Method and device for risk assessment of urban lifeline pipe network

Abstract

The application relates to a risk assessment method and device for an urban lifeline pipe network, and relates to the field of urban pipe network system safety monitoring; the method can accurately quantify pipe network risks and provide data support for urban risk decision-making. The method comprises the following steps: acquiring attribute parameters of each pipe section in a pipe network system and soil body attributes, and constructing a finite element model of the pipe network system; acquiring seismic motion records of a region where the pipe network system is located, performing dynamic analysis based on amplitude-modulated acceleration time histories, and obtaining stress data of the pipe sections under different seismic motion peak accelerations; determining the damage levels of the pipe sections through the stress data, and counting the number of times when the damage levels reach preset levels to obtain failure probabilities of the pipe sections; training a machine learning agent model through the seismic motion records and the failure probabilities, and determining the failure probabilities of to-be-measured pipe sections through the model; determining importance indexes of the to-be-measured pipe sections through a weighted graph theory model; and determining risk assessment results of the to-be-measured pipe sections based on the failure probabilities and the importance indexes.

Description

Technical Field

[0001] This application relates to the field of safety monitoring of urban pipeline systems, and in particular to a risk assessment method and apparatus for urban lifeline pipeline networks. Background Technology

[0002] Urban lifeline networks include infrastructure systems such as gas, water supply, and drainage that ensure the safe operation of cities. Earthquakes, geological disasters, and extreme weather events pose serious threats to these networks. Therefore, conducting scientific and accurate risk assessments to shift from reactive response to proactive prevention is a crucial foundation for improving urban safety resilience.

[0003] Current pipeline risk assessment methods often rely on expert experience or limited data, failing to fully consider multi-source heterogeneous information such as pipeline structural attributes, real-time monitoring data, geospatial information, and external environmental loads. This makes it impossible to comprehensively reflect the true risks of the system. Furthermore, the lack of quantitative analysis on the importance of pipeline topology makes it difficult to identify critical pipeline sections that are essential for system functionality and whose failure would cause significant damage to the overall system function. In terms of assessment models, mainstream methods are mostly static or quasi-static, failing to reflect the risk evolution of the pipeline throughout its entire life cycle. This results in a lack of foresight in risk decision-making and makes it difficult to support rapid assessment and resource allocation in emergency scenarios such as after earthquakes. Summary of the Invention

[0004] This application provides a risk assessment method and apparatus for urban lifeline pipeline networks, which can predict the damage level and failure probability of the pipeline network system under different seismic intensities. It integrates the actual structural response into the risk assessment system, ensuring the physical rationality and interpretability of the prediction results, and provides a core tool for the rapid risk quantification of large-scale pipeline network systems.

[0005] Firstly, this application provides a risk assessment method for urban lifeline pipeline networks, including: Obtain the attribute parameters of each pipe segment and the soil properties in the pipeline network system, and construct the finite element model of the pipeline network system based on the attribute parameters of the pipe segments and the soil properties. Seismic ground motion records of the area where the pipeline network system is located are obtained, and the seismic ground motion records are amplitude-modulated to obtain acceleration time histories under different peak ground accelerations. Based on the amplitude-modulated acceleration time histories, dynamic analysis is performed to obtain stress data of the pipe section under different peak ground accelerations. The failure level of the pipe section is determined by stress data, and the failure probability of the pipe section is obtained by counting the number of times the failure level of the pipe section reaches the preset level under the same ground motion record. A machine learning proxy model is obtained by training the ground motion records and the corresponding failure probabilities of the ground motion records, and the failure probability of the pipe section to be tested is determined by the machine learning proxy model. Determine the weighted graph theory model of the pipeline network system, and use the weighted graph theory model to determine the importance index of the pipe segment to be tested; The risk assessment result of the pipe section under test is determined based on the failure probability and importance index of the pipe section under test.

[0006] In the method provided in this embodiment, the response relationship between peak ground acceleration (PGA) and pipe segment damage level and failure probability is determined using multi-source information such as pipe segment attributes and soil. A machine learning surrogate model is then constructed to learn this response relationship. The trained model is used to assess the failure probability of the pipe segment, enabling efficient and accurate prediction of the risk situation of urban lifeline pipelines under the influence of seismic motion, thus improving the accuracy of risk quantification. Furthermore, this surrogate model allows for dynamic risk perception and prediction, facilitating proactive risk prevention and control, and addressing the issues of passive response and delayed decision-making in urban lifeline pipelines. In addition, this solution also considers the topological relationship of the pipeline system, using a weighted graph theory model to quantify the topological importance of pipe segments within the system. This integrates multi-source data and pipeline topology to quantify risk, further improving the accuracy of risk quantification and providing data support for risk decision-making.

[0007] Secondly, this application provides a risk assessment device for urban lifeline pipeline networks, comprising: The finite element model construction module is used to obtain the attribute parameters of each pipe segment and the soil properties in the pipeline network system, and to construct the finite element model of the pipeline network system based on the attribute parameters of the pipe segments and the soil properties. The stress calculation module is used to acquire ground motion records of the area where the pipeline system is located, perform amplitude modulation on the ground motion records to obtain acceleration time histories under different peak ground accelerations, and perform dynamic analysis based on the amplitude-modulated acceleration time histories to obtain stress data of the pipe section under different peak ground accelerations. The failure determination module is used to determine the damage level of the pipe section through stress data and count the number of times the damage level of the pipe section reaches the preset level under the same ground motion record to obtain the failure probability of the pipe section. The seismic motion risk prediction module is used to train a machine learning proxy model through the seismic motion records and the failure probabilities corresponding to the seismic motion records, and to determine the failure probability of the pipe section to be tested through the machine learning proxy model. The topology processing module is used to determine the weighted graph theory model of the pipeline network system, and to determine the importance index of the pipe segment to be tested through the weighted graph theory model; The assessment result determination module is used to determine the risk assessment result of the pipe section under test based on the failure probability and importance index of the pipe section under test.

[0008] Thirdly, this application provides an electronic device including a memory and one or more processors. The memory stores one or more computer programs, each including instructions that, when executed by the processor, cause the electronic device to perform a risk assessment method for urban lifeline networks as described in the first aspect.

[0009] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on an electronic device, cause the electronic device to perform the risk assessment method for urban lifeline networks as described in the first aspect.

[0010] Fifthly, this application provides a computer program product that, when run on an electronic device, causes the electronic device to perform the risk assessment method for urban lifeline pipeline networks as described in the first aspect.

[0011] Understandably, the beneficial effects that the aforementioned risk assessment devices, electronic devices, computer-readable storage media, and computer program products for urban lifeline pipeline networks can achieve are similar to the beneficial effects described in the first aspect, and will not be repeated here. Attached Figure Description

[0012] Figure 1 A flowchart illustrating the risk assessment method for urban lifeline pipeline networks provided in this application embodiment; Figure 2 A schematic diagram of the structure of a risk assessment device for urban lifeline pipeline networks provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0013] To facilitate a clear description of the technical solutions in the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish identical or similar items with substantially the same function and effect. For example, "first chip" and "second chip" are only used to distinguish different chips and do not limit their order. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" do not necessarily imply that they are different. It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate that they are examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in this application should not be construed as being better or more advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner. In the embodiments of this application, "at least one" means one or more, and "more than one" means two or more.

[0014] It should be noted that "at the time of..." in the embodiments of this application can be either at the instant when a certain situation occurs, or for a period of time after the occurrence of a certain situation. The embodiments of this application do not make specific limitations on this.

[0015] The implementation of this embodiment will now be described in detail with reference to the accompanying drawings.

[0016] This embodiment provides a risk assessment method for urban lifeline pipeline networks. For example, this risk assessment method for urban lifeline pipeline networks can be applied to various electronic devices such as computers (PCs), virtual reality / augmented reality devices, wearable devices, and industrial computers; it can also be applied to servers, cloud computing, server clusters, etc. This embodiment does not impose any special limitations on it.

[0017] Figure 1 A flowchart illustrating the risk assessment method for urban lifeline pipeline networks provided in this application embodiment is shown.

[0018] like Figure 1 As shown, the risk assessment method for this city's lifeline pipeline network may include the following steps: Step 101: Obtain the attribute parameters of each pipe segment and the soil properties in the pipeline network system, and construct the finite element model of the pipeline network system based on the attribute parameters of the pipe segments and the soil properties.

[0019] The pipeline system includes urban gas pipelines, water supply pipelines, drainage pipelines, etc. This embodiment takes the gas pipeline as an example. For the gas pipeline network in urban areas, the pipe segment properties include pipe segment properties such as diameter, wall thickness, length, burial depth, and material constitutive properties; soil properties include soil dimensions, damping ratio, elastic modulus, expansion angle, friction angle, and cohesion; this embodiment is not limited to these. A refined finite element model of pipe and soil is constructed through the property parameters and soil properties. Specifically, for each pipe segment, including the pipe segment (shell element) and the soil (solid element), the pipe-soil interaction is defined as surface-to-surface contact, with hard contact in the normal direction and a penalty function used to define the friction coefficient in the tangential direction.

[0020] In an exemplary implementation, based on the geometric properties, material properties, pressure ratings, and other attribute parameters of each pipe segment in the pipeline network system, a stratified random sampling method is used to select several statistically representative pipe segments from the entire system. This ensures that the samples cover the key variation characteristics of the system, forming a sample library. A finite element model is then constructed based on the pipe segment sample library.

[0021] Step 102: Obtain ground motion records for the area where the pipeline system is located, perform amplitude modulation on the ground motion records to obtain acceleration time histories under different peak ground accelerations, and perform dynamic analysis based on the amplitude-modulated acceleration time histories to obtain stress data of the pipe section under different peak ground accelerations.

[0022] Based on the area where the pipeline network is located, seismic ground motion records for that area are acquired. The peak ground acceleration (PGA) is amplitude-adjusted to a series of target intensity levels to obtain acceleration time histories for different intensity levels. Infinite element boundaries are established at the bottom and sides of the soil to prevent boundary reflection waves from interfering with the results. The amplitude-adjusted acceleration time histories are applied as seismic ground motion inputs to all nodes at the bottom of the soil. Nonlinear time history solutions are obtained using the explicit dynamics analysis module of the ABAQUS tool to obtain the dynamic response of the pipeline segment under PGA, and the combined stress peak value is extracted as stress data.

[0023] Step 103: Determine the damage level of the pipe section through stress data, and count the number of times the damage level of the pipe section reaches the preset level under the same ground motion record to obtain the failure probability of the pipe section.

[0024] In this embodiment, the damage level is divided into three categories: basically completed, moderately damaged, and completely destroyed. The combined stress peak value of the pipe segment in the acceleration time history is extracted. The stress data is obtained; if the combined stress peak value is... If the damage level is basically intact, then the combined stress peak value is... If the combined stress peak value is [missing information], then the damage level is moderate. The damage level is then devastating, where, For yield strength, It represents the ultimate strength.

[0025] Step 104: A machine learning proxy model is trained using the ground motion records and the failure probabilities corresponding to the ground motion records. The failure probability of the pipe section to be tested is determined using the machine learning proxy model.

[0026] The pipe segment under test is any segment in the pipeline network system. By acquiring real-time monitoring data of the pipeline network system and real-time monitored seismic motion data, and inputting the attribute parameters of the pipe segment under test and the seismic motion records into a trained machine learning surrogate model, the failure probability of the pipe segment under test can be predicted. This failure probability represents the risk of damage to the pipe segment under test under the influence of seismic motion; the higher the failure probability, the higher the risk of damage.

[0027] The training process of the above model specifically includes: constructing a multi-dimensional feature vector using soil properties and pipe segment attribute parameters; constructing a structural response dataset using the multi-dimensional feature vector and the damage level and failure probability of the pipe segment under different peak ground accelerations; training a machine learning proxy model using the structural response dataset, wherein the machine learning proxy model includes a multi-modal feature fusion layer, a multi-scale temporal feature extraction module, and a multi-task cascaded prediction module.

[0028] The multimodal feature fusion layer calculates multidimensional feature vectors to obtain mechanically derived features. Mechanically derived features refer to physical features, which can be calculated using attribute parameters and soil properties. For example, the yield strength ratio is calculated from the yield strength and ultimate strength parameters of a pipe segment; the yield strength ratio is the ratio of yield strength to ultimate strength. Another example is calculating the product of PGA and yield strength to obtain the seismic intensity factor. The multi-scale temporal feature extraction module extracts features from the multidimensional feature vectors and acceleration time histories under different peak ground accelerations (PGAs) using a gated loop unit, obtaining multi-scale temporal features. A multi-head self-attention mechanism is then used to weight the multi-scale temporal features to obtain the output features. The multi-task cascaded prediction module includes three hierarchically connected sub-models. The first level inputs the output features into a deep neural network to determine the probability distribution vector of the damage level of the pipe segment. The second level inputs the probability distribution vector, multi-dimensional feature vectors, and mechanically derived features into the XGBoot ensemble model to output the failure probability regression prediction result. The third level uses the probability distribution vectors and failure probability regression prediction results output from the first two levels as input features and inputs them into a lightweight Bayesian linear regression model to generate the final failure probability result and confidence interval. The machine learning proxy model is adjusted by the difference between the predicted failure probability result and the failure probability in the structural response dataset, so that the difference between the model's output prediction result and the structural response dataset meets the preset conditions, thus completing the training.

[0029] The trained machine learning proxy model is used to predict the damage level of the pipe segment to be tested. The input of the model is the multi-dimensional feature vector of the pipe segment to be tested and the real-time monitored ground motion data. The output is the damage level and failure probability.

[0030] In an exemplary embodiment, an adversarial training approach is introduced. The output of the aforementioned machine learning proxy model is used as the prediction result. A discriminator network distinguishes the model's predicted response distribution from the actual finite element simulation response distribution, ensuring that the proxy model can stably output fragility assessment results that conform to engineering mechanics logic. During training, the failure probability result is input into the discriminator network, which determines the true probability of the failure probability result. The prediction difference of the true probability is calculated using a loss function, and adversarial training is performed on the machine learning proxy model and the discriminator network based on the prediction difference.

[0031] In the above steps, a machine learning surrogate model is used to quantify the risk of pipe segments under the influence of seismic motion. This implementation also quantifies the risk based on the importance of the pipe segment to the topology of the overall pipeline network system to measure the impact of pipe segment damage on the overall pipeline network system. Next, step 105: Determine the weighted graph theory model of the pipeline network system, and determine the importance index of the pipe segment to be tested through the weighted graph theory model.

[0032] In a pipeline network system, gas source points, pressure regulating stations, user access points, and pipeline connection points are treated as nodes, and the pipeline segments connecting these nodes are treated as edges. The weight vectors of these edges are determined, and a weighted graph theory model is constructed using the nodes, edges, and their weight vectors. The weighted graph theory model is represented as follows: ,in, V This represents the set of key nodes in a pipeline network system. E The pipe segment representing the connecting node; W Represents each edge The weight vector, any weight vector This represents the physical and functional attributes of the pipeline segment, which may include operating pressure, historical average flow rate, and transportation costs. The weight vector is:

[0033] in, This is the functional attribute weighting coefficient, with a value of [0,1], used to adjust the relative weight between function and cost; , The pressure component weights and flow component weights satisfy the following conditions: This is used to determine the relative importance of internal pressure and flow rate in the reaction function; The design working pressure for pipe section e; The maximum design working pressure for the pipeline section; The historical average flow rate of pipe segment e; The historical average flow rate of pipe segment e is the maximum value. The unit time transportation cost for pipe segment e; Let be the physical length of pipe segment e.

[0034] To simplify the calculation of various indicators, an equivalent length is used instead of the physical length of the pipe segment. The determination of the importance indicators of the pipe segment to be tested through a weighted graph theory model specifically includes: determining the equivalent length of the pipe segment based on its physical length; calculating the normalized edge betweenness centrality and the network efficiency degradation index of each pipe segment based on the weighted graph theory model and the equivalent length; and standardizing the normalized edge betweenness centrality and the network efficiency degradation index to obtain the importance indicators of each pipe segment.

[0035] The equivalent length is:

[0036] in, The design working pressure for pipe section e; The maximum design working pressure for the pipeline section; The historical average flow rate of pipe segment e; The historical average flow rate of pipe segment e is the maximum value. The unit time transportation cost for pipe segment e; , Let e ​​be the physical length and equivalent length of pipe segment e; , , The weights of pressure, flow, and cost components.

[0037] Importance metrics include normalized edge betweenness centrality (EBC) and the network efficiency degradation index (NERI). To measure the criticality of a segment within the network system, the normalized edge betweenness centrality (EBC) of each segment is calculated. This index quantifies the proportion of the shortest paths between all pairs of nodes in the network that pass through that edge; to quantify the impact of pipe segment failure on the overall system function, the network efficiency decline index for each pipe segment is calculated. Network efficiency measures the relative loss of overall system function (transportation efficiency) caused by pipeline failure, and is defined as network efficiency (…). The relative decline rate of ).

[0038] Normalized edge betweenness centrality is:

[0039] in, Let be the number of all shortest paths from node s to node t. Let e ​​be the number of shortest paths that pass through edge e. It represents the total number of nodes.

[0040] The Network Efficiency Reduction Index (NERI) is:

[0041] in, This represents the average connectivity efficiency of the system in the weighted graph theory model. This represents the average connectivity efficiency of the system after removing edge e;

[0042] Let be the shortest path length from node s to node t, calculated using the equivalent length. If the nodes are not connected, then... .

[0043] The above evaluation indicators were standardized to obtain the ranking of the importance of each pipe section under different evaluation indicators.

[0044] Step 106: Determine the risk assessment result of the pipe section under test based on the failure probability and importance index of the pipe section under test.

[0045] In this embodiment, the failure probability is divided into five levels, and the importance index is also divided into five levels. The risk level corresponding to the failure probability of the tested pipe segment, the risk level corresponding to the normalized edge betweenness centrality (EBC), and the risk level corresponding to the network efficiency degradation index (NERI) are determined by the value range of each level. The combined risk level of these three factors is used as the final risk assessment result. For example, by summing or weighted summing the above three risk levels, a total risk value or a total risk level can be obtained as the final risk assessment result. Therefore, the risk assessment result can be a specific total risk value, a total risk level, or it can include risk level results from multiple different factors.

[0046] In one exemplary embodiment, considering multiple evaluation factors affecting the safe operation of the pipeline network, the final risk assessment result is determined by comprehensively considering multiple evaluation factors, thereby further improving the accuracy and comprehensiveness of the risk assessment. Specifically, this includes: determining various evaluation factors in a preset evaluation index system based on real-time monitoring data of the pipeline network system. This evaluation index system includes evaluation factors at multiple levels for pipe sections; failure probability and importance indicators can each be one of the evaluation factors, and failure probability and importance indicators are divided into different levels. For example, evaluation factors can be divided into multiple levels such as human factors, pipeline factors, environmental factors, operation and maintenance factors, and pipeline topology importance. Among these, evaluation factors at the environmental factors level can include seismic motion, and evaluation factors at the pipeline topology importance level can include the aforementioned importance indicators.

[0047] Real-time monitoring data includes pipeline pressure, flow rate, and other monitoring data. Based on this data, the current value of each evaluation factor can be determined. It should be noted that this evaluation index system quantifies the risk situation at levels other than the aforementioned environmental level (seismic motion) and pipeline topology importance. The risk levels corresponding to the environmental factor level and the pipeline topology importance level are determined using the failure probability and importance index calculated in the above steps. That is, the risk level corresponding to the failure probability can be used as the risk level for seismic motion factors or the risk level for environmental factors. The importance index, namely the normalized edge betweenness centrality and the network efficiency decline index, can be used as evaluation factors for the pipeline topology importance level. Similarly, the values ​​of the normalized edge betweenness centrality and the network efficiency decline index calculated in the above steps can determine their corresponding risk levels.

[0048] For other risk levels, the calculation is performed as follows: Construct a factor comparison matrix for evaluation factors at the same level. The factor comparison matrix is ​​represented as follows:

[0049] in,

[0050] 1≤j≤n, 1≤i≤n, where n is the number of evaluation factors at the same level.

[0051] Based on the factor comparison matrix, a transformation judgment matrix is ​​solved. The risk weight vector for each evaluation factor is then calculated using the transformation judgment matrix. The risk level of each evaluation factor is determined using the risk weight vector. Finally, the risk assessment result for the tested pipe section is obtained by combining the risk levels of each evaluation factor with the failure probability and importance index of the pipe section under test. The specific solution process is as follows: Solve the transformation judgment matrix :

[0052]

[0053] In the formula, , , .

[0054] Solving for the optimal consistency matrix :

[0055]

[0056]

[0057] Next, we solve for the risk weight vector of n factors. :

[0058]

[0059]

[0060] Determining the risk level of each evaluation factor using a risk weight vector specifically includes: dividing the risk level into five levels, using the normal distribution as the membership function, and determining the level weight vector for each level through the membership function; combining the risk weight vector and the level weight vector to determine the risk level of each evaluation factor. Finally, the risk levels of all evaluation factors are combined to determine the risk level of the corresponding level.

[0061] For example, the risk levels of factors are divided into five levels (I, II, III, IV, V). A normal distribution is used as the membership function, specifically defined as:

[0062] x This refers to one possible value for the evaluation factor, where σ is the standard deviation of the membership function. It is the center of the membership function. By solving for the rank weight vector through the membership function and normalizing the membership, we can obtain the weight vector for each rank, specifically defined as: A Ⅰ = (0.4444, 0.3422, 0.1600, 0.0444, 0.0089) A Ⅱ = (0.2567, 0.3333, 0.2567, 0.1200, 0.0333) A Ⅲ = (0.1104, 0.2362, 0.3067, 0.2362, 0.1104) A Ⅳ = (0.0333, 0.1200, 0.2567, 0.3333, 0.2567) A Ⅴ = (0.0089, 0.0444, 0.1600, 0.3422, 0.4444) The membership matrix R consists of five level vectors, R = (A Ⅰ A Ⅱ A Ⅲ A Ⅳ A Ⅴ )T The evaluation factors are evaluated, and the expression is: Then, the risk level of each evaluation factor is solved by using the risk weight vector of each evaluation factor and the level vector of each factor, thereby obtaining the risk level of the level to which it belongs.

[0063] In one implementation, the method may include the following steps: S1: Construction of a representative pipe segment sample library. Based on the geometric properties, material properties, pressure ratings, and other attribute parameters of each pipe segment in the pipeline network system, a stratified random sampling method is used to select several statistically representative pipe segments from the entire system to ensure that the samples cover the key variation characteristics of the system and form a sample library.

[0064] S2: High-fidelity structural response dataset. Based on the sample library established in step S1, seismic analysis is performed using the finite element nonlinear time history analysis method. By using PGA amplitude modulation, the dynamic response of each representative pipe segment under different seismic intensities is calculated. Based on the combined stress results, the damage level of the pipe segment and the failure probability are determined, thus constructing a high-fidelity structural response dataset of "pipe segment characteristics-PGA-damage level-failure probability".

[0065] S3: Vulnerability Prediction of Pipeline System Based on Machine Learning Surrogate Model. Based on the high-fidelity dataset from step S2, a machine learning surrogate model is constructed and trained. This model takes the multi-dimensional feature vectors of the pipe segments (attribute parameters from step S1) and PGA as input, and outputs damage level and failure probability. The model learns a complex nonlinear mapping relationship from the feature space to the response space, predicting the damage level and failure probability of all pipe segments in the entire pipeline system. The trained surrogate model can replace computationally expensive numerical simulations, achieving efficient and high-precision prediction of the vulnerability of each pipe segment in the pipeline system under different seismic intensities.

[0066] S4: Pipeline Network System Topological Importance Analysis. Based on the actual distribution of urban pipeline networks, it is abstracted into a weighted graph theory model, and the structural importance of each pipeline segment in maintaining the connectivity and functional efficiency of the pipeline network system is quantified through evaluation indicators (i.e., importance indicators).

[0067] S5: Multi-level Comprehensive Risk Assessment. This integrates the pipe segment failure probability obtained in step S3 and the topological importance obtained in step S4, and accesses real-time monitoring data such as pipeline pressure and flow, as well as information on personnel, environment, operation and maintenance, and expert risk evaluation results. It adopts the classic paradigm of Risk = Probability × Consequence to construct a risk assessment framework that includes probability... U ,as a result of V A multi-level comprehensive risk assessment system. Among these, probability includes human error. U 1. Pipeline U 2. Environment U 3. Operation and Maintenance U4. Importance of network topology U 5. Consequences include casualties. V 1. Economic losses V 2. Environmental impact V 3. Social impact V 5. Finally, the fuzzy comprehensive evaluation method is introduced to determine the weights of the indicators, enabling dynamic and refined risk assessment of all pipeline sections and supporting risk tracing and prevention decisions.

[0068] S2 specifically includes: S2-1: Establish a refined finite element model of the pipe-soil interaction. For each representative pipe segment, including the pipe segment (shell element) and the soil (solid element), the pipe segment properties include diameter, wall thickness, length, burial depth, and material constitutive properties; the soil properties include soil dimensions, damping ratio, elastic modulus, expansion angle, friction angle, and cohesion; the pipe-soil interaction is defined as surface-to-surface contact, with hard contact in the normal direction and a penalty function used to define the friction coefficient in the tangential direction.

[0069] S2-2: Conduct nonlinear time history analysis. Based on the area where the pipeline network is located, select a series of representative seismic ground motion records from the local area, and adjust the peak ground acceleration (PGA) to a series of target intensity levels. Establish infinite element boundaries (CIN3D8R elements) at the bottom and sides of the soil mass to prevent boundary reflection waves from interfering with the results; apply the adjusted acceleration time history as the seismic ground motion input to all nodes at the bottom of the soil mass. Use the ABAQUS display dynamics analysis module to solve the nonlinear time history.

[0070] S2-3: Result Extraction and Damage Level Determination. Under each operating condition, extract the peak combined stress of the pipe segment throughout the entire time history (…). Based on the combined stress, the pipe section was classified as Level III, essentially intact. ), moderate damage ( ) and destruction ( ),in For yield strength, It represents the ultimate strength.

[0071] S2-4: Failure Probability Acquisition. For each representative pipe segment and each PGA level, the number of times the structural response reaches or exceeds a certain damage level under all seismic motions is counted to obtain the conditional failure probability of the pipe segment at the current PGA level and corresponding damage level. The damage levels and failure probabilities of all representative pipe segments under all PGA levels are summarized to finally form a high-fidelity structural response dataset of "pipe segment characteristics - PGA - damage level - failure probability".

[0072] S3 specifically includes: S3-1: Multimodal Feature Fusion and Physically Guided Feature Engineering Layer. The input multidimensional feature vectors (attribute parameters from step S1) undergo structured regularization and enhancement. Through the physically guided feature construction module, high-order derived features with clear mechanical meaning are automatically generated, including cross-sectional moments of inertia based on pipe diameter and wall thickness, theoretical elastic buckling stress based on material constitutive and geometric dimensions, and equivalent lateral earth pressure based on burial depth and soil parameters. This provides more physically interpretable input parameters for subsequent data-driven models.

[0073] S3-2: Multi-scale temporal feature extraction module based on gated recurrent unit and attention mechanism. Considering the time series characteristics of seismic action and the differences in dynamic response of different pipe segment properties, a dedicated feature extractor is designed. This extractor uses a gated recurrent unit (GRU) network to capture the long-short-term dependencies in the PGA time history curve and attribute sequence.

[0074] Meanwhile, a multi-head self-attention mechanism is introduced, which enables the model to dynamically and differentially focus on the key input features of the current prediction task, automatically complete the weighting of feature importance, and improve the interpretability of the model.

[0075] S3-3: Phased-Multi-Task Cascaded Prediction Model. A cascaded hybrid proxy model is constructed, consisting of three levels of professional sub-models connected in sequence, satisfying the structure of "qualitative screening - quantitative calculation - uncertainty quantification".

[0076] Level 1: Coarse assessment of damage mode. Using a deep neural network, the task is modeled as a multi-label classification problem to quickly and initially determine the probability distribution vector of each level of damage (basically intact, moderate damage, and complete destruction) of the pipe segment under given conditions.

[0077] Level 2: Conditional Failure Probability Calculation Model. Based on the probability distribution vector output from Level 1, combined with the original attribute parameters and higher-order derived features, an XGBoost ensemble model is used to specifically optimize the regression prediction of conditional failure probability. Physical constraints (e.g., failure probability monotonically non-decreasing with PGA) are embedded in the objective function to ensure that the prediction results conform to basic mechanical laws.

[0078] Level 3: Prediction Fusion and Uncertainty Quantification. The outputs of the first two levels are used as features and input into a lightweight Bayesian linear regression model to generate the final failure probability result. The confidence interval or uncertainty measure of the prediction result is output, providing a reliable basis for risk decision-making.

[0079] S3-4: System Deduction and Dynamic Update Mechanism. The trained cascaded hybrid agent model can predict the damage level and failure probability of any pipe segment within the pipeline network system. The model also supports an online learning mechanism, which can be updated when new representative pipe segment finite element analysis data is acquired, enabling continuous evolution of the assessment capability.

[0080] In S3-3, based on the problem of prediction deviation caused by the scarcity of high-fidelity sample data and the uneven distribution of samples, the adversarial training idea of ​​Generative Adversarial Network (GAN) is introduced. During the training process, the discriminator network distinguishes the model's predicted response distribution from the actual finite element simulation response distribution, ensuring that the surrogate model can stably output vulnerability assessment results that conform to the logic of engineering mechanics.

[0081] S4 specifically includes: The actual urban pipe network distribution is abstracted into a weighted graph theory model. .in V This refers to the set of key nodes in a pipeline network system, including gas source points, pressure regulating stations, user access points, and pipeline connection nodes in an urban pipeline network system. E The pipe segment representing the connecting node; W Represents each edge The set of weight vectors, any weight vector The physical and functional attributes of a pipeline segment can include operating pressure, historical average flow rate, and transportation costs, and are specifically defined as follows:

[0082] in, This is the functional attribute weighting coefficient, with a value of [0,1], used to adjust the relative weight between function and cost; , The pressure component weights and flow component weights satisfy the following conditions: This is used to determine the relative importance of internal pressure and flow rate in the reaction function; The design working pressure for pipe section e; The maximum design working pressure for the pipeline section; The historical average flow rate of pipe segment e; The historical average flow rate maximum value for pipe segment e; The unit time transportation cost for pipe segment e; Let be the physical length of pipe segment e.

[0083] To measure the criticality of a pipe segment in the network system, the normalized edge betweenness centrality of each pipe segment is calculated. This metric quantifies the proportion of shortest paths between all pairs of nodes in the network that pass through that edge. To quantify the impact of pipe segment failure on the overall system function, the network efficiency degradation index for each pipe segment is calculated. Network efficiency measures the relative loss of overall system function (transportation efficiency) caused by pipeline failure, and is defined as network efficiency (…). The relative decline rate of an edge failure. The higher the NERI value, the greater the damage to the network's functional integrity caused by the edge failure.

[0084] The above evaluation indicators were standardized to obtain the ranking of the importance of each pipe section under different evaluation indicators.

[0085] S5 specifically includes: Multi-source data standardization processing. Based on information such as failure probability, network topology importance indicators, real-time monitoring data, and expert risk evaluation results, consistency processing is performed to determine the segment plan layer division.

[0086] Factor ranking and weighting evaluation. Based on expert evaluation, the importance of factors at the same level is compared, and a factor comparison matrix is ​​established. A The transformation judgment matrix is ​​obtained by solving the factor comparison matrix A. Further solve for the quasi-consistent matrix Solve for the risk weight vector of n factors. .

[0087] Fuzzy evaluation of risk level weights. Factor risk levels are divided into five levels (I, II, III, IV, V). A normal distribution is used as the membership function to solve for the level weight vector. After normalizing the membership, the weight vector for each level can be obtained, specifically defined as: A Ⅰ = (0.4444, 0.3422, 0.1600, 0.0444, 0.0089) A Ⅱ = (0.2567, 0.3333, 0.2567, 0.1200, 0.0333) A Ⅲ = (0.1104, 0.2362, 0.3067, 0.2362, 0.1104) A Ⅳ = (0.0333, 0.1200, 0.2567, 0.3333, 0.2567) A Ⅴ = (0.0089, 0.0444, 0.1600, 0.3422, 0.4444) Comprehensive risk calculation. The membership matrix R consists of five level vectors, R = (A Ⅰ A Ⅱ A Ⅲ A Ⅳ AⅤ ) T The evaluation factors are evaluated, and the expression is: Then, the evaluation result vector for each level is calculated separately, specifically defined as:

[0088]

[0089] in, Let E be the evaluation result vector of the j-th evaluation factor in a level, and let E be the total evaluation result vector of that level.

[0090] Finally, based on the evaluation result vector, and following the principle of maximum membership, the level containing the maximum value in the vector represents the risk level for that level. The final risk assessment result is obtained by combining the risk levels of each level. For example, the risk assessment result may include the risk level corresponding to the failure probability of the tested pipe section, the risk level corresponding to the importance indicator, and the risk levels of each level in the evaluation indicator system.

[0091] In this embodiment: (1) By abstracting the complex pipeline network system into a weighted graph theory model and comprehensively using multiple indicators such as normalized edge betweenness centrality (EBC) and network efficiency degradation index (NERI), the structural importance of each pipeline segment in the overall function of the system is quantitatively evaluated from the two dimensions of connectivity and functional efficiency. This makes up for the limitation of traditional risk assessment that only focuses on individual failures and ignores the system correlation effect. It can accurately identify the key pipeline segments that have a huge impact on the integrity of network function, and provides key decision-making basis for accurate risk assessment.

[0092] (2) A high-precision and high-efficiency vulnerable cascade hybrid proxy model was constructed, which integrates advanced machine learning technologies such as physical guidance, multi-scale temporal feature extraction and multi-task learning, and introduces the adversarial training idea. It can predict the damage level and failure probability of the pipeline network system under different ground motion intensities, integrate the actual structural response into the risk assessment system, and ensure the physical rationality and interpretability of the prediction results. It provides a core tool for the rapid risk quantification of large-scale pipeline network systems.

[0093] (3) It provides a multi-level comprehensive risk assessment method, which systematically integrates pipeline attribute parameters, structural vulnerability data, geospatial information, external environmental loads, as well as multi-source heterogeneous information such as expert knowledge, real-time monitoring data, human factors, and environment. It establishes a multi-level comprehensive risk assessment model with dynamic weight allocation, which is applicable to scenarios with multiple disasters and multiple factors coupled together. It realizes a fundamental shift from passive response to proactive prevention and control, and significantly improves the comprehensiveness, scientificity, and accuracy of risk assessment of urban lifeline systems.

[0094] Furthermore, this embodiment also provides a risk assessment device for urban lifeline pipeline networks, which can be used to perform the aforementioned risk assessment method for urban lifeline pipeline networks. For example... Figure 2 As shown, the risk assessment device 200 for the city's lifeline pipeline network includes: a finite element model construction module 201, used to acquire the attribute parameters of each pipe segment and soil properties in the pipeline network system, and construct a finite element model of the pipeline network system based on the attribute parameters of the pipe segments and soil properties; a stress calculation module 202, used to acquire seismic ground motion records of the area where the pipeline network system is located, perform amplitude modulation on the seismic ground motion records to obtain acceleration time histories under different peak ground accelerations, perform dynamic analysis based on the amplitude-modulated acceleration time histories to obtain stress data of the pipe segments under different peak ground accelerations; and a failure determination module 203, used to determine the failure of the pipe segments through the stress data. The system calculates the failure level of a pipe segment and counts the number of times the failure level reaches a preset level under the same seismic motion record to obtain the failure probability of the pipe segment. A seismic motion risk prediction module 204 is used to train a machine learning proxy model using the seismic motion record and the corresponding failure probability, and then uses the machine learning proxy model to determine the failure probability of the pipe segment under test. A topology processing module 205 is used to determine a weighted graph theory model of the pipeline network system, and then uses the weighted graph theory model to determine the importance index of the pipe segment under test. An evaluation result determination module 206 is used to determine the risk assessment result of the pipe segment under test based on the failure probability and importance index.

[0095] The specific details of each module or unit in the aforementioned risk assessment device for urban lifeline pipeline networks have been described in detail in the corresponding risk assessment method for urban lifeline pipeline networks, so they will not be repeated here.

[0096] This application also provides an electronic device. Figure 3 A schematic diagram of the structure of an electronic device suitable for implementing embodiments of the present disclosure is shown. Figure 3 The electronic device 600 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0097] like Figure 3 As shown, the electronic device 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes based on a program stored in a read-only memory (ROM) 602 or a program loaded from a storage section 608 into a random access memory (RAM) 603. The RAM 603 also stores various programs and data required for system operation. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0098] The following components are connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 610 as needed so that computer programs read from it can be installed into storage section 608 as needed.

[0099] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611. When the computer program is executed by central processing unit (CPU) 601, it performs the functions defined in the embodiments of this application.

[0100] For example, when the computer program is executed by the central processing unit (CPU) 601, it can perform the following: acquire the attribute parameters of each pipe segment and the soil properties in the pipeline network system; construct a finite element model of the pipeline network system based on the attribute parameters of the pipe segments and the soil properties; acquire seismic ground motion records of the area where the pipeline network system is located; perform amplitude modulation on the seismic ground motion records to obtain acceleration time histories under different peak ground accelerations; perform dynamic analysis based on the amplitude-modulated acceleration time histories to obtain stress data of the pipe segments under different peak ground accelerations; determine the damage level of the pipe segments through the stress data, and count the number of times the damage level of the pipe segments reaches a preset level under the same seismic ground motion record to obtain the failure probability of the pipe segments; train a machine learning surrogate model through the seismic ground motion records and the failure probabilities corresponding to the seismic ground motion records; determine the failure probability of the pipe segment to be tested through the machine learning surrogate model; determine a weighted graph theory model of the pipeline network system; determine the importance index of the pipe segment to be tested through the weighted graph theory model; and determine the risk assessment result of the pipe segment to be tested based on the failure probability and importance index of the pipe segment to be tested.

[0101] It should be noted that the computer-readable medium disclosed herein may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0102] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0103] The units described in the embodiments of this disclosure can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the unit itself.

[0104] In another aspect, this application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The computer-readable medium carries one or more programs, which include instructions that, when executed by the electronic device, cause the electronic device to perform the methods described in the above embodiments.

[0105] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0106] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A risk assessment method for urban lifeline pipeline networks, characterized in that, include: Obtain the attribute parameters of each pipe segment and the soil properties in the pipeline network system, and construct the finite element model of the pipeline network system based on the attribute parameters of the pipe segments and the soil properties. Seismic ground motion records of the area where the pipeline network system is located are obtained, and the seismic ground motion records are amplitude-modulated to obtain acceleration time histories under different peak ground accelerations. Based on the amplitude-modulated acceleration time histories, dynamic analysis is performed to obtain stress data of the pipe section under different peak ground accelerations. The failure level of the pipe section is determined by stress data, and the failure probability of the pipe section is obtained by counting the number of times the failure level of the pipe section reaches the preset level under the same ground motion record. A machine learning proxy model is obtained by training the ground motion records and the corresponding failure probabilities of the ground motion records, and the failure probability of the pipe section to be tested is determined by the machine learning proxy model. Determine the weighted graph theory model of the pipeline network system, and use the weighted graph theory model to determine the importance index of the pipe segment to be tested; The risk assessment result of the pipe section under test is determined based on the failure probability and importance index of the pipe section under test.

2. The risk assessment method for urban lifeline pipeline networks according to claim 1, characterized in that, The step of training a machine learning proxy model using the seismic motion records and the corresponding failure probabilities includes: A multidimensional feature vector is constructed by combining soil properties and pipe segment properties. A structural response dataset is then constructed by combining the multidimensional feature vector with the damage level and failure probability of the pipe segment under different peak ground accelerations. The machine learning agent model is trained using a structured response dataset. The machine learning agent model includes a multimodal feature fusion layer, a multi-scale temporal feature extraction module, and a multi-task cascaded prediction module. The multimodal feature fusion layer calculates multidimensional feature vectors to obtain mechanically derived features. The multi-scale temporal feature extraction module extracts features from the multi-dimensional feature vector and the acceleration time history under different peak ground accelerations through a gated loop unit to obtain multi-scale temporal features. The multi-head self-attention mechanism is used to assign weights to the multi-scale temporal features to obtain the output features. The multi-task cascaded prediction module includes three hierarchical sub-models connected in sequence. The first level inputs the output features into a deep neural network to determine the probability distribution vector of the damage level of the pipe segment. The second level inputs the probability distribution vector, multi-dimensional feature vector, and mechanically derived features into the XGBoot ensemble model and outputs the failure probability regression prediction result. The third level uses the probability distribution vector and failure probability regression prediction result output from the first two levels as input features and inputs them into a lightweight Bayesian linear regression model to generate the final failure probability result and confidence interval. The machine learning proxy model is adjusted by comparing the predicted failure probability with the failure probability in the structural response dataset.

3. The risk assessment method for urban lifeline pipeline networks according to claim 2, characterized in that, The step of adjusting the machine learning proxy model based on the difference between the predicted failure probability and the failure probability in the structural response dataset includes: The failure probability result is input into the discriminator network, and the discriminator network determines the true probability of the failure probability result. The predicted difference of the true probability is calculated using a loss function, and the machine learning agent model and discriminator network are then subjected to adversarial training based on the predicted difference.

4. The risk assessment method for urban lifeline pipeline networks according to claim 1, characterized in that, The weighted graph theory model for determining the pipeline network system includes: The gas source point, pressure regulating station, user access point and pipeline connection point in the pipeline network system are taken as nodes, and the pipeline segments connecting the nodes are taken as edges. The weight vector of the edge is determined, and a weighted graph theory model is constructed through the nodes, edges and edge weight vectors. The weight vector is: in, This is the functional attribute weighting coefficient, with a value of [0,1], used to adjust the relative weight between function and cost; , The pressure component weights and flow component weights satisfy the following conditions: This is used to determine the relative importance of internal pressure and flow rate in the reaction function; The design working pressure for pipe section e; The maximum design working pressure for the pipeline section; The historical average flow rate of pipe segment e; The historical average flow rate of pipe segment e is the maximum value. The unit time transportation cost for pipe segment e; Let be the physical length of pipe segment e.

5. The risk assessment method for urban lifeline pipeline networks according to claim 4, characterized in that, The determination of importance indicators for the pipe segment under test using a weighted graph theory model includes: The equivalent length of the pipe segment is determined based on its physical length, and the equivalent length is: in, The design working pressure for pipe section e; The maximum design working pressure for the pipeline section; The historical average flow rate of pipe segment e; The historical average flow rate of pipe segment e is the maximum value. The unit time transportation cost for pipe segment e; , Let e ​​be the physical length and equivalent length of pipe segment e; , , The weights of pressure, flow, and cost components; Based on the weighted graph theory model and the equivalent length, the normalized edge betweenness centrality and network efficiency decline index of each pipe segment are calculated. The normalized edge betweenness centrality and network efficiency degradation index are standardized to obtain the importance index of each pipe segment.

6. The risk assessment method for urban lifeline pipeline networks according to claim 1, characterized in that, The risk assessment result for the pipe section under test, determined based on the failure probability and importance index, includes: Based on real-time monitoring data of the pipeline network system, the evaluation factors in the preset evaluation index system are determined. The evaluation index system includes evaluation factors at multiple levels for the pipeline section. Construct a factor comparison matrix for evaluation factors at the same level. The factor comparison matrix is ​​represented as follows: in, Based on the aforementioned factor comparison matrix, the transformation judgment matrix is ​​solved, and the risk weight vector of each evaluation factor is solved through the transformation judgment matrix. The risk level of each evaluation factor is determined by a risk weight vector. The risk assessment result of the pipe section under test is obtained by combining the risk level of each evaluation factor, the failure probability of the pipe section under test, and the importance index.

7. The risk assessment method for urban lifeline pipeline networks according to claim 6, characterized in that, The process of determining the risk level of each evaluation factor through a risk weight vector includes: The risk level is divided into five levels, and the normal distribution is used as the membership function. The level weight vector of each level is determined by the membership function. The risk level of each evaluation factor is determined by combining the risk weight vector and the level weight vector.

8. The risk assessment method for urban lifeline pipeline networks according to claim 5, characterized in that, The normalized edge betweenness centrality is: in, Let be the number of all shortest paths from node s to node t. Let e ​​be the number of shortest paths that pass through edge e. It is the total number of nodes; The network efficiency degradation index is: in, This represents the average connectivity efficiency of the system in the weighted graph theory model. This represents the average connectivity efficiency of the system after removing edge e; Let be the shortest path length from node s to node t, calculated using the equivalent length. If the nodes are not connected, then... .

9. The risk assessment method for urban lifeline pipeline networks according to claim 1, characterized in that, The determination of the damage level of the pipe section using stress data includes: Extract the peak combined stress of the pipe section in the acceleration time history. Stress data is obtained; If the combined stress peak If the damage level is basically intact, then the combined stress peak value is... If the combined stress peak value is [missing information], then the damage level is moderate. The damage level is then devastating, where, For yield strength, It represents the ultimate strength.

10. A risk assessment device for urban lifeline pipeline networks, characterized in that, include: The finite element model construction module is used to obtain the attribute parameters of each pipe segment and the soil properties in the pipeline network system, and to construct the finite element model of the pipeline network system based on the attribute parameters of the pipe segments and the soil properties. The stress calculation module is used to acquire ground motion records of the area where the pipeline system is located, perform amplitude modulation on the ground motion records to obtain acceleration time histories under different peak ground accelerations, and perform dynamic analysis based on the amplitude-modulated acceleration time histories to obtain stress data of the pipe section under different peak ground accelerations. The failure determination module is used to determine the damage level of the pipe section through stress data and count the number of times the damage level of the pipe section reaches the preset level under the same earthquake record to obtain the failure probability of the pipe section. The seismic motion risk prediction module is used to train a machine learning proxy model through the seismic motion records and the failure probabilities corresponding to the seismic motion records, and to determine the failure probability of the pipe section to be tested through the machine learning proxy model. The topology processing module is used to determine the weighted graph theory model of the pipeline network system, and to determine the importance index of the pipe segment to be tested through the weighted graph theory model; The assessment result determination module is used to determine the risk assessment result of the pipe section under test based on the failure probability and importance index of the pipe section under test.

Images