of final accident reasons corresponding to each layer in the current case, the accident reason tree corresponding to the current case is constructed according to the quality room structure diagram and the accident reason tree of each similar accident case, and the accident reason tree corresponding to the current case is used for identifying the accident reason of the current case, which specifically comprises the following steps:
establishing a correlation matrix of each final accident reason in two adjacent layers of the current case according to the quality room structure chart and the accident reason tree of each similar accident case;
establishing an autocorrelation correlation matrix of a plurality of final accident reasons at the h level in the current case, wherein h is 1, 2.
Establishing self or correlation matrixes of a plurality of final accident reasons of the h layer in the current case;
sequencing according to the correlation matrix, the autocorrelation matrix and the autocorrelation or the correlation degree to construct an accident reason tree corresponding to the current case;
and identifying the accident reason of the current case according to the accident reason tree corresponding to the current case.
6. A special equipment accident cause identification system, the system comprising:
the accident cause representation model building module is used for building an accident cause representation model of the special equipment by utilizing the accident cause model and the fault tree;
the accident reason base construction module is used for extracting accident reasons of all historical cases based on the accident reason representation model to form an accident reason base;
the initial importance calculating module is used for calculating the initial importance corresponding to each accident reason in each historical case by adopting a fuzzy analytic hierarchy process based on the accident reason library; each historical case comprises a plurality of layers, and each layer comprises a plurality of accident causes;
the accident scene element system construction module is used for analyzing the accident scene elements of each historical case by utilizing a rooting-factor analysis method to form an accident scene element system of the special equipment;
the similar accident case determination module is used for selecting a historical case similar to the current case from the historical cases as a similar accident case according to the accident scene elements in the accident scene element system; each similar accident case comprises a plurality of layers, and each layer comprises a plurality of similar accident reasons;
the final accident reason determining module is used for determining a plurality of final accident reasons corresponding to each layer in the current case according to the initial importance degree corresponding to each similar accident reason in each similar accident case;
and the accident reason identification module is used for constructing an accident reason tree corresponding to the current case according to the quality room structure diagram and the accident reason tree of the similar accident case based on a plurality of final accident reasons corresponding to each layer in the current case, and identifying the accident reason of the current case by using the accident reason tree corresponding to the current case.
7. The special equipment accident cause identification system according to claim 6, wherein the initial importance calculation module specifically comprises:
the fuzzy complementary judgment matrix determining unit is used for establishing a fuzzy complementary judgment matrix corresponding to each historical case by adopting a fuzzy analytic hierarchy process based on the accident cause library;
and the initial importance calculating unit is used for calculating the initial importance corresponding to each accident reason in each history case based on the fuzzy complementary judgment matrix corresponding to each history case.
8. The special equipment accident cause identification system according to claim 6, wherein the similar accident case determination module specifically comprises:
the effective characteristic word set determining unit is used for performing word segmentation and stay word removal processing on the descriptive text of each accident situation element in the current case by utilizing the ICTCCLAS and the stay word list to obtain a current effective characteristic word set; performing word segmentation and stay word removal processing on descriptive texts of accident scene elements in the historical case by using an ICTS and a stay word list to obtain a historical effective feature word set;
the scene element similarity calculation unit is used for determining the scene element similarity corresponding to each accident scene element of the current case and each historical case by using the current effective characteristic word set and the historical effective characteristic word set;
the scene similarity calculation unit is used for performing information fusion on the scene element similarity corresponding to each accident scene element under different scenes to obtain the scene similarity of the current case and each historical case under different scenes;
the scene similarity threshold determining unit is used for determining a scene similarity threshold SQ;
a first judgment unit for judging Sim*dWhether the SQ is greater than or equal to the SQ; if Sim*dWhen the SQ is greater than or equal to the SQ, the historical case Z is compareddAs a similar accident case of the current case, a historical case ZdTaking the accident reason of each layer as the similar accident reason of each layer of the similar accident case; wherein, Sim*dFor current case and historical case Z under different scenesdD ═ 1, 2.., D represents the total number of history cases.
9. The special equipment accident cause identification system according to claim 6, wherein the final accident cause determination module specifically comprises:
the similarity importance calculating unit is used for calculating the similarity importance corresponding to each similar accident reason in each similar accident case based on the initial importance corresponding to each similar accident reason in each similar accident case and the scene similarity of the current case and each similar accident case;
the merging and duplicate removal unit is used for merging and duplicate removal processing on a plurality of similar accident reasons on each layer of each similar accident case to obtain a final similar accident reason set corresponding to each layer of the current case;
the system comprises an adding processing unit, a calculating unit and a calculating unit, wherein the adding processing unit is used for adding initial importance degrees corresponding to a plurality of same similar accident reasons on each layer of each similar accident case to obtain a final importance degree vector set corresponding to each layer of the current case;
a final importance threshold determination unit for determining a final importance threshold CQh；
A second judgment unit for judging
Whether or not it is greater than or equal to CQh(ii) a If it is not
CQ or greaterhIf the final similar accident reason corresponding to the tth final importance of the h layer of the current case is the final accident reason corresponding to the h layer of the current case,
represents the tth final importance of the h layer of the current case, T1, 2h，ThH is the total number of final importance of the h-th layer of the current case, 1, 2.
10. The special equipment accident cause identification system according to claim 6, wherein the accident cause identification module specifically comprises:
the correlation matrix establishing unit is used for establishing a correlation matrix of each final accident reason in two adjacent layers of the current case according to the quality room structure chart and the accident reason tree of each similar accident case;
an autocorrelation matrix establishing unit, configured to establish an autocorrelation matrix of a plurality of final accident causes at a h-th level in a current case, where h is 1, 2.
The system comprises an auto or correlation matrix establishing unit, a correlation matrix calculating unit and a correlation matrix calculating unit, wherein the auto or correlation matrix establishing unit is used for establishing auto or correlation matrices of a plurality of final accident reasons of the h layer in the current case;
the accident reason tree establishing unit is used for sequencing according to the correlation matrix, the autocorrelation matrix and the autocorrelation or the correlation degree to establish an accident reason tree corresponding to the current case;
and the accident reason identification unit is used for identifying the accident reason of the current case according to the accident reason tree corresponding to the current case.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110691047.7A CN113420147B (en) | 2021-06-22 | 2021-06-22 | Special equipment accident cause identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110691047.7A CN113420147B (en) | 2021-06-22 | 2021-06-22 | Special equipment accident cause identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113420147A true CN113420147A (en) | 2021-09-21 |
| CN113420147B CN113420147B (en) | 2024-01-26 |
Family
ID=77789841
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110691047.7A Active CN113420147B (en) | 2021-06-22 | 2021-06-22 | Special equipment accident cause identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113420147B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020107721A1 (en) * | 2000-10-24 | 2002-08-08 | International Business Machines Corporation | Story-based organizational assessment and effect system |
| US20080133288A1 (en) * | 2006-11-30 | 2008-06-05 | Microsoft Corporation | Grouping Failures to Infer Common Causes |
| CN103500283A (en) * | 2013-10-11 | 2014-01-08 | 国家电网公司 | Power transformer risk assessment method based on fault tree |
| CN106920040A (en) * | 2017-03-01 | 2017-07-04 | 西南交通大学 | Freeway tunnel street accidents risks appraisal procedure based on Fuzzy Level Analytic Approach |
| CN110580580A (en) * | 2019-09-02 | 2019-12-17 | 长沙理工大学 | Bridge hanging basket construction risk assessment method based on fuzzy analytic hierarchy process |
| CN112508416A (en) * | 2020-12-11 | 2021-03-16 | 华南理工大学 | Oil and gas storage and transportation station safety grade evaluation method based on cloud fuzzy analytic hierarchy process |
-
2021
- 2021-06-22 CN CN202110691047.7A patent/CN113420147B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020107721A1 (en) * | 2000-10-24 | 2002-08-08 | International Business Machines Corporation | Story-based organizational assessment and effect system |
| US20080133288A1 (en) * | 2006-11-30 | 2008-06-05 | Microsoft Corporation | Grouping Failures to Infer Common Causes |
| CN103500283A (en) * | 2013-10-11 | 2014-01-08 | 国家电网公司 | Power transformer risk assessment method based on fault tree |
| CN106920040A (en) * | 2017-03-01 | 2017-07-04 | 西南交通大学 | Freeway tunnel street accidents risks appraisal procedure based on Fuzzy Level Analytic Approach |
| CN110580580A (en) * | 2019-09-02 | 2019-12-17 | 长沙理工大学 | Bridge hanging basket construction risk assessment method based on fuzzy analytic hierarchy process |
| CN112508416A (en) * | 2020-12-11 | 2021-03-16 | 华南理工大学 | Oil and gas storage and transportation station safety grade evaluation method based on cloud fuzzy analytic hierarchy process |
Non-Patent Citations (2)
| Title |
|---|
| 谭章禄;宋庆正;: "基于扎根理论的煤矿安全事故致因分析", 煤矿安全, no. 09 * |
| 邹永广;林炜铃;: ""驴友"安全事故影响因素重要度评价研究", 旅游导刊, no. 04 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113420147B (en) | 2024-01-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200208786A1 (en) | Method for tracking and locating contamination sources in water distribution systems based on consumer complaints | |
| DiMattia et al. | Determination of human error probabilities for offshore platform musters | |
| CN111475655B (en) | Power distribution network knowledge graph-based power scheduling text entity linking method | |
| CN108537259A (en) | Train control on board equipment failure modes and recognition methods based on Rough Sets Neural Networks model | |
| CN111709244A (en) | Deep learning method for identifying causal relationship of contradictory dispute events | |
| CN112199496A (en) | Power grid equipment defect text classification method based on multi-head attention mechanism and RCNN (Rich coupled neural network) | |
| Karimi et al. | A deep model for predicting online course performance | |
| CN112488371A (en) | Pipeline intelligent early warning method and system based on big data | |
| CN115909692A (en) | Management method, platform, equipment and medium for expressway alarm event | |
| CN114254102B (en) | Natural language-based collaborative emergency response SOAR script recommendation method | |
| CN114757557A (en) | On-site operation risk assessment prediction method and device based on electric work ticket | |
| CN113743750A (en) | Nuclear power plant process system process risk assessment system and method | |
| Katz et al. | VISIT: Visualizing and interpreting the semantic information flow of transformers | |
| Lei et al. | A finer-grain universal dialogue semantic structures based model for abstractive dialogue summarization | |
| CN117667495B (en) | Association rule and deep learning integrated application system fault prediction method | |
| CN113420147A (en) | Special equipment accident reason identification method and system | |
| CN117313742A (en) | Intelligent fault diagnosis method for aviation maintenance text records | |
| Tran et al. | BARTPhoBEiT: Pre-trained Sequence-to-Sequence and Image Transformers Models for Vietnamese Visual Question Answering | |
| Čepin | Comparison of methods for dependency determination between human failure events within human reliability analysis | |
| Wang | Subway crowded and stampede neural networks safety assessment basing on shel model | |
| CN115618297A (en) | Method and device for identifying abnormal enterprise | |
| CN112990599A (en) | Urban emergency fire fighting optimization method based on similarity calculation of emergency fire events | |
| CN117194633B (en) | Dam emergency response knowledge question-answering system based on multi-level multipath and implementation method | |
| Lopez et al. | Paying Attention to the Insider Threat. | |
| Xu et al. | Real-time Classification of Ship Cabins Fire Hazard Levels Based on CNN |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |