A water diversion tunnel safety assessment method, system, device and medium

Abstract

The present invention discloses a water diversion tunnel safety assessment method, system, device and medium. First, the prior probability of the Bayesian network is obtained by combining inspection data and model simulation; then, based on fuzzy set theory, the system's fault tree is converted into a corresponding Bayesian network model, and a conditional probability table of different node characteristics is obtained; finally, based on the Bayesian network's bidirectional reasoning calculation and sensitivity analysis capabilities, the probability of safety risk occurrence is calculated and key risk factors are identified, while accident cause diagnosis is performed. The advantages of the present invention are: by analyzing historical data, the probability of each node in the Bayesian network is accurately obtained, and using fuzzy set theory, a logical relationship between the failure probability of network events and the different characteristics between event nodes is constructed, effectively improving the accuracy of water diversion tunnel safety assessment.

Description

Technical Field

[0001] The present invention relates to the technical field of water diversion tunnel safety evaluation, and in particular to a water diversion tunnel safety evaluation method, system, device and medium based on fuzzy Bayesian network. Background Art

[0002] A diversion tunnel is a hydraulic tunnel that diverts water from a water source. Currently, a large number of diversion tunnels are involved in the construction of hydropower stations and other projects. Due to the complex geological environment, their safety status is highly complex and uncertain. However, due to the limitations of current technical measurement methods, the safety status of diversion tunnels cannot be directly explored and analyzed. [1] .

[0003] Fault tree analysis, as a common method for analyzing system safety status, is unable to evaluate the safety status of polymorphic and uncertain event characteristics. However, Bayesian networks (BN) based on probability theory can more clearly describe the polymorphism and logical relationships of events, have better expression and analysis capabilities for uncertain event characteristics, and can also perform network probability updates and reverse reasoning to a certain extent. Bayesian networks have been used in dam reliability analysis. [2] , Tunnel Engineering Analysis [3] , foundation pit risk assessment [4,5] Although the Bayesian network method has many advantages mentioned above, it is still limited by the traditional probability method in the uncertainty reasoning based on accurate probability. [6-8] .

[0004] In fact, due to the lack of historical data and other reasons, the node probability of the Bayesian network is usually not accurately obtained, and the failure probability of network events and the logical relationship between different characteristics of event nodes are also fuzzy, resulting in the existing technology having defects such as poor accuracy and low efficiency in evaluating the safety of water diversion tunnels. [9,10] .

[0005] References:

[0006] [1] Li Dongsheng. Research on condition monitoring and fault diagnosis of stacker crane[J]. Journal of Science and Technology Innovation, 2020, 17(01): 66-67. DOI: 10.16660 / j.cnki.1674-098X.2020.01.066;

[0007] [2] Tian Shirun. Reliability analysis of ZPW-2000A track circuit system based on Bayesian network[J]. Railway Communication and Signal Engineering Technology, 2020, 17(S1): 106-109;

[0008] [3] Liu Guangyu, Zhang Chunyou. Improved Bayesian network hydraulic system reliability analysis based on evidence theory[J]. Machine Tools and Hydraulics, 2019, 47(24): 17-23+46;

[0009] [4] Wang Chengtang, Wang Hao, Qin Weimin, et al. Evaluation of the possibility of collapse of deep foundation pit of subway station based on polymorphic fuzzy Bayesian network[J]. Rock and Soil Mechanics, 2020, 41(05): 1670-1679+1689;

[0010] [5] Yue Jianwei, Zhong Haolei, Gu Lihua, et al. Application of Bayesian regularized neural network in deep foundation pit deformation prediction[J]. Journal of Henan University, 2022, 52(02): 200-209. DOI: 10.15991 / j.cnki.411100.2022.02.004;

[0011] [6] Bai Xu, Tang Rongkeng, Luo Xiaofang, et al. Multi-state reliability analysis of semi-submersible drilling platform system based on fault tree analysis and Bayesian network method [J]. China Shipbuilding, 2020, 61(02): 220-228;

[0012] [7] Hua Ling, Tang Tao, Qing Linbo, et al. DVC-HEVC fast transcoding based on naive Bayesian classification [J]. Journal of Terahertz Science and Electronic Information, 2020, 18(02): 235-240;

[0013] [8] Han Fengxia, Wang Hongjun, Qiu Cheng. Reliability evaluation of production line system based on fuzzy Bayesian network[J]. Manufacturing Technology and Machine Tools, 2020(09):45-49.DOI:10.19287 / j.cnki.1005-2402.2020.09.008;

[0014] [9] Chen Dongning, Yao Chengyu. Reliability analysis of multi-state systems based on fuzzy Bayesian networks and its application in hydraulic systems [J]. Journal of Mechanical Engineering, 2012, 48(16): 175-183;

[0015]

[10] Wang Chengtang, Wang Hao, Qin Weimin, et al. Evaluation of the possibility of collapse of deep foundation pit of subway station based on polymorphic fuzzy Bayesian network [J]. Rock and Soil Mechanics, 2020, 41(5): 1670-1679,1689. Summary of the Invention

[0016] Aiming at the complexity and uncertainty of the existing water diversion tunnel safety evaluation, the present invention provides a water diversion tunnel safety evaluation method, system, device and medium.

[0017] In order to achieve the above object of the invention, the technical solution adopted by the present invention is as follows:

[0018] A water diversion tunnel safety assessment method includes: first, combining inspection data and model simulation to obtain a priori probabilities of a Bayesian network; then, based on fuzzy set theory, converting the system's fault tree into a corresponding Bayesian network model and obtaining a conditional probability table for different node characteristics; finally, based on the Bayesian network's bidirectional reasoning and sensitivity analysis capabilities, calculating the probability of safety risk occurrence and identifying key risk factors, while also performing accident cause diagnosis.

[0019] Furthermore, the specific steps of the water diversion tunnel safety assessment method are as follows:

[0020] S1: On-site inspection to collect data on wall safety hazards;

[0021] S2: Establish Bayesian network structure;

[0022] Taking "tunnel wall safety hazard warning" as the top event, a safety evaluation model fault tree is established, and then it is converted into a Bayesian network model according to the Bayesian network construction process based on the fault tree.

[0023] S3: calculation of root node failure probability;

[0024] To collect data on wall safety hazards, we first used fuzzy feature representation, using semantic descriptions of {none, mild, severe} and corresponding values ​​of {0, 0.5, 1}. We then calculated the membership of each root node based on the likelihood of occurrence based on on-site inspection data.

[0025] S4: Conditional probability calculation;

[0026] After obtaining the root node failure probability, the conditional probabilities of the corresponding intermediate nodes and even leaf nodes are calculated and directly converted from the fault tree.

[0027] Once the conditional probabilities of the intermediate nodes and leaf nodes are obtained, the posterior probability of the leaf node T can be calculated. The fuzzy fault probabilities of each root node are then calculated using the on-site inspection data. The fuzzy likelihood of various fault states is then obtained using the conditional probability table.

[0028] Furthermore, the event names in the Bayesian network model of S2 are shown in Table 1:

[0029] Table 1

[0030]

[0031] The leaf node T represents the safety hazard of the tunnel wall, the intermediate nodes are: M1~M4, and the root nodes are: X1~X8.

[0032] Furthermore, the specific process for constructing a Bayesian network based on a fault tree in S2 is as follows: the bottom event of the fault tree can be mapped to the root node of the fuzzy Bayesian network, and the intermediate and top events can be mapped to intermediate nodes and leaf nodes, respectively. The input-output relationship between events in the fault tree can be mapped to the logical relationship between the root node and leaf nodes in the fuzzy Bayesian network; the probability of the bottom event of the fault tree corresponds to the failure probability of the root node in the fuzzy Bayesian network; and the discriminant of the logic gate in the fault tree can be expressed using the conditional probability table in the fuzzy Bayesian network.

[0033] Furthermore, in S3, the membership degree of each root node is calculated based on its occurrence probability, as follows:

[0034] in, Represents node x i a i Fault conditions:

[0035]

[0036]

[0037]

[0038] in, Represents node x i a i Fault status.

[0039] Furthermore, the fuzzy probability of various fault states is obtained in S4, and the formula is as follows:

[0040]

[0041] Among them, λ(T) and λ(y j ) are leaf nodes T and intermediate nodes y respectively j The parent node set of is the root node x i In state probability of failure.

[0042] The present invention also discloses a water diversion tunnel safety evaluation system, comprising: a wall safety hazard data module, a fuzzy Bayesian network module, a root node failure probability calculation module and a conditional probability calculation module;

[0043] Wall safety hazard data module: used to collect wall safety hazard data and used for fuzzy Bayesian network.

[0044] Fuzzy Bayesian network module: Predicts the occurrence probability of each root node based on collected environmental information and on-site inspection data. Then, based on the forward inference algorithm of the Bayesian network, it predicts the probability of occurrence of safety hazards in the tunnel wall. It calculates the critical importance of each root node to identify key risk factors.

[0045] Root Node Failure Probability Calculation Module: Data on wall safety hazards is collected and fuzzy representation is performed, using semantic descriptions of {none, mild, severe} and corresponding values ​​of {0, 0.5, 1}. The probability of occurrence of each root node is then calculated based on on-site inspection data.

[0046] Conditional probability calculation module: After obtaining the root node fault probability, the conditional probability of the corresponding intermediate nodes and even leaf nodes is calculated and directly converted from the fault tree.

[0047] The present invention also discloses a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the above-mentioned water diversion tunnel safety assessment method is implemented.

[0048] The present invention also discloses a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned water diversion tunnel safety evaluation method is implemented.

[0049] Compared with the prior art, the advantages of the present invention are:

[0050] (1) Aiming at the uncertainty of various event states in the process of water diversion tunnel safety evaluation, the present invention adopts fuzzy set theory and combines it with Bayesian network to propose a water diversion tunnel safety evaluation model analysis method based on fuzzy Bayesian network.

[0051] (2) The conditional probability table of the Bayesian network is used to describe the logical relationship between different events, which solves the problem of fuzziness of event states and uncertainty of logical relationships between different states. It also broadens the scope of application of the Bayesian reliability analysis method.

[0052] (3) Combined with actual inspection data, the Bayesian network was used to process fuzzy and uncertain information, and an example analysis of the water diversion tunnel safety evaluation model was carried out to verify the feasibility of this method.

[0053] (4) By analyzing historical data, the probability of each node in the Bayesian network is accurately obtained. Using fuzzy set theory, the logical relationship between the failure probability of network events and the different characteristics between event nodes is constructed, effectively improving the accuracy of water diversion tunnel safety evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1Conditional probability plots for two-state Bayes;

[0055] Figure 2 Conditional probability graphs for fuzzy multi-state Bayes;

[0056] Figure 3 This is a safety evaluation flow chart based on fuzzy Bayesian network according to an embodiment of the present invention;

[0057] Figure 4 This is a schematic diagram of the structure of the fuzzy function for fault events according to an embodiment of the present invention;

[0058] Figure 5 Schematic diagram of the Bayesian network structure of the water diversion tunnel wall safety evaluation model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and examples.

[0060] 1. Introduction to the Fuzzy Bayesian Network Based on This Example

[0061] 1. Subjective Bayes

[0062] The basic rule in subjective Bayesian reasoning is: IF A THEN B(LS,LN). LS represents the sufficient condition for the rule to be true, and LN represents the necessary condition for the rule to be true. The calculation formulas for LS and LN are:

[0063]

[0064] Where P represents the probability function and O represents the probability function. The relationship between probability and probability is: O = P / (1-P).

[0065] In formula (1), when A is true, the effect on B is expressed as LS. If LS becomes larger, then Ο(B| A ) becomes larger, making P(B|A) larger, which also shows that B relies more on A. When A is false, the impact on B is expressed as LN. If LN=0, then This also shows When is true, B must be false. However, in practical systems, LS and LN are usually difficult to obtain directly.

[0066] When condition A is known, the prior probability of condition B is P(B). According to formula (2), the conditional probability of B P(B|A) and They are:

[0067]

[0068] Since the conclusion of a rule usually has multiple different sufficient conditions, the following equation holds:

[0069]

[0070] Finally, based on the conversion relationship between probability and probability, we can get the node probability of B under multiple sufficient conditions.

[0071] 2. Two-state Bayesian evaluation

[0072] Normally, the two states of a two-state system (fault or normal) can be represented by two numerical values, 0 and 1. In fault tree analysis, traditional logic gates can directly express the conditional probabilities in Bayesian networks. However, this method can only express fixed logical relationships between components. When the logical relationships between components are uncertain, the conditional probabilities of the child nodes need to be changed accordingly. Figure 1 As shown, the conditional probability table of the two-state Bayes can be expressed as:

[0073] exist Figure 1 In the example, when the fault state of node x1 is 0 and the fault state of node x2 is 0, the fault probability of node y when the state is 1 is 0.1, which can be expressed as P(y=1|x1=0,x2=0)=0.1.

[0074] 3. Fuzzy Polymorphic Bayesian Evaluation

[0075] Bayesian networks can be used to handle the polymorphism of nodes, using different conditional probabilities for different variables. By modifying the conditional probability table of the corresponding node, the polymorphism of the variable can be represented. Figure 2 As shown, the conditional probability table of the fuzzy polymorphic Bayesian network can be expressed as:

[0076] exist Figure 1 and Figure 2 In this paper, the variable symbols in the conditional probability table can represent different fault states of the corresponding nodes and the probability of failure under these fault states. The fault logic relationships between the corresponding nodes can also be transformed into the fault logic relationships of the corresponding nodes in the fuzzy Bayesian network based on the conditional probability table. Furthermore, by modifying the conditional probability table between the corresponding nodes, it is also possible to represent the impact that changes in environmental conditions, etc., may have on system failures.

[0077] 2. Security evaluation based on fuzzy Bayesian network;

[0078] The embodiment of the present invention provides a water diversion tunnel safety assessment method, such as Figure 3 shown.

[0079] 1. Construction of fuzzy Bayesian network based on fault tree;

[0080] By mapping the different events in a fault tree, the corresponding structure of a fuzzy Bayesian network can be constructed. The bottom event of the fault tree can be mapped to the root node of the fuzzy Bayesian network, while the intermediate and top events can be mapped to intermediate and leaf nodes, respectively. The input-output relationship between events in the fault tree can be mapped to the logical relationship between the root and leaf nodes in the fuzzy Bayesian network. The probability of the bottom event of the fault tree corresponds to the failure probability of the root node in the fuzzy Bayesian network. The discriminants of the logic gates in the fault tree can be expressed using the conditional probability table in the fuzzy Bayesian network. In fact, the modeling accuracy of the fuzzy Bayesian network depends on the accurate solution of the root node failure probability.

[0081] 2. Fuzzification of node polymorphism fault events;

[0082] For a two-state system, the state of the system or component can usually be described as normal and faulty. However, in practice, due to the fuzziness of system variables, the system or component will usually be in multiple failure modes and different failure states. Therefore, as time goes by, the system state will also show polymorphic characteristics and have a certain degree of fuzziness. By using linguistic variables to describe the fault states of different nodes, the node polymorphic fault events can be fuzzified. Use the language value set {none, mild, severe} to describe the three fault states of the system nodes, and use fuzzy numbers 0, 0.5 and 1 to replace the corresponding language values ​​in turn, and calculate the conditional probability. Determine the node failure probability by constructing the corresponding fault event fuzzification function. The constructed fault event fuzzification function is as follows: Figure 4 shown.

[0083] Depend on Figure 4 It can be seen that:

[0084]

[0085]

[0086]

[0087] When the fault state of a node is 0.3, it can be obtained from equations (4) to (6) that the membership of the node determined as a fault-free state is 1 / 3, the membership of the node determined as a mild fault state is 2 / 3, and the membership of the node determined as a severe fault state is 0, and μ0(0.3)+μ 0.5 (0.3)+μ1(0.3)=1.

[0088] 3. Determination of conditional probability table;

[0089] The fault probability state of the root node can be obtained by the above method, while the conditional probabilities of the intermediate nodes and leaf nodes are usually obtained by direct conversion from the fault tree.

[0090] In fact, by mapping the logical gate relationships between different events in a fault tree, a conditional probability table for a traditional Bayesian network can be directly obtained. However, the shortcomings of traditional fault trees limit the Bayesian network's ability to describe event states and the connections between them. To better integrate practical experience and expert knowledge, a fuzzy multi-state Bayesian network based on fuzzy set theory is used to describe the polymorphism between events and the uncertainty of their connections. This allows the generation of a conditional probability table that represents the fault states of each child node caused by various fault states of a multi-state root node.

[0091] 4. Forward reasoning failure probability;

[0092] If the failure probability of each state of the root node is known, then forward reasoning can be performed based on the joint probability distribution of BN to obtain the failure state of the leaf node T as T q The probability of occurrence of (q=1,2,...,t) is as follows:

[0093]

[0094] Among them, λ(T) and λ(y j ) are leaf nodes T and intermediate nodes y respectively j The parent node set of is the root node x i In state probability of failure.

[0095] 5. Calculation of root node importance;

[0096] When the root node x i When the failure probability of i=1,2,...,n changes, the failure probability of the leaf node will also change accordingly. At this time, the importance of the root node to the leaf node can be expressed as:

[0097]

[0098] Where M i Represents x i The number of states, P(T=T q ) indicates that the leaf node T has state T q The probability of Represents x i In state The probability of is x i The probability importance of It can be understood that when the root node x i The status is When the state of leaf node T is T qThe greater the critical importance of the corresponding root node to the leaf node, the greater the possibility that the root node will trigger an unsafe state in the leaf node. When a leaf node is in an unsafe state, the root node should be given priority.

[0099] 3. Example Analysis

[0100] 1. Establish Bayesian network structure

[0101] This embodiment takes the water diversion tunnel wall safety hazard warning as an example. After field research, it was found that wall safety hazards usually manifest themselves in three aspects: an increase in wall cracks, crack expansion, and an increase in sediment in the water diversion tunnel. In terms of the increase in wall cracks, it is mainly manifested in a large number of algae plants attached to the wall and new cracks in the tunnel; in terms of crack expansion, it is mainly manifested in the extension of crack length, increase in crack width and depth, and peeling of the wall at the crack; in terms of the increase in sediment in the water diversion tunnel, it is mainly manifested in the accumulation of mud and sand at the bottom, tree stumps, etc. inside the tunnel. After analysis, the "tunnel wall safety hazard warning" is used as the top event to establish a safety evaluation model fault tree, and then according to the Bayesian network construction process based on the fault tree, it is converted into the following: Figure 5 The Bayesian network model shown in Figure 5 As shown in the figure, leaf node T represents a tunnel wall safety hazard. The event names corresponding to intermediate nodes M1–M4 and root nodes X1–X8 are shown in Table 1. The occurrence probability of each root node is predicted based on collected environmental data and on-site inspection data. The probability of tunnel wall safety hazard occurrence is then predicted using a forward inference algorithm using a Bayesian network. The critical importance of each root node is calculated to identify key risk factors, thereby better analyzing the water diversion tunnel safety assessment model.

[0102] Table 1 Event codes, names and unsafe status scores

[0103]

[0104] Table 2 Fault status membership of each node based on on-site inspection data

[0105]

[0106]

[0107] 2. Calculation of root node failure probability

[0108] In the field environment, for faults such as new tunnel cracks, fuzzy feature representation is first performed, using semantic descriptions of {none, mild, severe} and corresponding values ​​of {0, 0.5, 1}. Based on the field inspection data, the membership degree of each root node is calculated for the probability of occurrence. As shown in Table 2.

[0109] 3. Conditional Probability Calculation

[0110] Once the root node failure probability is obtained, the conditional probabilities of the corresponding intermediate nodes and even leaf nodes can be calculated. However, due to the polymorphism of the root node state, the conditional probabilities of the intermediate nodes and leaf nodes cannot be directly calculated. Figure 5 As shown, the relationship between M1 and x1 and x2 can be represented by an OR logic relationship. When x1 or x2 is "critical," M1 is always in "critical" status; when both x1 and x2 are "none," M1 is in "none." The conditional probabilities of intermediate nodes M1, M2, M3, and M4 are shown in Tables 3, 4, 5, and 6.

[0111] Table 3 M1 conditional probability table

[0112]

[0113] Table 4 M2 conditional probability table

[0114]

[0115] Table 5 M3 conditional probability table

[0116]

[0117]

[0118] Table 6 M4 conditional probability table

[0119]

[0120] Table 7 Leaf node T conditional probability table

[0121]

[0122]

[0123] 4. Posterior Probability Calculation

[0124] After obtaining the conditional probabilities of the intermediate nodes and leaf nodes, the posterior probability of the leaf node T can be calculated. Table 7 shows the conditional probabilities of the leaf node T. Then, the fuzzy fault probabilities of each root node are calculated using the field inspection data. Using the conditional probability table and Equation (7), the fuzzy probabilities of various fault states are obtained as follows: P(T=0)=0.79, P(T=0.5)=0.12, and P(T=1)=0.09.

[0125] 5. Root node key importance calculation

[0126] According to formula (8), the critical importance of each root node fault state with respect to leaf node T can be obtained. Figure 5 As shown in the figure, x4 has the highest critical importance, followed by x1, x3, x7, and x8. They can be sorted according to the critical importance of the root nodes to determine the order of water diversion tunnel safety evaluation.

[0127] In one embodiment of the present invention, a water diversion tunnel safety assessment system is provided. The system can be used to implement the above-mentioned water diversion tunnel safety assessment method. Specifically, the system includes: a wall safety hazard data module, a fuzzy Bayesian network module, a root node failure probability calculation module, and a conditional probability calculation module.

[0128] Wall safety hazard data module: used to collect wall safety hazard data and used for fuzzy Bayesian network.

[0129] Fuzzy Bayesian network module: Predicts the occurrence probability of each root node based on collected environmental information and on-site inspection data. Then, based on the forward inference algorithm of the Bayesian network, it predicts the probability of occurrence of safety hazards in the tunnel wall. It calculates the critical importance of each root node to identify key risk factors.

[0130] Calculating root node failure probability: Data on wall safety hazards is collected and fuzzy representation is performed, using semantic descriptions of {none, mild, severe} and corresponding values ​​of {0, 0.5, 1}. The probability of occurrence of each root node is then calculated based on on-site inspection data.

[0131] Conditional probability calculation module: After obtaining the root node fault probability, the conditional probability of the corresponding intermediate nodes and even leaf nodes is calculated and directly converted from the fault tree.

[0132] In another embodiment of the present invention, a terminal device is provided, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the water diversion tunnel safety assessment method, including the following steps:

[0133] First, the prior probability of the Bayesian network is obtained by combining inspection data and model simulation. Then, based on fuzzy set theory, the system's fault tree is converted into a corresponding Bayesian network model, and a conditional probability table of different node characteristics is obtained. Finally, based on the Bayesian network's bidirectional reasoning and sensitivity analysis capabilities, the probability of safety risk occurrence is calculated and key risk factors are identified, while accident causes are diagnosed.

[0134] In another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory.

[0135] The processor may load and execute one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the water diversion tunnel safety assessment system method in the above embodiment. The processor may load and execute the following steps:

[0136] First, the prior probability of the Bayesian network is obtained by combining inspection data and model simulation. Then, based on fuzzy set theory, the system's fault tree is converted into a corresponding Bayesian network model, and a conditional probability table of different node characteristics is obtained. Finally, based on the Bayesian network's bidirectional reasoning and sensitivity analysis capabilities, the probability of safety risk occurrence is calculated and key risk factors are identified, while accident causes are diagnosed.

[0137] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0138] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0139] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0141] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the implementation methods of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific descriptions and embodiments. Those skilled in the art can make various other specific variations and combinations based on the technical teachings disclosed in the present invention without departing from the essence of the present invention, and such variations and combinations are still within the scope of protection of the present invention.

Claims

1. A water diversion tunnel safety assessment method, characterized in that: include: First, the inspection data and model simulation are combined to obtain the prior probability of the Bayesian network; Then, based on fuzzy set theory, the system's fault tree is converted into a corresponding Bayesian network model, and a conditional probability table of different node characteristics is obtained. Finally, based on the Bayesian network's bidirectional reasoning and sensitivity analysis capabilities, the probability of safety risk occurrence is calculated, key risk factors are identified, and accident causes are diagnosed. The specific steps of the diversion tunnel safety assessment method are as follows: S1: On-site inspection to collect data on wall safety hazards; S2: Establish Bayesian network structure; Taking "tunnel wall safety hazard warning" as the top event, a safety evaluation model fault tree is established, and then converted into a Bayesian network model based on the Bayesian network construction process based on the fault tree; The Bayesian network construction process based on the fault tree is specifically as follows: the bottom event of the fault tree is mapped to the root node of the fuzzy Bayesian network, and the intermediate event and the top event are mapped to the intermediate node and the leaf node respectively; the input-output relationship between the events in the fault tree is mapped to the logical relationship between the root node and the leaf node in the fuzzy Bayesian network; the probability of the bottom event of the fault tree corresponds to the fault probability of the root node in the fuzzy Bayesian network; the logic gate discriminant in the fault tree is expressed using the conditional probability table in the fuzzy Bayesian network; S3: calculation of root node failure probability; To collect data on wall safety hazards, we first perform fuzzy feature representation, using semantic descriptions of {none, mild, severe}, with corresponding values ​​of {0, 0.5, 1}; And based on the on-site inspection data, the respective membership degree of each root node is calculated for its occurrence probability; S4: Conditional probability calculation; After obtaining the root node failure probability, the conditional probabilities of the corresponding intermediate nodes and even leaf nodes are calculated and directly converted from the fault tree; After obtaining the conditional probabilities of the intermediate nodes and leaf nodes, the posterior probability of the leaf node T is calculated; then, the fuzzy fault probability of each root node is calculated using the on-site inspection data, and then the fuzzy probability of various fault states is obtained through the conditional probability table.

2. A water diversion tunnel safety assessment method according to claim 1, characterized in that: The event names in the Bayesian network model described in S2 are shown in Table 1: Table 1 The leaf node T represents the safety hazard of the tunnel wall, the intermediate nodes are: M1~M4, and the root nodes are: X1~X8.

3. A water diversion tunnel safety assessment method according to claim 1, characterized in that: In S3, the membership degree of each root node is calculated based on its occurrence probability, as shown in the following formula: in, Represents node x i a i Fault conditions: in, Represents node x i a i Fault status.

4. A water diversion tunnel safety assessment method according to claim 1, characterized in that: In S4, the fuzzy probability of various fault states is obtained, and the formula is as follows: Among them, λ(T) and λ(y j ) are leaf nodes T and intermediate nodes y respectively j The parent node set of is the root node x i In state probability of failure.

5. A water diversion tunnel safety assessment system, characterized in that: The system can be used to implement the water diversion tunnel safety assessment method according to any one of claims 1 to 4, specifically comprising: a wall safety hazard data module, a fuzzy Bayesian network module, a root node failure probability calculation module, and a conditional probability calculation module; Wall safety hazard data module: used to collect wall safety hazard data and use it for fuzzy Bayesian network; Fuzzy Bayesian network module: This module predicts the probability of occurrence of each root node based on collected environmental information and on-site inspection data. It then uses the Bayesian network's forward inference algorithm to predict the probability of occurrence of tunnel wall safety hazards and calculates the critical importance of each root node to identify key risk factors. Root node failure probability calculation module: To collect data on wall safety hazards, fuzzy feature representation is first performed, using semantic descriptions of {none, mild, severe} and corresponding values ​​of {0, 0.5, 1}. Based on on-site inspection data, the membership degree of each root node is calculated for its probability of occurrence. Conditional probability calculation module: After obtaining the root node fault probability, the conditional probability of the corresponding intermediate nodes and even leaf nodes is calculated and directly converted from the fault tree.

6. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the water diversion tunnel safety assessment method according to one of claims 1 to 4 is implemented.

7. A computer-readable storage medium, characterized in that: A computer program is stored, and when the program is executed by a processor, the water diversion tunnel safety evaluation method according to one of claims 1 to 4 is implemented.

Images