CN115034648B - Bridge engineering risk assessment method based on BP neural network under condition of few samples - Google Patents
Bridge engineering risk assessment method based on BP neural network under condition of few samples Download PDFInfo
- Publication number
- CN115034648B CN115034648B CN202210727541.9A CN202210727541A CN115034648B CN 115034648 B CN115034648 B CN 115034648B CN 202210727541 A CN202210727541 A CN 202210727541A CN 115034648 B CN115034648 B CN 115034648B
- Authority
- CN
- China
- Prior art keywords
- risk
- level
- bridge engineering
- construction
- probability
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000012502 risk assessment Methods 0.000 title claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 29
- 238000010276 construction Methods 0.000 claims abstract description 65
- 238000012549 training Methods 0.000 claims abstract description 24
- 238000011156 evaluation Methods 0.000 claims abstract description 13
- 238000011002 quantification Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 59
- 239000011159 matrix material Substances 0.000 claims description 44
- 238000013139 quantization Methods 0.000 claims description 32
- 238000009430 construction management Methods 0.000 claims description 21
- 239000004035 construction material Substances 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 16
- 239000000463 material Substances 0.000 claims description 12
- 238000003062 neural network model Methods 0.000 claims description 12
- 238000013459 approach Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 3
- 238000011835 investigation Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000013210 evaluation model Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 7
- 238000011161 development Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/08—Construction
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Primary Health Care (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a bridge engineering risk assessment method based on BP neural network under the condition of few samples, which comprises the steps of firstly, identifying risk factors for bridge construction safety and establishing a risk assessment system; secondly, preprocessing sample data obtained by an expert experience assessment method by adopting an analytic hierarchy process, and training the processed sample by combining a BP neural network to construct a hybrid assessment model; and finally, applying the hybrid evaluation model to bridge construction safety risk evaluation. According to the invention, an AHP is utilized to construct an evaluation index system meeting engineering construction, and meanwhile, the nonlinear characteristic of the BP neural network is adopted to solve the problem that the risk level quantification is artificial subjectivity too large by an expert evaluation method. The method is more concise, practical, scientific and strict, is convenient for compiling a standardized algorithm program, and has certain reference value and popularization value. The invention can obtain the evaluation result scientifically and accurately without a large number of data samples.
Description
Technical Field
The invention belongs to the field of bridge engineering, relates to a bridge engineering risk assessment method, and in particular relates to a bridge engineering risk assessment method based on a BP neural network under a few sample condition.
Background
In recent years, road traffic in China is in a rapid development stage, and the construction requirement for western high-speed railway bridges is urgent. Most western parts of China are in mountain river zones, so a large number of high-speed railway bridges are required to be built. The engineering safety risk influence factors of the high-speed railway bridge are complex, and in the development of the high-speed railway, the engineering construction of the high-speed railway bridge is focused. In order to ensure the rapid development and construction of the high-speed railway bridge, a set of complete and scientific safety risk assessment system aiming at the construction of the high-speed railway bridge is needed.
The national safety risk assessment research on bridge engineering construction mainly adopts a traditional expert experience assessment method, the main research direction is concentrated on the aspects of roads, tunnels and the like, and the assessment on each risk factor is too subjective, so that the safety risk assessment result is influenced.
In recent years, quantitative security risk assessment is started to be carried out on the high-speed railway bridge risk assessment, and although the scientificity of a result is met, the calculation process is complex and not concise enough. Aiming at complex and tedious links, heavier man-made subjectivity, complex calculation and low informatization degree of an engineering safety evaluation system, the evaluation result is deviated and the timeliness is poor.
In the prior art, many problems can be solved by adopting neural network artificial intelligence learning, but the artificial intelligence learning often needs enough training samples, and because of bridge engineering specificity, the construction period is long, and the enough training samples are often difficult to collect, so that the bridge engineering risk assessment by using the neural network is greatly limited.
In summary, at present, a very complex nonlinear relationship exists among all influencing factors in the risk assessment of the construction safety of the high-speed railway bridge, and the adoption of the BP neural network can truly reduce artificial interference factors and has strong nonlinear approximation capability, so that the method has objectivity, but the premise of using the BP neural network is that enough data samples are needed to train a network model. At present, samples are generally obtained by an expert according to experience by adopting an expert evaluation method, and sample data are easy to be subject to artificial subjective interference; the Analytic Hierarchy Process (AHP) integrally judges the matrix through the relative relation components of the importance of the risk factors, so that the problem that the sample is subjected to artificial subjective interference is solved to a certain extent, and the index system of the engineering process flow components is easily combined.
Disclosure of Invention
The invention aims to provide a bridge engineering risk assessment method based on a BP neural network under the condition of few samples, which is characterized in that the learning accuracy is high enough through limited samples, so that quantification and objectification requirements of high-speed railway bridge construction safety risk assessment are met. Firstly, constructing a high-speed railway bridge construction safety risk assessment system, identifying bridge construction risk indexes, and quantitatively scoring the occurrence probability (P) of risk events and the occurrence consequence degree S of the risk events of all risk factors; preprocessing the data by an AHP method to obtain a final risk level; and finally, building a BP neural network to train data and obtaining a complete scientific and accurate model. The method integrates an AHP method and a BP neural network algorithm, and constructs a high-speed railway bridge construction safety risk assessment system so as to meet the requirements of scientificity, conciseness and generalizability of high-speed railway bridge construction risk assessment.
In order to solve the technical problems, the invention adopts the following technical scheme:
The bridge engineering risk assessment method based on the BP neural network under the condition of few samples is characterized by comprising the following steps of:
S1, constructing a bridge engineering construction safety risk assessment system, and identifying risk factors of bridge construction to obtain a plurality of first-level risk factor indexes, wherein each first-level risk factor index comprises a plurality of second-level risk factor indexes;
S2, aiming at the identified risk factor secondary indexes, carrying out preliminary risk level quantification on each corresponding risk investigation item to obtain a safety risk assessment data set K consisting of quantified values of all the secondary indexes;
S3, grading after quantitatively scoring the risk event occurrence probability P and the risk event occurrence result degree S of the risk factors, and grading the risk grade of the bridge engineering according to the risk event occurrence probability P and the risk event occurrence result degree S;
S4, carrying out data preprocessing on the quantized scores of the occurrence probability P of the risk event and the occurrence consequence degree S of the risk event by using a analytic hierarchy process to obtain a first weight vector W and a second weight vector W *, wherein the first weight vector W consists of the occurrence probability weights of the risk event and the occurrence consequence degree weights of the risk event of each secondary index;
S5, multiplying each secondary index quantized value by a corresponding risk event occurrence probability weight W and accumulating to obtain a risk event occurrence probability P of the risk factors, and multiplying each secondary index quantized value by a corresponding risk event occurrence result degree weight W * and accumulating to obtain a risk event occurrence result degree S of the risk factors;
S6, taking the safety risk assessment data set K for identifying the bridge engineering in the step S1 as a sample, taking the risk event occurrence probability P and the risk event occurrence result degree S obtained in the step S5 as labels, constructing a training sample, collecting safety risk assessment data of different bridge engineering according to the methods of the steps S1-S5, and constructing a training sample set;
s7, establishing a BP neural network model;
S8, inputting the sample set for bridge engineering risk assessment into a BP neural network model to perform self-learning training of the network;
S9, carrying out risk factor identification on the bridge engineering to be evaluated according to the methods of the steps S1 and S2 to obtain a safety risk evaluation data set K of the bridge engineering to be evaluated; and (3) inputting the identified safety risk assessment data set K into the BP neural network model after training is completed, obtaining the occurrence probability P of the risk event and the occurrence result degree S of the risk event, and obtaining the safety risk level of the bridge engineering through the bridge engineering risk level standard divided in the step (S3).
Further, in step S1, the four primary indexes of the identified risk factors are personnel quality K 1, mechanical equipment and construction materials K 2, construction management K 3 and construction environment K 4 respectively; wherein,
Personnel quality K 1 includes: safety awareness K 11 of personnel, operation standardization degree K 12, personnel academy K 13, personnel experience K 14, pre-post safety training K 15, working time K 16, site protection measure K 17, safety equipment use experience K 18 and salon spirit K 19;
Mechanical equipment and construction material K 2 include: the equipment approach quality detection K 21, the safety operation standard K 22, whether the mechanical strength is suitable for the construction strength K 23, the equipment maintenance K 24, the material approach detection K 25, the material arrangement and storage layout K 26, the in-field material transportation K 27 and the material use meet the design requirement K 28;
Construction management K 3 includes: whether the security manager works as a full job K 31, whether the security regulation system is complete in specification K 32, whether the security supervision periodically checks K 33 according to the regulations, the configuration of the personnel is unreasonable K 34, and the quality of the construction commander K 35 comprises;
Construction environment K 4 includes: construction environment transportation K 41, surrounding environment complexity K 42, reasonable operation environment layout K 43, construction-independent personnel flow K 44, construction environment severity K 45, natural disaster possibility K 46 and surrounding buildings K 47.
Further, in step S2, the preliminary risk level quantization method is to scale the risk range of the risk factor to a specific value between 0 and 1, and the safer the value is.
Further, in step S2, the membership degree in the level is determined by using the fuzzy mathematical theory and using accurate mathematics, and the quantization standard and the corresponding quantization value of the preliminary risk level quantization are as follows:
And is dangerous: the quantization value is 1 to 0.8;
risk: the quantization value is 0.8-0.6;
The safety is compared: the quantization value is 0.6-0.4;
safety: the quantization value is 0.4-0.2;
Is very safe: the quantization value is 0.2-0.
Further, in step S3, the level of the risk event occurrence probability P is quantized as follows:
Frequent occurrences are: the probability level is 5, and the quantization value is 1-0.8;
it may happen that: the probability level is 4, and the quantization value is 0.8-0.6;
happens accidentally: the probability level is 3, and the quantization value is 0.6-0.4;
rarely occurs: probability level is 2, and quantization value is 0.4-0.2;
it is highly unlikely that: the probability level is 1, and the quantization value is 0.2-0;
the level of the degree S of consequences of the occurrence of a risk event is quantified as follows:
Catastrophic: the probability grade is E, and the quantized value is 1-0.8
Very serious: the probability grade is D, and the quantized value is 0.8-0.6
Serious: the probability grade is C, and the quantized value is 0.6-0.4
Larger: the probability grade is B, and the quantized value is 0.4-0.2
Slight: the probability grade is A, and the quantized value is 0.2-0.
Further, in step S4, the bridge engineering risk class is classified into four risk classes of high, medium and low by combining the risk event occurrence probability class and the risk event occurrence result class, which are specifically shown in the following table:
and determining the final risk level according to the table after obtaining the risk event occurrence probability P and the risk event occurrence result degree S.
Further, the specific method of step S4 is as follows:
s4.1, firstly constructing a judgment matrix, exploring and judging the importance of each risk factor through expert discussion, comparing each factor with each other under certain standard, and grading according to the importance degree, wherein the general formula of the judgment matrix is as follows:
Wherein: a ij is the relative importance of the ith and jth risk factors, where a ij＝1/aji; n is the total number of secondary risk factors;
The judging matrix A of the probability P of occurrence of the risk event comprises a first-level index judging matrix and a corresponding second-level index judging matrix, wherein the first-level index judging matrix is marked as A 0, the judging matrix of personnel quality is marked as A 1, and the judging matrix of mechanical equipment and construction materials is marked as A 2; the judgment matrix of construction management is marked as A 3; the judgment matrix of the construction environment is marked as A 4;
The judgment matrix A * of the occurrence result degree S of the risk event comprises a first-level index judgment matrix and a corresponding second-level index judgment matrix, wherein the first-level index judgment matrix is marked as A 0 *, the judgment matrix of personnel quality is marked as A 1 *, and the judgment matrix of mechanical equipment and construction materials is marked as A 2 *; the judgment matrix of construction management is marked as A 3 *; the judgment matrix of the construction environment is marked as A 4 *;
S4.2, carrying out average geometric calculation and normalization on each judgment matrix to obtain a weight vector of each risk factor, wherein the calculation formula is as follows:
in the above formula, u is an angle sign, u=0 represents a weight vector of the first-level index, and u=1-4 represents weight vectors of each second-level index respectively;
For the probability of occurrence of a risk event P, the weight vector of the primary index is denoted as W 0＝(a1,a2,a3,a4), where a 1,a2,a3,a4 is the weight of each primary index; the weight vector of the personnel quality is marked as W 1; the weight vector of the mechanical equipment and the construction material is marked as W 2; the weight vector of construction management is marked as W 3; the weight vector of the construction environment is marked as W 4;
for the occurrence result degree S of the risk event, the weight vector of the first-level index is recorded as W0 *＝(a1 *,a2 *,a3 *,a4 *),, wherein a 1 *,a2 *,a3 *,a4 * is the weight of each first-level index; the weight vector of personnel quality is marked as W 1; the weight vector of the mechanical equipment and the construction material is marked as W 2; the weight vector of construction management is marked as W 3; the weight vector of the construction environment is marked as W 4;
S4.3, multiplying the weight of each secondary index by the weight of the corresponding primary index, and obtaining a result which is the risk event occurrence probability weight and the risk event occurrence result degree weight of the corresponding secondary index quantized value, wherein the calculation formula is as follows:
w= (a 1·W1,a2·W2,a3·W3,a4·W4) formula (3)
In the above formula, W is a first weight vector composed of probability weights of occurrence of risk events, and W * is a second weight vector composed of degree weights of occurrence of risk events.
Further, in step S5, the calculation formula of the risk event occurrence probability P and the risk event occurrence result degree S is as follows:
K= (K 11,K12,…,K47) equation (5)
P=w.k T formula (6)
S=w *·KT formula (7).
According to the invention, risk factors are identified aiming at bridge construction safety, a bridge construction safety risk assessment system is constructed, a risk factor grade data sample is obtained by combining an analytic hierarchy process with expert experience, and data preprocessing is performed, so that the purpose of overcoming the subjectivity of the traditional assessment system is achieved, a BP neural network is adopted to train the processed data sample, and a scientific and accurate assessment result can be obtained under the condition of a small number of data samples.
Compared with the prior art, the invention has the following beneficial effects:
1. According to the invention, an AHP is utilized to construct an evaluation index system meeting engineering construction, and meanwhile, the nonlinear characteristic of the BP neural network is adopted to solve the problem that the risk level quantification is artificial subjectivity too large by an expert evaluation method.
2. According to the invention, a sample library required by BP neural network training is constructed by utilizing an AHP method, engineering verification shows that the method is suitable for risk level assessment of high-speed railway engineering projects, and compared with the traditional assessment method which is complex in process and subjective in conclusion, the method is more concise, practical, scientific and strict, is convenient for compiling a standardized algorithm program, and has a certain reference value and popularization value.
3. The invention can obtain the evaluation result scientifically and accurately without a large number of data samples.
Drawings
In order to more clearly illustrate the embodiments of the present invention, or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art 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 flowchart of a bridge engineering risk assessment method based on a BP neural network under a few-sample condition in an embodiment of the present invention.
FIG. 2 is a system diagram of a security risk assessment index for high-speed railway bridge construction in an embodiment of the invention.
Fig. 3 is a schematic diagram of a BP neural network model according to an embodiment of the present invention.
Fig. 4 is a diagram of a training process of a BP neural network model in an embodiment of the present invention.
Detailed Description
Embodiments of the present invention are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the invention but are not intended to limit the scope of the invention.
The present invention is illustrated below by taking a security risk assessment of high-speed railway bridge construction as an example, and it should be clearly explained that the present invention is not limited to high-speed railway bridge construction, but can be applied to any other bridge engineering construction risk assessment.
As shown in fig. 1, the invention provides a bridge engineering risk assessment method based on a BP neural network under a few sample condition, which comprises the following steps:
S1, constructing a bridge engineering construction safety risk assessment system, and identifying risk factors of bridge construction to obtain a plurality of first-level risk factor indexes, wherein each first-level risk factor index comprises a plurality of second-level risk factor indexes;
As shown in fig. 2, the risk factors are identified from the first-level indexes of aspects 4, namely: personnel quality (K 1), mechanical equipment and construction materials (K 2), construction management (K 3) and construction environment (K 4).
Each risk factor primary indicator comprises a plurality of risk factor secondary indicators, and the identified risk factors comprise 29, namely 29 secondary indicators, including:
Personnel quality (K 1): safety awareness of personnel (K 11), degree of standardization of operation (K 12), academic of personnel (K 13), experience of personnel (K 14), safety training before post (K 15), time of use (K 16), site protection measures (K 17), experience of safety equipment use (K 18), and mental in the trade (K 19);
Mechanical equipment and construction materials (K 2): quality detection of equipment approach (K 21), safety operation specification (K 22), whether mechanical strength is suitable for construction strength (K 23), equipment maintenance (K 24), material approach detection (K 25), material arrangement and storage layout (K 26), in-field material transportation (K 27) and whether material use meets design requirements (K 28);
Construction management (K 3): whether the security manager works fully (K 31), whether the security regulation is complete (K 32), whether the security supervision is checked regularly according to the regulations (K 33), the personnel configuration is unreasonable (K 34), the quality of the construction commander (K 35);
Construction environment (K 4): construction environment transportation (K 41), surrounding environment complexity (K 42), whether the operation environment layout is reasonable (K 43), personnel flow irrelevant to construction (K 44), construction environment severity (K 45), possibility of natural disasters (K 46) and whether buildings exist around (K 47).
S2, aiming at the identified risk factor secondary indexes, carrying out preliminary risk level quantification on each corresponding risk investigation item to obtain a safety risk assessment data set K consisting of quantified values of all the secondary indexes;
And (3) carrying out preliminary risk level quantification on each secondary index (K 11、K12……K47) through a mathematical fuzzy method, wherein quantification rules are shown in table 1:
Table 1 preliminary risk level quantization scoring table
S3, grading after quantitatively scoring the risk event occurrence probability P and the risk event occurrence result degree S of the risk factors, and grading the risk grade of the bridge engineering according to the risk event occurrence probability P and the risk event occurrence result degree S;
the quantitative division rule for risk factors comprises the following steps:
table 2 risk event occurrence probability P-level quantization table
Table 3 risk event occurrence outcome level S rank quantization table
The final bridge engineering security risk level is four levels in total, including: low risk, medium risk, high risk, extremely high risk. The grading rules are shown in Table 4.
TABLE 4 bridge engineering risk level criteria
S4, carrying out data preprocessing on the quantized scores of the occurrence probability P of the risk event and the occurrence consequence degree S of the risk event by using an analytic hierarchy process (AHP method) to obtain a first weight vector W and a second weight vector W *, wherein the first weight vector W consists of the occurrence probability weights of the risk event and the occurrence consequence degree weights of the risk event of each secondary index;
specifically, in step S4, the step of performing data preprocessing on the quantitative score of the occurrence probability P of the risk event and the occurrence result degree S of the risk event by using the AHP method is as follows:
s4.1, firstly constructing a judgment matrix, exploring and judging the importance of each risk factor through expert discussion, comparing each factor with each other under certain standard, and grading according to the importance degree, wherein the general formula of the judgment matrix is as follows:
Wherein: a ij is the relative importance of the ith and jth risk factors, where a ij＝1/aji; n is the total number of secondary risk factors.
The judgment matrix a for the risk event occurrence probability P is as follows:
Note that: a 0 represents a first-order index; a 1 represents personnel quality; a 2 represents mechanical equipment and construction materials; a 3 represents construction management; a 4 represents a construction environment.
The judgment matrix a * for the occurrence result degree S of the risk event is as follows:
Note that: a 0 * represents a first-order index; a 1 * represents personnel quality; a 2 * represents mechanical equipment and construction materials; a 3 * represents construction management; a 4 * represents a construction environment.
S4.2, carrying out average geometric calculation and normalization on each judgment matrix to obtain a weight vector of each risk factor, wherein the calculation formula is as follows:
In the above formula, u is an angle sign, wherein u is 0, 1, 2,3 and 4, and represent weight vectors of primary indexes, personnel quality, mechanical equipment, construction materials, construction management and construction environment respectively;
For the probability of occurrence of a risk event P, the weight vector of the primary index is denoted as W 0＝(a1,a2,a3,a4), where a 1,a2,a3,a4 is the weight of each primary index; the weight vector of the personnel quality is marked as W 1; the weight vector of the mechanical equipment and the construction material is marked as W 2; the weight vector of construction management is marked as W 3; the weight vector of the construction environment is marked as W 4;
for the occurrence result degree S of the risk event, the weight vector of the first-level index is recorded as W0 *＝(a1 *,a2 *,a3 *,a4 *),, wherein a 1 *,a2 *,a3 *,a4 * is the weight of each first-level index; the weight vector of personnel quality is marked as W 1; the weight vector of the mechanical equipment and the construction material is marked as W 2; the weight vector of construction management is marked as W 3; the weight vector of the construction environment is marked as W 4;
the weight vector W for the risk event occurrence probability P is calculated as follows:
Note that: w 0 represents a first-order index; w 1 represents personnel quality; w 2 represents mechanical equipment and construction materials; w 3 represents construction management; w 4 represents a construction environment.
W1＝(0.148,0.128,0.077,0.141,0.085,0.075,0.081,0.156,0.109)
W2＝(0.204,0.178,0.009,0.012,0.182,0.087,0.183,0.145)
W3＝(0.246,0.169,0.268,0.148,0.169)
W4＝(0.168,0.173,0.151,0.133,0.133,0.117,0.125)
W0＝(0.3563,0.1100,0.3219,0.2118)
The weight vector W * for the level S of the risk event occurrence result is as follows:
Note that: w 0 * represents a first-order index; w 1 * represents personnel quality; w 2 * represents mechanical equipment and construction materials; w 3 * represents construction management; w 4 * represents a construction environment.
W1*＝(0.095,0.229,0.042,0.069,0.096,0.100,0.247,0.090,0.032)
W2*＝(0.099,0.200,0.027,0.026,0.114,0.141,0.145,0.248)
W3*＝(0.223,0.076,0.125,0.237,0.339)
W4*＝(0.097,0.008,0.307,0.189,0.088,0.244,0.067)
W0 *＝(0.193,0.359,0.123,0.325)
S4.3, multiplying the weight of each secondary index by the weight of the corresponding primary index, and obtaining a result which is the risk event occurrence probability weight and the risk event occurrence result degree weight of the corresponding secondary index quantized value, wherein the calculation formula is as follows:
w= (a 1·W1,a2·W2,a3·W3,a4·W4) formula (3)
In the above formula, W is a first weight vector composed of probability weights of occurrence of risk events, and W * is a second weight vector composed of degree weights of occurrence of risk events.
S5, multiplying each secondary index quantized value by a corresponding risk event occurrence probability weight W and accumulating to obtain a risk event occurrence probability P of a risk factor, multiplying each secondary index quantized value by a corresponding risk event occurrence result degree weight W * and accumulating to obtain a risk event occurrence result degree S of the risk factor, and obtaining a final safety risk level according to the bridge engineering risk level divided in the step S3; specifically, the risk event occurrence probability P and the risk event occurrence result degree S are substituted into table 4 to obtain the final security risk level.
The calculation formula of the risk event occurrence probability P and the risk event occurrence result degree S is as follows:
K= (K 11,K12,…,K47) equation (5)
P=w.k T formula (6)
S=w *·KT formula (7).
S6, taking the bridge engineering safety risk assessment data set K identified in the step S1 as a sample, taking the risk event occurrence probability P and the risk event occurrence result degree S obtained in the step S5 as labels (output), constructing a training sample, collecting different bridge engineering safety risk assessment data according to the steps S1-S5, and constructing a training sample set;
s7, establishing a BP neural network model;
As shown in fig. 3, the established BP neural network model includes: an input layer, an hidden layer, and an output layer; the network model only sets one hidden layer as a single hidden layer network.
Specifically, the basic idea learning process of the BP neural network algorithm consists of two processes of forward propagation of signals and backward propagation of errors. The forward propagation is to input an input sample into an input layer, process the input sample by an implicit layer, and then transmit the data to an output layer, and if the output actual data of the output layer does not accord with the expected data, the backward propagation stage of the error is shifted. The process of adjusting the weights of the layers of forward propagation of the signal and backward propagation of the error is performed repeatedly, and is called a self-learning phase of the BP network. This process is continued until the error in the network output is reduced to an acceptable level or until a predetermined number of studies have been performed, as schematically shown in fig. 3.
In the algorithm network of the BP neural network, there is often an output error E when the actual output is different from the expected output. The formula is as follows:
It can be seen that the weight w jk、vij of each layer can affect the change error E, and the adjustment of the weight can continuously reduce the error, so that the weight adjustment calculation formula of each layer is as follows:
deltaw jk＝μ(dk-ok)ok(1-ok)yj equation (9)
Wherein: j=0, 1,2,3, …, m; i=0, 1,2,3, …, m; k=0, 1,2,3, …, l;
the constant mu epsilon (0, 1) represents the scaling factor, reflecting the training rate and delta is the error signal.
Wherein, the input vector is:
X= (X 1,x2,…,xi,…,xn)T formula (11)
The hidden layer output vector is:
Y= (Y 1,y2,…yj,…,ym)T formula (12)
The desired output vector is:
D= (D 1,d2,…dk,…,dl)T formula (13)
The output layer output vector is:
O= (O 1,o2,…ok,…,ol)T formula (14)
As shown in fig. 4, the standard BP network calculation flow is as follows:
initializing, inputting training samples, calculating network output errors, calculating error signals of each layer, adjusting weight of each layer, checking whether the network reaches a preset round, checking whether the total error of the network reaches the precision requirement, and outputting a result.
S8, inputting the sample set for bridge engineering risk assessment into a BP neural network model to perform self-learning training of the network;
29 secondary indexes of bridge engineering (high-speed railway bridge) construction risk factors are used as input sets (K 11～K47), the level quantized value of the risk event occurrence probability P of the high-speed railway bridge and the level quantized value of the risk event occurrence result degree S are 2 output sets, the node number of an hidden layer is determined by a large amount of experience and test, a fixed theoretical formula is not available, and 35 hidden node numbers which are the best hidden node number of the network model are selected through multiple attempts and verification.
The hidden layer transfer function and the output layer transfer function selected by the network model are respectively adopting 'tansig' and 'logsig' as optimal functions, and the specific formula is as follows:
tansig function formula:
log sig (x) =1/(1+e (-x)) formula (15)
Logsig function formula:
tansig (x) =2/((1+e (-2x)) -1) formula (16)
Wherein: the x implies the values of the input variables of the layer or the layer below the output layer.
The performance function adopts MSE (mean square error) function, and the specific formula is as follows:
Wherein: m is the number of samples, Y i is the true value, Is a predicted value.
Network parameter setting, training function: trainbr (bayesian regularization algorithm); maximum number of iterations: 1000; network training target error: 0.00001; learning rate: 0.0001.
S9, carrying out risk factor identification on the bridge engineering to be evaluated according to the methods of the steps S1 and S2 to obtain a safety risk evaluation data set K of the bridge engineering to be evaluated; inputting the identified safety risk assessment data set K into the BP neural network model after training is completed to obtain the risk event occurrence probability P and the risk event occurrence result degree S, obtaining the bridge engineering safety risk level through the bridge engineering risk level standard divided in the step S3, specifically substituting the risk event occurrence probability P and the risk event occurrence result degree S into the table 4, and obtaining the final bridge engineering safety risk level.
The above embodiments are only for illustrating the present invention, and are not limiting of the present invention. While the invention has been described in detail with reference to the embodiments, those skilled in the art will appreciate that various combinations, modifications, and substitutions can be made thereto without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (8)
1. The bridge engineering risk assessment method based on the BP neural network under the condition of few samples is characterized by comprising the following steps of:
S1, constructing a bridge engineering construction safety risk assessment system, and identifying risk factors of bridge construction to obtain a plurality of first-level risk factor indexes, wherein each first-level risk factor index comprises a plurality of second-level risk factor indexes;
S2, aiming at the identified risk factor secondary indexes, carrying out preliminary risk level quantification on each corresponding risk investigation item to obtain a safety risk assessment data set K consisting of quantified values of all the secondary indexes;
S3, grading after quantitatively scoring the risk event occurrence probability P and the risk event occurrence result degree S of the risk factors, and grading the risk grade of the bridge engineering according to the risk event occurrence probability P and the risk event occurrence result degree S;
S4, carrying out data preprocessing on the quantized scores of the occurrence probability P of the risk event and the occurrence consequence degree S of the risk event by using a analytic hierarchy process to obtain a first weight vector W and a second weight vector W *, wherein the first weight vector W consists of the occurrence probability weights of the risk event and the occurrence consequence degree weights of the risk event of each secondary index;
S5, multiplying each secondary index quantized value by a corresponding risk event occurrence probability weight W and accumulating to obtain a risk event occurrence probability P of the risk factors, and multiplying each secondary index quantized value by a corresponding risk event occurrence result degree weight W * and accumulating to obtain a risk event occurrence result degree S of the risk factors;
S6, taking the safety risk assessment data set K of the bridge engineering identified in the step S1 as a sample, taking the risk event occurrence probability P and the risk event occurrence result degree S obtained in the step S5 as labels, constructing a training sample, collecting safety risk assessment data of different bridge engineering according to the method of the steps S1-S5, and constructing a training sample set;
s7, establishing a BP neural network model;
S8, inputting the sample set for bridge engineering risk assessment into a BP neural network model to perform self-learning training of the network;
S9, carrying out risk factor identification on the bridge engineering to be evaluated according to the methods of the steps S1 and S2 to obtain a safety risk evaluation data set K of the bridge engineering to be evaluated; and (3) inputting the identified safety risk assessment data set K into the BP neural network model after training is completed, obtaining the occurrence probability P of the risk event and the occurrence result degree S of the risk event, and obtaining the safety risk level of the bridge engineering through the bridge engineering risk level standard divided in the step (S3).
2. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 1, wherein the bridge engineering risk assessment method is characterized by comprising the following steps of: in the step S1, four primary indexes of the identified risk factors are personnel quality K 1, mechanical equipment and construction materials K 2, construction management K 3 and construction environment K 4 respectively; wherein,
Personnel quality K 1 includes: safety awareness K 11 of personnel, operation standardization degree K 12, personnel academy K 13, personnel experience K 14, pre-post safety training K 15, working time K 16, site protection measure K 17, safety equipment use experience K 18 and salon spirit K 19;
Mechanical equipment and construction material K 2 include: the equipment approach quality detection K 21, the safety operation standard K 22, whether the mechanical strength is suitable for the construction strength K 23, the equipment maintenance K 24, the material approach detection K 25, the material arrangement and storage layout K 26, the in-field material transportation K 27 and the material use meet the design requirement K 28;
Construction management K 3 includes: whether the security manager works as a full job K 31, whether the security regulation system is complete in specification K 32, whether the security supervision periodically checks K 33 according to the regulations, the configuration of the personnel is unreasonable K 34, and the quality of the construction commander K 35 comprises;
Construction environment K 4 includes: construction environment transportation K 41, surrounding environment complexity K 42, reasonable operation environment layout K 43, construction-independent personnel flow K 44, construction environment severity K 45, natural disaster possibility K 46 and surrounding buildings K 47.
3. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 2, wherein the bridge engineering risk assessment method is characterized in that: in step S2, the preliminary risk level quantization method is to scale the risk course according to the risk factor to a specific value between 0 and 1, and the safer the smaller the value.
4. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 3, wherein the bridge engineering risk assessment method is characterized by comprising the following steps of: in step S2, the membership degree in the level is determined by using the fuzzy mathematical theory and using accurate mathematics, and the quantization standard and the corresponding quantization value of the preliminary risk level quantization are as follows:
And is dangerous: the quantization value is 1 to 0.8;
risk: the quantization value is 0.8-0.6;
The safety is compared: the quantization value is 0.6-0.4;
safety: the quantization value is 0.4-0.2;
Is very safe: the quantization value is 0.2-0.
5. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 4, wherein the bridge engineering risk assessment method is characterized in that: in step S3, the level quantization of the risk event occurrence probability P is as follows:
Frequent occurrences are: the probability level is 5, and the quantization value is 1-0.8;
it may happen that: the probability level is 4, and the quantization value is 0.8-0.6;
happens accidentally: the probability level is 3, and the quantization value is 0.6-0.4;
rarely occurs: probability level is 2, and quantization value is 0.4-0.2;
it is highly unlikely that: the probability level is 1, and the quantization value is 0.2-0;
the level of the degree S of consequences of the occurrence of a risk event is quantified as follows:
Catastrophic: the probability grade is E, and the quantized value is 1-0.8
Very serious: the probability grade is D, and the quantized value is 0.8-0.6
Serious: the probability grade is C, and the quantized value is 0.6-0.4
Larger: the probability grade is B, and the quantized value is 0.4-0.2
Slight: the probability grade is A, and the quantized value is 0.2-0.
6. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 5, wherein the bridge engineering risk assessment method is characterized in that: in step S4, the risk class of the bridge engineering is classified into four risk classes of high, medium and low by combining the risk event occurrence probability class and the risk event occurrence result class, which are specifically shown in the following table:
and determining the final risk level according to the table after obtaining the risk event occurrence probability P and the risk event occurrence result degree S.
7. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 2, wherein the bridge engineering risk assessment method is characterized in that: the specific method of step S4 is as follows:
s4.1, firstly constructing a judgment matrix, exploring and judging the importance of each risk factor through expert discussion, comparing each factor with each other under certain standard, and grading according to the importance degree, wherein the general formula of the judgment matrix is as follows:
Wherein: a ij is the relative importance of the ith and jth risk factors, where a ij＝1/aji; n is the total number of secondary risk factors;
The judging matrix A of the probability P of occurrence of the risk event comprises a first-level index judging matrix and a corresponding second-level index judging matrix, wherein the first-level index judging matrix is marked as A 0, the judging matrix of personnel quality is marked as A 1, and the judging matrix of mechanical equipment and construction materials is marked as A 2; the judgment matrix of construction management is marked as A 3; the judgment matrix of the construction environment is marked as A 4;
The judgment matrix A * of the occurrence result degree S of the risk event comprises a first-level index judgment matrix and a corresponding second-level index judgment matrix, wherein the first-level index judgment matrix is marked as A 0 *, the judgment matrix of personnel quality is marked as A 1 *, and the judgment matrix of mechanical equipment and construction materials is marked as A 2 *; the judgment matrix of construction management is marked as A 3 *; the judgment matrix of the construction environment is marked as A 4 *;
S4.2, carrying out average geometric calculation and normalization on each judgment matrix to obtain a weight vector of each risk factor, wherein the calculation formula is as follows:
in the above formula, u is an angle sign, u=0 represents a weight vector of the first-level index, and u=1-4 represents weight vectors of each second-level index respectively;
For the probability of occurrence of a risk event P, the weight vector of the primary index is denoted as W 0＝(a1,a2,a3,a4), where a 1,a2,a3,a4 is the weight of each primary index; the weight vector of the personnel quality is marked as W 1; the weight vector of the mechanical equipment and the construction material is marked as W 2; the weight vector of construction management is marked as W 3; the weight vector of the construction environment is marked as W 4;
For the occurrence result degree S of the risk event, the weight vector of the first-level index is recorded as W0 *＝(a1 *,a2 *,a3 *,a4 *),, wherein a 1 *,a2 *,a3 *,a4 * is the weight of each first-level index; the weight vector of personnel quality is marked as W 1; the weight vector of the mechanical equipment and the construction material is marked as W 2; the weight vector of construction management is marked as W 3; the weight vector of the construction environment is marked as W 4;
S4.3, multiplying the weight of each secondary index by the weight of the corresponding primary index, and obtaining a result which is the risk event occurrence probability weight and the risk event occurrence result degree weight of the corresponding secondary index quantized value, wherein the calculation formula is as follows:
w= (a 1·W1,a2·W2,a3·W3,a4·W4) formula (3)
W*＝(a1 *·W1 *,a2 *·W2 *,a3 *·W3 *,a4 *·W4 *) Formula (4)
In the above formula, W is a first weight vector composed of probability weights of occurrence of risk events, and W * is a second weight vector composed of degree weights of occurrence of risk events.
8. The bridge engineering risk assessment method based on the BP neural network algorithm according to claim 7, wherein the bridge engineering risk assessment method is characterized in that: in step S5, the calculation formula of the risk event occurrence probability P and the risk event occurrence result degree S is as follows:
K= (K 11,K12,…,K47) equation (5)
P=w.k T formula (6)
S=w *·KT formula (7).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210727541.9A CN115034648B (en) | 2022-06-24 | 2022-06-24 | Bridge engineering risk assessment method based on BP neural network under condition of few samples |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210727541.9A CN115034648B (en) | 2022-06-24 | 2022-06-24 | Bridge engineering risk assessment method based on BP neural network under condition of few samples |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115034648A CN115034648A (en) | 2022-09-09 |
| CN115034648B true CN115034648B (en) | 2024-05-17 |
Family
ID=83127151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210727541.9A Active CN115034648B (en) | 2022-06-24 | 2022-06-24 | Bridge engineering risk assessment method based on BP neural network under condition of few samples |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115034648B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105335818A (en) * | 2015-10-21 | 2016-02-17 | 江苏省电力公司 | Power transmission and transformation project cost risk assessment and forecasting method based on BP neural algorithm |
| CN107085644A (en) * | 2017-04-24 | 2017-08-22 | 武汉理工大学 | A kind of concrete bridge Cantilever Construction methods of risk assessment under complicated risk source |
| CN111598352A (en) * | 2020-05-25 | 2020-08-28 | 哈尔滨工业大学 | Concrete beam type bridge comprehensive evaluation method based on Bayesian network |
| CN111861238A (en) * | 2020-07-27 | 2020-10-30 | 东北财经大学 | Expressway bridge engineering risk assessment method and device and computer equipment |
-
2022
- 2022-06-24 CN CN202210727541.9A patent/CN115034648B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105335818A (en) * | 2015-10-21 | 2016-02-17 | 江苏省电力公司 | Power transmission and transformation project cost risk assessment and forecasting method based on BP neural algorithm |
| CN107085644A (en) * | 2017-04-24 | 2017-08-22 | 武汉理工大学 | A kind of concrete bridge Cantilever Construction methods of risk assessment under complicated risk source |
| CN111598352A (en) * | 2020-05-25 | 2020-08-28 | 哈尔滨工业大学 | Concrete beam type bridge comprehensive evaluation method based on Bayesian network |
| CN111861238A (en) * | 2020-07-27 | 2020-10-30 | 东北财经大学 | Expressway bridge engineering risk assessment method and device and computer equipment |
Non-Patent Citations (2)
| Title |
|---|
| 川藏铁路桥隧施工安全风险评价;张锦;徐君翔;;安全与环境学报;20200225(第01期);全文 * |
| 桥梁风险评估中BP神经网络算法研究;黄豪;;福建交通科技;20130820(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115034648A (en) | 2022-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111582718B (en) | Cable channel fire risk assessment method and device based on network analytic hierarchy process | |
| CN106844663B (en) | Ship collision risk assessment method and system based on data mining | |
| CN107886235A (en) | A kind of Fire risk assessment method for coupling certainty and uncertainty analysis | |
| CN110348752B (en) | Large industrial system structure safety assessment method considering environmental interference | |
| CN102456158A (en) | Security assessment method for air traffic management (ATM) information system based on ANNBP (Artificial Neural Network Blood Pressure) model | |
| CN110619467B (en) | Power equipment state evaluation method based on alarm big data information | |
| CN113887908A (en) | Bridge risk assessment method considering subjective and objective cross fusion weight | |
| Li et al. | Real-time warning and risk assessment of tailings dam disaster status based on dynamic hierarchy-grey relation analysis | |
| CN111178725A (en) | Protective equipment state early warning method based on analytic hierarchy process | |
| CN111861238A (en) | Expressway bridge engineering risk assessment method and device and computer equipment | |
| CN112257914B (en) | Aviation safety causal prediction method based on random forest | |
| CN115018384A (en) | Building site security risk assessment method and system | |
| CN112149976A (en) | Power grid accurate investment project decision method based on artificial intelligence | |
| CN115471097A (en) | Data-driven underground local area safety state evaluation method | |
| CN115018291A (en) | Large bridge engineering risk assessment method | |
| CN114862267A (en) | Evaluation method and system of oil and gas pipeline alarm management system | |
| CN107256445A (en) | Analysis method, the device and system of production environment security | |
| Zheng et al. | A hybrid approach for evaluating faulty behavior risk of high‐risk operations using ANP and evidence theory | |
| CN115034648B (en) | Bridge engineering risk assessment method based on BP neural network under condition of few samples | |
| Zhao et al. | A grey BP neural network-based model for prediction of court decision service rate | |
| CN117009923A (en) | Tunnel advanced geological disaster intelligent judging and identifying method based on multi-source information fusion | |
| CN116596346A (en) | Expressway tunnel operation safety evaluation method | |
| CN114266370A (en) | Method and system for generating fault handling plan of power grid equipment in typhoon meteorological environment on line and storage medium | |
| CN114065624A (en) | Intelligent evaluation method for pipe network integrity management level of refining device | |
| Wang et al. | Comprehensive safety risk evaluation of fireworks production enterprises using the frequency-based ANP and BPNN |
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 |