CN112464576A - Dam risk assessment method and equipment based on Bayesian network - Google Patents

Dam risk assessment method and equipment based on Bayesian network Download PDF

Info

Publication number
CN112464576A
CN112464576A CN202011497870.6A CN202011497870A CN112464576A CN 112464576 A CN112464576 A CN 112464576A CN 202011497870 A CN202011497870 A CN 202011497870A CN 112464576 A CN112464576 A CN 112464576A
Authority
CN
China
Prior art keywords
dam
risk
node
occurrence
bayesian network
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
Application number
CN202011497870.6A
Other languages
Chinese (zh)
Other versions
CN112464576B (en
Inventor
杨海云
唐贤琪
施玉群
陈泽钦
罗志华
何金平
吴凡
吴在强
林亚涛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan University WHU
Electric Power Research Institute of State Grid Fujian Electric Power Co Ltd
State Grid Fujian Electric Power Co Ltd
Original Assignee
Wuhan University WHU
Electric Power Research Institute of State Grid Fujian Electric Power Co Ltd
State Grid Fujian Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Wuhan University WHU, Electric Power Research Institute of State Grid Fujian Electric Power Co Ltd, State Grid Fujian Electric Power Co Ltd filed Critical Wuhan University WHU
Priority to CN202011497870.6A priority Critical patent/CN112464576B/en
Publication of CN112464576A publication Critical patent/CN112464576A/en
Application granted granted Critical
Publication of CN112464576B publication Critical patent/CN112464576B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
    • 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
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
    • Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Evolutionary Computation (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Development Economics (AREA)
  • Geometry (AREA)
  • Educational Administration (AREA)
  • Medical Informatics (AREA)
  • Game Theory and Decision Science (AREA)
  • Software Systems (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a dam risk assessment method based on a Bayesian network, which comprises the following steps: selecting a plurality of main risk factors as nodes of the Bayesian network according to the actual condition of the dam, and dividing the state of each node according to the occurrence characteristics of each risk factor; dividing each node into an input node, an intermediate node and an output node according to the relevance among the risk factors, wherein the input node is a risk source influencing the dam safety, the output node is an object of dam risk assessment, and the intermediate node is an object carrying out risk transmission on the output node under the influence of the input node; assigning the prior probability of the input node, and assigning the conditional probability of the input node and the intermediate node and the conditional probability between the intermediate node and the output node; and carrying out risk transmission through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and carrying out risk evaluation of the dam.

Description

Dam risk assessment method and equipment based on Bayesian network
Technical Field
The invention relates to a dam risk assessment method and device based on a Bayesian network, and belongs to the technical field of hydraulic engineering.
Background
With the continuous deepening and development of dam safety problem research, dam safety management is shifting from a traditional mode to a modern mode. In a traditional dam safety management mode, whether a dam structure has potential safety hazards or not and whether danger removal reinforcement is needed or not are measured by taking a dam safety coefficient with certainty significance as a standard; the modern dam safety management mode takes dam acceptable risk with uncertain significance as a standard to determine whether the dam risk is within an acceptable risk range and whether measures need to be taken to reduce the dam risk. The basic content of dam risk analysis mainly comprises risk identification, risk assessment, risk standard, risk management and control and the like. The risk identification is to identify dam risk elements and the influence thereof, the risk assessment is to estimate the dam accident probability and the accident consequence, the risk standard is to determine the acceptable risk degree of the dam, the risk control is to manage and control the dam risk, and the risk assessment is the core of dam risk analysis.
Although the research on dam risk analysis starts late in China, abundant research results are obtained. Li Lei, Wang ren clock, etc. combine the specific situation and foreign experience of using of our country, have proposed the conversion relation between probability and qualitative description suitable for event occurrence of our country; a comprehensive evaluation model of dam break consequences is provided based on a risk matrix and a Borda sequence value method, such as Chenyue and considering rush time, so that a certain reference basis is provided for dam risk management; the Zhoujian is equal to the national and international risk standards of system research, and the social acceptable risk standard of the cascade reservoir group is provided, so that the blank of the engineering design safety standard of the extra-high dam is filled; based on the reliability theory, Chenzuyu and the like develop research aiming at the safety coefficient standards of the normal working condition and the earthquake working condition of the ultra-high earth-rock dam and provide a safety criterion for the anti-skidding stability of the dam slope of the ultra-high earth-rock dam; fiberella asiatica and the like establish an earth and rockfill dam safety comprehensive evaluation model based on uncertain measurement and an analytic hierarchy process aiming at subjectivity and unknown in the judgment of risk factors of the earth and rockfill dam, and provide a new idea for the safety management decision of the earth and rockfill dam. The method enriches and innovates the dam safety risk analysis theory from different angles, provides scientific basis for dam risk management, and has less consideration to the uncertainty and relevance problems of the dam system risk source. The traditional dam failure probability analysis method, such as an event tree method, a fault tree method and the like, has the advantages of visual expression, strong logicality and the like, but has certain limitation when a complex system formed by a plurality of risk factors is analyzed.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a dam risk assessment method based on a Bayesian network, which can overcome the defects of event trees, fault trees and other methods by using the Bayesian network, solve the problems of uncertainty, relevance, event polymorphism and the like of risk factors, accurately and efficiently realize bidirectional reasoning and provide an effective way for dam safety risk analysis.
The technical scheme of the invention is as follows:
the first technical scheme is as follows:
a dam risk assessment method based on a Bayesian network comprises the following steps:
selecting a plurality of main risk factors as nodes of the Bayesian network according to the actual situation of the dam, and dividing the states of the nodes according to the occurrence characteristics of the risk factors;
determining a Bayesian network structure, and dividing each node into an input node, an intermediate node and an output node according to the relevance among the risk factors, wherein the input node is a risk source influencing the safety of the dam, the output node is an object of risk evaluation of the dam, and the intermediate node is an object of risk transmission to the output node under the influence of the input node;
probability assignment, namely assigning the prior probability of the input node according to the actual condition, and assigning the conditional probability of the input node and the intermediate node and the conditional probability of the intermediate node and the output node according to the association degree between the nodes;
and risk transfer, namely performing risk transfer through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and performing risk evaluation of the dam.
Further, the risk sources influencing the safety of the dam comprise over-standard flood, gate faults, bank instability and surge, rolling layer shearing damage, dam body overturning instability, dam foundation uplift pressure over-standard and seepage damage around the dam; the dam risk assessment object is dam accident; and carrying out dam overtopping, dam structure damage and dam penetration damage on the objects for risk transfer on the output node under the influence of the input node.
Further, the step of dividing the state of each node according to the occurrence characteristics of each risk factor specifically includes:
dividing the states of the overproof flood, the gate fault, the bank instability and the surge and the dam body overturning instability into occurrence and non-occurrence;
dividing the state of the rolling layer shearing damage into serious state and non-occurrence state or non-serious state;
dividing states of over-standard uplift pressure of the dam foundation and seepage damage around the dam into serious states and non-occurrence states or serious states;
dividing the states of the dam overtopping, dam structure damage and dam seepage damage into non-occurrence and non-occurrence;
dividing the state of the dam crash includes not occurring and not occurring.
Further, the method also comprises a step of determining main risk factors threatening the safety of the dam, and specifically comprises the following steps:
according to the formula
Figure BDA0002842713870000041
Calculating the importance of risk factors in each node;
wherein, the aiRepresenting the importance of the risk factor represented by the ith node, p0Representing the probability of occurrence, p, of the risk factor represented by the parent nodeiRepresenting the occurrence probability of the risk factor represented by the parent node when the risk factor represented by the ith node does not occur;
and (4) carrying out ranking according to the importance of the risk factors in each node to obtain the main risk factors threatening the safety of the dam.
The second technical scheme is as follows:
a bayesian network based dam risk assessment device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program implementing the steps of:
selecting a plurality of main risk factors as nodes of the Bayesian network according to the actual situation of the dam, and dividing the states of the nodes according to the occurrence characteristics of the risk factors;
determining a Bayesian network structure, and dividing each node into an input node, an intermediate node and an output node according to the relevance among the risk factors, wherein the input node is a risk source influencing the safety of the dam, the output node is an object of risk evaluation of the dam, and the intermediate node is an object of risk transmission to the output node under the influence of the input node;
probability assignment, namely assigning the prior probability of the input node according to the actual condition, and assigning the conditional probability of the input node and the intermediate node and the conditional probability of the intermediate node and the output node according to the association degree between the nodes;
and risk transfer, namely performing risk transfer through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and performing risk evaluation of the dam.
Further, the risk sources influencing the safety of the dam comprise over-standard flood, gate faults, bank instability and surge, rolling layer shearing damage, dam body overturning instability, dam foundation uplift pressure over-standard and seepage damage around the dam; the dam risk assessment object is dam accident; and carrying out dam overtopping, dam structure damage and dam penetration damage on the objects for risk transfer on the output node under the influence of the input node.
Further, the step of dividing the state of each node according to the occurrence characteristics of each risk factor specifically includes:
dividing the states of the overproof flood, the gate fault, the bank instability and the surge and the dam body overturning instability into occurrence and non-occurrence;
dividing the state of the rolling layer shearing damage into serious state and non-occurrence state or non-serious state;
dividing states of over-standard uplift pressure of the dam foundation and seepage damage around the dam into serious states and non-occurrence states or serious states;
dividing the states of the dam overtopping, dam structure damage and dam seepage damage into non-occurrence and non-occurrence;
dividing the state of the dam crash includes not occurring and not occurring.
Further, the method also comprises a step of determining main risk factors threatening the safety of the dam, and specifically comprises the following steps:
according to the formula
Figure BDA0002842713870000061
Calculating the importance of risk factors in each node;
wherein, the aiRepresenting the importance of the risk factor represented by the ith node, p0Representing the probability of occurrence, p, of the risk factor represented by the parent nodeiRepresenting the occurrence probability of the risk factor represented by the parent node when the risk factor represented by the ith node does not occur;
and (4) carrying out ranking according to the importance of the risk factors in each node to obtain the main risk factors threatening the safety of the dam.
The invention has the following beneficial effects:
1. according to the dam risk assessment method and device based on the Bayesian network, the Bayesian network node, the Bayesian network structure, the Bayesian network parameter, the risk factor, the relevance of the risk factor and the relevance degree of the risk factor are established, and the dam risk assessment model of the water-east hydropower station based on the Bayesian network is established. The model can visually reflect the relevance among risk factors, obtain the total safety risk of the large dam of the Francis hydropower station, help to identify the main risk source of the dam accident, and can quickly and effectively realize the safety risk assessment and management decision of the Francis hydropower station large dam.
2. The dam risk assessment method and the dam risk assessment equipment based on the Bayesian network calculate the importance of risk factors in each node related to dam failure, perform ranking according to the importance, determine main risk factors threatening the dam safety, and provide a theoretical basis for dam safety decision.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention;
FIG. 2 is a diagram of a Bayesian network configuration of an embodiment of the present invention;
FIG. 3 is an exemplary diagram of a basic Bayesian network architecture;
fig. 4 is an exemplary diagram of a risk transfer calculation result of the dam a according to the embodiment of the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and the specific embodiments.
The first embodiment is as follows:
referring to fig. 1, a method for evaluating dam risk based on bayesian network includes the following steps:
node selection and state division, wherein for a dam, a plurality of risk factors influencing the safety of the dam need to be considered at the same time, main risk factors are selected from the dam as nodes of the Bayesian network, and the states of the nodes are divided according to the occurrence characteristics of the risk factors;
determining a Bayesian network structure, wherein nodes of the Bayesian network can be divided into input nodes, intermediate nodes and output nodes according to a causal control relationship; correspondingly, according to the relevance among the risk factors, dividing the risk factors in the network into basic risks, intermediate risks and target risks, and respectively corresponding to input nodes, intermediate nodes and output nodes of the Bayesian network one by one; the basic risk is a risk source influencing the safety of the dam, the target risk is an object of risk assessment, and the intermediate risk plays a role in risk transfer under the influence of the basic risk; and carrying out classification induction on the dam risk factors by utilizing expert knowledge and engineering experience to obtain basic risk, intermediate risk and target risk, thereby determining the basic structure of the Bayesian network.
Probability assignment, namely assigning the prior probability of the input node according to the actual condition, wherein the prior probability of the input node corresponds to the edge probability of a risk source in the dam risk system and can be determined by analyzing the dam operation state; assigning values to the conditional probabilities of the input node and the intermediate node and the conditional probabilities of the intermediate node and the output node according to the association degree between the nodes, wherein the conditional probabilities reflect the association degree of the father node and the son node, can be regarded as the un-normalized joint probabilities of the father node and the son node, and can be obtained according to the historical accident statistical data of the dam and the calculation and analysis result of the dam structure;
and risk transfer, namely performing risk transfer through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and performing comparative analysis on the state and the corresponding probability with the existing dam risk standard to obtain a corresponding evaluation result.
The embodiment utilizes the Bayesian network to overcome the defects of event trees, fault trees and other methods, solves the problems of uncertainty, relevance, event polymorphism and the like of risk factors, accurately and efficiently realizes bidirectional reasoning, and provides an effective way for dam safety risk analysis.
Example two:
referring to fig. 2 specifically, in this embodiment, on the basis of the first embodiment, further selected risk sources affecting the safety of the dam include overproof flood, gate failure, bank instability and surge, rolling bed shearing damage, dam body overturning instability, dam foundation uplift pressure overproof, and seepage damage around the dam; the dam risk assessment object is dam accident; and carrying out dam overtopping, dam structure damage and dam penetration damage on the objects for risk transfer on the output node under the influence of the input node.
Further, referring to table 1, in this embodiment, the step of dividing the state of each node according to the occurrence characteristics of each risk factor specifically includes:
dividing the states of the overproof flood, the gate fault, the bank instability and the surge and the dam body overturning instability into occurrence and non-occurrence;
dividing the state of the rolling layer shearing damage into serious state and non-occurrence state or non-serious state;
dividing states of over-standard uplift pressure of the dam foundation and seepage damage around the dam into serious states and non-occurrence states or serious states;
dividing the states of the dam overtopping, dam structure damage and dam seepage damage into non-occurrence and non-occurrence;
dividing the state of the dam crash includes not occurring and not occurring.
TABLE 1 Bayesian network node State and Risk partitioning
Figure BDA0002842713870000091
In this embodiment, taking the dam a as an example, probability assignment of each node is performed, and by analyzing the operating state of the dam a and combining historical statistical data of a crash of the dam a, the prior probability of each input node and the conditional probability value (taking the structure-destroyed node as an example) between parent and child nodes are obtained, respectively, and the results are shown in tables 2 and 3.
TABLE 2 prior probability values of the respective nodes
Figure BDA0002842713870000092
TABLE 3 conditional probability distribution of structural failure nodes
Figure BDA0002842713870000093
Figure BDA0002842713870000101
Referring specifically to fig. 3, fig. 3 shows the basic structure of a bayesian network, where nodes a, b, c respectively represent different variables in the theoretical domain, and directed arcs L1、L2The association between variables is expressed, and the node a controls the nodes b and c, and is called a parent node of the nodes b and c, the nodes b and c are called a child node of the node a, and the node without the parent node is called a root node, such as the node a in fig. 1. The strength of the dependency between nodes is quantified by a conditional probability, e.g., P (b | a) represents the degree of influence of an event represented by the parent node a on an event represented by the child node.
The main purpose of bayesian network modeling is to obtain a joint probability distribution of all node variables in the network. Assume that there is a set of variables X ═ X1,x2,...,xNThe variable x is obtained according to a Bayes formula and a total probability formulaiAnd variable xjThe probability relationship between the two is shown in formula (1); by Bayesian networksThe joint probabilities of all variables in the set X can be obtained by multiplying the respective local conditional probabilities, see equation (2).
Figure BDA0002842713870000102
Figure BDA0002842713870000103
In the formula, P (x)i)、P(xj) Are respectively a variable xiAnd variable xjA priori of, P (x)j|xi) Is a variable xiAnd variable xjConditional probability of P (x)i|xj) Is a variable xiN is the variable xiThe number of possible states;
Figure BDA0002842713870000104
represents the variable xiAll cause events set.
Based on the above principle, the joint probability distribution of the simple bayesian network shown in fig. 1 is:
P(a,b,c)＝P(a)P(b|a)P(c|a) (3)
in addition, the Bayesian network can also update the occurrence probability of other nodes through network propagation according to the observation result of the known node, so as to realize the function of reverse reasoning. For example, for the bayesian network shown in fig. 1, assuming that the possible occurrence states of the node c are c 1-cn, the posterior probabilities of the nodes a and b when the occurrence state of the node c is "c 1" are:
Figure BDA0002842713870000111
referring to fig. 4 specifically, in this embodiment, based on the basic principle of the bayesian network, inference software (for example, Hugin software) is used to perform risk transfer calculation on the dam a, so that the dam failure probability of the dam a is 6.4 × 10-5, the risk values of the nodes are shown in fig. 3, and the probabilities in the graph are percentages;
according to the research result of the risk standard, the tolerable risk standard of the hydropower station dam in China is 10-4/year, and the acceptable risk standard of the hydropower station dam in China is 10-6/year, so that the current total risk level of the dam A is between the tolerable risk and the acceptable risk.
Further, in this embodiment, the method further includes a step of determining a main risk factor threatening the safety of the dam, specifically:
according to the formula
Figure BDA0002842713870000112
Calculating the importance of risk factors in each node;
wherein, the aiRepresenting the importance of the risk factor represented by the ith node, p0Representing the probability of occurrence, p, of the risk factor represented by the parent nodeiRepresenting the occurrence probability of the risk factor represented by the parent node when the risk factor represented by the ith node does not occur;
and (4) carrying out ranking according to the importance of the risk factors in each node to obtain the main risk factors threatening the safety of the dam. The importance of the node reflects the degree of influence of the risk factors represented by the node on the system. The node importance degree can be rapidly and accurately analyzed by utilizing the reverse reasoning function of the Bayesian network, so that the main risk factors threatening the dam safety are determined, and a theoretical basis is provided for the dam safety decision.
The importance of the risk factors in each node calculated in the present embodiment is shown in table 4;
TABLE 4 importance of crash risk factors for large dam of hydro-east hydropower station
Figure BDA0002842713870000121
As can be seen from table 4, the risk factors with the greatest node importance are rolling bed shear failure and dam overturning instability, which are 43.63%, and it can be seen that the rolling bed shear failure and the dam overturning instability are the main risks that may cause the failure of the dam a. Based on the above, in the operation process of the dam A, the two risk factors are focused, the reinforcing treatment is carried out on the rolling layer surface of the dam body if necessary, and the maintenance treatment is carried out on the downstream pit flushing.
Example three:
referring to fig. 1, a bayesian network based dam risk assessment device comprises a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the program:
node selection and state division, wherein for a dam, a plurality of risk factors influencing the safety of the dam need to be considered at the same time, main risk factors are selected from the dam as nodes of the Bayesian network, and the states of the nodes are divided according to the occurrence characteristics of the risk factors;
determining a Bayesian network structure, wherein nodes of the Bayesian network can be divided into input nodes, intermediate nodes and output nodes according to a causal control relationship; correspondingly, according to the relevance among the risk factors, dividing the risk factors in the network into basic risks, intermediate risks and target risks, and respectively corresponding to input nodes, intermediate nodes and output nodes of the Bayesian network one by one; the basic risk is a risk source influencing the safety of the dam, the target risk is an object of risk assessment, and the intermediate risk plays a role in risk transfer under the influence of the basic risk; and carrying out classification induction on the dam risk factors by utilizing expert knowledge and engineering experience to obtain basic risk, intermediate risk and target risk, thereby determining the basic structure of the Bayesian network.
Probability assignment, namely assigning the prior probability of the input node according to the actual condition, wherein the prior probability of the input node corresponds to the edge probability of a risk source in the dam risk system and can be determined by analyzing the dam operation state; assigning values to the conditional probabilities of the input node and the intermediate node and the conditional probabilities of the intermediate node and the output node according to the association degree between the nodes, wherein the conditional probabilities reflect the association degree of the father node and the son node, can be regarded as the un-normalized joint probabilities of the father node and the son node, and can be obtained according to the historical accident statistical data of the dam and the calculation and analysis result of the dam structure;
and risk transfer, namely performing risk transfer through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and performing comparative analysis on the state and the corresponding probability with the existing dam risk standard to obtain a corresponding evaluation result.
The embodiment utilizes the Bayesian network to overcome the defects of event trees, fault trees and other methods, solve the problems of uncertainty, relevance, event polymorphism and the like of risk factors, accurately and efficiently realize bidirectional reasoning, and provide an effective way for dam safety risk analysis
Example four:
referring to fig. 2 specifically, in this embodiment, on the basis of the third embodiment, further selected risk sources affecting the safety of the dam include overproof flood, gate failure, bank instability and surge, rolling bed shearing damage, dam body overturning instability, dam foundation uplift pressure overproof, and seepage damage around the dam; the dam risk assessment object is dam accident; and carrying out dam overtopping, dam structure damage and dam penetration damage on the objects for risk transfer on the output node under the influence of the input node.
Further, referring to table 1, in this embodiment, the step of dividing the state of each node according to the occurrence characteristics of each risk factor specifically includes:
dividing the states of the overproof flood, the gate fault, the bank instability and the surge and the dam body overturning instability into occurrence and non-occurrence;
dividing the state of the rolling layer shearing damage into serious state and non-occurrence state or non-serious state;
dividing states of over-standard uplift pressure of the dam foundation and seepage damage around the dam into serious states and non-occurrence states or serious states;
dividing the states of the dam overtopping, dam structure damage and dam seepage damage into non-occurrence and non-occurrence;
dividing the state of the dam crash includes not occurring and not occurring.
TABLE 1 Bayesian network node State and Risk partitioning
Figure BDA0002842713870000141
In this embodiment, taking the dam a as an example, probability assignment of each node is performed, and by analyzing the operating state of the dam a and combining historical statistical data of a crash of the dam a, the prior probability of each input node and the conditional probability value (taking the structure-destroyed node as an example) between parent and child nodes are obtained, respectively, and the results are shown in tables 2 and 3.
TABLE 2 prior probability values of the respective nodes
Figure BDA0002842713870000151
TABLE 3 conditional probability distribution of structural failure nodes
Figure BDA0002842713870000152
Referring specifically to fig. 3, fig. 3 shows the basic structure of a bayesian network, where nodes a, b, c respectively represent different variables in the theoretical domain, and directed arcs L1、L2The association between variables is expressed, and the node a controls the nodes b and c, and is called a parent node of the nodes b and c, the nodes b and c are called a child node of the node a, and the node without the parent node is called a root node, such as the node a in fig. 1. The strength of the dependency between nodes is quantified by a conditional probability, e.g., P (b | a) represents the degree of influence of an event represented by the parent node a on an event represented by the child node.
The main purpose of bayesian network modeling is to obtain a joint probability distribution of all node variables in the network. Assume that there is a set of variables X ═ X1,x2,...,xNThe variable x is obtained according to a Bayes formula and a total probability formulaiAnd variable xjThe probability relationship between the two is shown in formula (1); joint summary of all variables in set X by conditional independence assumptions in a Bayesian networkThe rate can be obtained by multiplying the respective local conditional probabilities, see equation (2).
Figure BDA0002842713870000161
Figure BDA0002842713870000162
In the formula, P (x)i)、P(xj) Are respectively a variable xiAnd variable xjA priori of, P (x)j|xi) Is a variable xiAnd variable xjConditional probability of P (x)i|xj) Is a variable xiN is the variable xiThe number of possible states;
Figure BDA0002842713870000163
represents the variable xiAll cause events set.
Based on the above principle, the joint probability distribution of the simple bayesian network shown in fig. 1 is:
P(a,b,c)＝P(a)P(b|a)P(c|a) (3)
in addition, the Bayesian network can also update the occurrence probability of other nodes through network propagation according to the observation result of the known node, so as to realize the function of reverse reasoning. For example, for the bayesian network shown in fig. 1, assuming that the possible occurrence states of the node c are c 1-cn, the posterior probabilities of the nodes a and b when the occurrence state of the node c is "c 1" are:
Figure BDA0002842713870000164
referring to fig. 4 specifically, in this embodiment, based on the basic principle of the bayesian network, inference software (for example, Hugin software) is used to perform risk transfer calculation on the dam a, so that the dam failure probability of the dam a is 6.4 × 10-5, the risk values of the nodes are shown in fig. 3, and the probabilities in the graph are percentages;
according to the research result of the risk standard, the tolerable risk standard of the hydropower station dam in China is 10-4/year, and the acceptable risk standard of the hydropower station dam in China is 10-6/year, so that the current total risk level of the dam A is between the tolerable risk and the acceptable risk.
Further, in this embodiment, the method further includes a step of determining a main risk factor threatening the safety of the dam, specifically:
according to the formula
Figure BDA0002842713870000171
Calculating the importance of risk factors in each node;
wherein, the aiRepresenting the importance of the risk factor represented by the ith node, p0Representing the probability of occurrence, p, of the risk factor represented by the parent nodeiRepresenting the occurrence probability of the risk factor represented by the parent node when the risk factor represented by the ith node does not occur;
and (4) carrying out ranking according to the importance of the risk factors in each node to obtain the main risk factors threatening the safety of the dam. The importance of the node reflects the degree of influence of the risk factors represented by the node on the system. The node importance degree can be rapidly and accurately analyzed by utilizing the reverse reasoning function of the Bayesian network, so that the main risk factors threatening the dam safety are determined, and a theoretical basis is provided for the dam safety decision.
The importance of the risk factors in each node calculated in the present embodiment is shown in table 4;
TABLE 4 importance of crash risk factors for large dam of hydro-east hydropower station
Figure BDA0002842713870000172
As can be seen from table 4, the risk factors with the greatest node importance are rolling bed shear failure and dam overturning instability, which are 43.63%, and it can be seen that the rolling bed shear failure and the dam overturning instability are the main risks that may cause the failure of the dam a. Based on the above, in the operation process of the dam A, the two risk factors are focused, the reinforcing treatment is carried out on the rolling layer surface of the dam body if necessary, and the maintenance treatment is carried out on the downstream pit flushing.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes performed by the present specification and drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims (8)

1. A dam risk assessment method based on a Bayesian network is characterized by comprising the following steps:
selecting a plurality of main risk factors as nodes of the Bayesian network according to the actual situation of the dam, and dividing the states of the nodes according to the occurrence characteristics of the risk factors;
determining a Bayesian network structure, and dividing each node into an input node, an intermediate node and an output node according to the relevance among the risk factors, wherein the input node is a risk source influencing the safety of the dam, the output node is an object of risk evaluation of the dam, and the intermediate node is an object of risk transmission to the output node under the influence of the input node;
probability assignment, namely assigning the prior probability of the input node according to the actual condition, and assigning the conditional probability of the input node and the intermediate node and the conditional probability of the intermediate node and the output node according to the association degree between the nodes;
and risk transfer, namely performing risk transfer through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and performing risk evaluation of the dam.
2. The Bayesian network-based dam risk assessment method according to claim 1, wherein:
the risk sources influencing the safety of the dam comprise overproof flood, gate faults, unstable surge of the reservoir bank, shearing damage of a rolling layer, overturning instability of a dam body, overproof uplift pressure of a dam foundation and seepage damage around the dam;
the dam risk assessment object is dam accident;
and carrying out dam overtopping, dam structure damage and dam penetration damage on the objects for risk transfer on the output node under the influence of the input node.
3. The dam risk assessment method based on the bayesian network according to claim 2, wherein the step of dividing the state of each node according to the occurrence characteristics of each risk factor specifically comprises:
dividing the states of the overproof flood, the gate fault, the bank instability and the surge and the dam body overturning instability into occurrence and non-occurrence;
dividing the state of the rolling layer shearing damage into serious state and non-occurrence state or non-serious state;
dividing states of over-standard uplift pressure of the dam foundation and seepage damage around the dam into serious states and non-occurrence states or serious states;
dividing the states of the dam overtopping, dam structure damage and dam seepage damage into non-occurrence and non-occurrence;
dividing the state of the dam crash includes not occurring and not occurring.
4. The dam risk assessment method based on the bayesian network as claimed in claim 1, further comprising a step of determining major risk factors threatening the safety of the dam, specifically:
according to the formula
Figure FDA0002842713860000021
Calculating the importance of risk factors in each node;
wherein, the aiRepresenting the importance of the risk factor represented by the ith node, p0Representing the probability of occurrence, p, of the risk factor represented by the parent nodeiIndicating the parent when the risk factor represented by the ith node does not occurProbability of occurrence of risk factors represented by the nodes;
and (4) carrying out ranking according to the importance of the risk factors in each node to obtain the main risk factors threatening the safety of the dam.
5. A bayesian network based dam risk assessment device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program performs the steps of:
selecting a plurality of main risk factors as nodes of the Bayesian network according to the actual situation of the dam, and dividing the states of the nodes according to the occurrence characteristics of the risk factors;
determining a Bayesian network structure, and dividing each node into an input node, an intermediate node and an output node according to the relevance among the risk factors, wherein the input node is a risk source influencing the safety of the dam, the output node is an object of risk evaluation of the dam, and the intermediate node is an object of risk transmission to the output node under the influence of the input node;
probability assignment, namely assigning the prior probability of the input node according to the actual condition, and assigning the conditional probability of the input node and the intermediate node and the conditional probability of the intermediate node and the output node according to the association degree between the nodes;
and risk transfer, namely performing risk transfer through an input node and an output node in the Bayesian network, calculating the state and the corresponding probability of the output node, and performing risk evaluation of the dam.
6. The Bayesian network-based dam risk assessment device according to claim 5, wherein:
the risk sources influencing the safety of the dam comprise overproof flood, gate faults, unstable surge of the reservoir bank, shearing damage of a rolling layer, overturning instability of a dam body, overproof uplift pressure of a dam foundation and seepage damage around the dam;
the dam risk assessment object is dam accident;
and carrying out dam overtopping, dam structure damage and dam penetration damage on the objects for risk transfer on the output node under the influence of the input node.
7. The Bayesian network-based dam risk assessment device according to claim 6, wherein the step of dividing the state of each node according to the occurrence characteristics of each risk factor specifically comprises:
dividing the states of the overproof flood, the gate fault, the bank instability and the surge and the dam body overturning instability into occurrence and non-occurrence;
dividing the state of the rolling layer shearing damage into serious state and non-occurrence state or non-serious state;
dividing states of over-standard uplift pressure of the dam foundation and seepage damage around the dam into serious states and non-occurrence states or serious states;
dividing the states of the dam overtopping, dam structure damage and dam seepage damage into non-occurrence and non-occurrence;
dividing the state of the dam crash includes not occurring and not occurring.
8. The Bayesian network-based dam risk assessment device according to claim 5, further comprising a step of determining major risk factors threatening dam safety, specifically:
according to the formula
Figure FDA0002842713860000041
Calculating the importance of risk factors in each node;
wherein, the aiRepresenting the importance of the risk factor represented by the ith node, p0Representing the probability of occurrence, p, of the risk factor represented by the parent nodeiRepresenting the occurrence probability of the risk factor represented by the parent node when the risk factor represented by the ith node does not occur;
and (4) carrying out ranking according to the importance of the risk factors in each node to obtain the main risk factors threatening the safety of the dam.
CN202011497870.6A 2020-12-17 2020-12-17 Dam risk assessment method and equipment based on Bayesian network Active CN112464576B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011497870.6A CN112464576B (en) 2020-12-17 2020-12-17 Dam risk assessment method and equipment based on Bayesian network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011497870.6A CN112464576B (en) 2020-12-17 2020-12-17 Dam risk assessment method and equipment based on Bayesian network

Publications (2)

Publication Number Publication Date
CN112464576A true CN112464576A (en) 2021-03-09
CN112464576B CN112464576B (en) 2023-01-24

Family

ID=74803049

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011497870.6A Active CN112464576B (en) 2020-12-17 2020-12-17 Dam risk assessment method and equipment based on Bayesian network

Country Status (1)

Country Link
CN (1) CN112464576B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113657599A (en) * 2021-08-20 2021-11-16 北京航空航天大学 Accident cause and effect reasoning method and device, electronic equipment and readable storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20150129487A (en) * 2014-05-12 2015-11-20 한국건설기술연구원 method of evaluating extent of damage of levee and system for the same
CN107491898A (en) * 2017-09-01 2017-12-19 中国电建集团成都勘测设计研究院有限公司 Bayesian network model and construction method for step power station risk analysis
CN107808263A (en) * 2017-12-01 2018-03-16 安徽中科超安科技有限公司 Nuclear power station multi-risk System integrated estimation system and method based on information fusion
CN108615114A (en) * 2018-04-28 2018-10-02 中国电建集团昆明勘测设计研究院有限公司 A kind of reservoir dam safety risk estimating method based on bowknot model

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20150129487A (en) * 2014-05-12 2015-11-20 한국건설기술연구원 method of evaluating extent of damage of levee and system for the same
CN107491898A (en) * 2017-09-01 2017-12-19 中国电建集团成都勘测设计研究院有限公司 Bayesian network model and construction method for step power station risk analysis
CN107808263A (en) * 2017-12-01 2018-03-16 安徽中科超安科技有限公司 Nuclear power station multi-risk System integrated estimation system and method based on information fusion
CN108615114A (en) * 2018-04-28 2018-10-02 中国电建集团昆明勘测设计研究院有限公司 A kind of reservoir dam safety risk estimating method based on bowknot model

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
徐强: "大坝的风险分析方法研究", 《中国优秀硕士学位论文全文数据库》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113657599A (en) * 2021-08-20 2021-11-16 北京航空航天大学 Accident cause and effect reasoning method and device, electronic equipment and readable storage medium
CN113657599B (en) * 2021-08-20 2024-05-28 北京航空航天大学 Accident cause and effect reasoning method, device, electronic equipment and readable storage medium

Also Published As

Publication number Publication date
CN112464576B (en) 2023-01-24

Similar Documents

Publication Publication Date Title
CN112003277B (en) Transmission and distribution cooperative load recovery optimization control method and system
CN112001626A (en) Method for evaluating toughness of power distribution network in typhoon weather, storage medium and equipment
CN107292502B (en) Power distribution network reliability assessment method
CN103077447A (en) Subway construction safety management decision-making method based on Bayesian network
CN103116804A (en) Fuzzy neural network model and intelligent prediction method for deep excavation deformation
CN112464575A (en) Dam group risk assessment method and equipment based on Bayesian network
CN112464576B (en) Dam risk assessment method and equipment based on Bayesian network
Wang et al. Practical reliability analysis and design by Monte Carlo Simulation in spreadsheet
CN116307709B (en) Comprehensive assessment method and system for flood control capacity of transformer substation based on information gain fusion
CN112540159A (en) Nuclear power plant atmospheric diffusion prediction correction method, system, medium and electronic equipment
CN116911594A (en) Method and device for evaluating leakage emergency repair risk of gas pipeline
CN110021933B (en) Power information system control function reliability assessment method considering component faults
King Using a systems approach to analyze the operational safety of dams
CN116452041A (en) Power distribution network reliability assessment method based on cross entropy important sampling
CN110336280B (en) Power system cascading failure analysis method based on dictionary set acceleration
CN112751334A (en) Power grid online modeling method and system based on memory computing architecture
CN114548739A (en) Transformer substation flood prevention risk combination evaluation method
Ma et al. A reliability allocation method based on Bayesian networks and Analytic Hierarchy Process
Zavisca et al. A bayesian network approach to accident management and estimation of source terms for emergency planning
CN103197564A (en) Digitalized human-machine interface monitoring unit number optimization method and system
CN112862211A (en) Method and device for assigning orders of dynamic ring defects of communication management system
Zhu et al. Fault analysis of wind turbine gearbox based on vague set and fault tree
CN117273465B (en) Risk transfer method for cascade hydropower junction group
Li et al. Research on wind turbine system reliability modeling and preventive maintenance policy considering performance degradation and shock
Fan et al. Risk Analysis of Discrete Dynamic Event Tree Based on Dynamic Bayesian Network

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