Dynamic risk analysis method and system based on knowledge graph and digital twinborn model

Abstract

The invention relates to the technical field of risk analysis, in particular to a dynamic risk analysis method and system based on a knowledge graph and a digital twin model. The method comprises the following steps: acquiring a preset three-dimensional twin model of a hydrogen cylinder; determining a key weak area of the hydrogen cylinder according to a preset three-dimensional twinborn model; predicting a potential failure mode of the hydrogen cylinder according to the key weak area on the basis of a preset knowledge graph; and according to the failure mode, simulating a risk evolution state of the hydrogen cylinder, and according to a simulation result, determining a transportation strategy. Meanwhile, a foundation is laid for the risk evolution state, and it is ensured that the input data is accurate and timely. The overhead of comprehensive scanning is reduced, knowledge graph query and risk simulation are ensured to be concentrated on high-probability failure points, and the overall analysis efficiency is improved. Operable input is provided for risk evolution state simulation, and the reliability and pertinence of risk identification are enhanced. The dynamic association of risk prevention and control and transportation decision is realized, and the initiative of safety management is improved.

Description

Technical Field

[0001] This application relates to the field of risk analysis technology, and in particular to a dynamic risk analysis method and system based on knowledge graphs and digital twin models. Background Technology

[0002] Hydrogen, as a clean and efficient energy carrier, is increasingly used in the transportation sector. However, hydrogen cylinders, as key components for storing and transporting hydrogen, face various failure risks under complex operating conditions, including hydrogen embrittlement, delamination of composite materials, and bulging and collapse of the plastic liner.

[0003] However, existing risk analysis methods for hydrogen cylinders can only predict approximate failure areas based on knowledge graphs. This approach leads to crude strategy formulation in complex scenarios, failing to support real-time decision-making and increasing the risk of safety incidents. Summary of the Invention

[0004] This application provides a dynamic risk analysis method and system based on knowledge graphs and digital twin models to solve the above problems.

[0005] Firstly, this application provides a dynamic risk analysis method based on knowledge graphs and digital twin models, the method comprising: Obtain a pre-defined 3D twin model of the hydrogen cylinder; Based on the preset three-dimensional twin model, the key weak areas of the hydrogen cylinder are determined; Based on a pre-defined knowledge graph, the potential failure modes of the hydrogen cylinder are predicted according to the key weak areas. Based on the failure modes, the risk evolution state of the hydrogen cylinder is simulated, and a transportation strategy is determined based on the simulation results.

[0006] This solution acquires a pre-defined 3D twin model of the hydrogen cylinder, ensuring a unified digital representation and avoiding data fragmentation. It also lays the foundation for understanding risk evolution and ensures accurate and timely input data. Based on the pre-defined 3D twin model, key weak areas of the hydrogen cylinder are identified, reducing the overhead of comprehensive scanning and ensuring that knowledge graph queries and risk simulations focus on high-probability failure points, thus improving overall analysis efficiency. Based on the pre-defined knowledge graph and key weak areas, potential failure modes of the hydrogen cylinder are predicted, providing actionable input for risk evolution simulation and enhancing the reliability and specificity of risk identification. Based on the failure modes, the risk evolution state of the hydrogen cylinder is simulated. Based on the simulation results, transportation strategies are determined, achieving a dynamic link between risk prevention and transportation decisions, and enhancing the proactiveness of safety management.

[0007] Optionally, obtaining the preset three-dimensional twin model of the hydrogen cylinder includes: Obtaining component information of the hydrogen cylinder; analyzing the component information to determine component geometric parameters and component material properties; According to the component geometric parameters and the component material properties, a stress-strain curve under a hydrogen environment is constructed; According to the component material properties, fiber orientation and matrix performance of a composite material are determined; According to the component geometric parameters and the component material properties, the thickness and hydrogen permeation coefficient of the plastic liner are determined; According to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient, the preset three-dimensional twin model is generated.

[0008] Through the scheme, the component information of the hydrogen cylinder is obtained, and complete initial information support is ensured for the preset three-dimensional twin model construction process. The component information is analyzed to determine the component geometric parameters and the component material properties, which provides accurate input parameters for constructing the stress-strain curve, eliminates data ambiguity and supports parameterized modeling. According to the component geometric parameters and the component material properties, a stress-strain curve under a hydrogen environment is constructed to reflect the actual deformation characteristics under a hydrogen environment. According to the component material properties, the fiber orientation and matrix performance of a composite material are determined to support accurate characterization of the failure behavior of the composite material. According to the component geometric parameters and the component material properties, the thickness and hydrogen permeation coefficient of the plastic liner are determined to ensure that the preset three-dimensional twin model accurately simulates the physical and chemical behavior of the liner. According to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient, the preset three-dimensional twin model is generated to support failure simulation and strategy optimization.

[0009] Optionally, the generating of the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient comprises: According to the stress-strain curve, hydrogen embrittlement sensitivity is determined; Obtaining guided wave monitoring data under cyclic load; Based on the fiber orientation and the matrix performance, the guided wave monitoring data under the cyclic load is used to determine the debonding probability of the composite layer; Obtaining the acoustic emission signal of the plastic liner; Based on the thickness and the hydrogen permeation coefficient, the acoustic emission signal is analyzed to determine the bulge collapse critical value; According to the hydrogen embrittlement sensitivity, the debonding probability and the bulge collapse critical value, the preset three-dimensional twin model is generated.

[0010] According to the stress-strain curve, the hydrogen embrittlement sensitivity is determined, the prediction accuracy of hydrogen embrittlement failure is improved, the guided wave monitoring data under cyclic load is obtained, and the response to dynamic working condition changes is ensured. Based on the fiber orientation and the matrix performance, the debonding probability of the composite layer is determined according to the guided wave monitoring data under cyclic load, and the dynamic capture ability of the delamination failure is enhanced. The acoustic emission signal of the plastic liner is obtained, and the actual degradation process of the plastic liner is reflected. Based on the thickness and the hydrogen permeation coefficient, the acoustic emission signal is analyzed, the bulging collapse critical value is determined, and the prediction reliability of the instability of the liner caused by hydrogen permeation is improved. According to the hydrogen embrittlement sensitivity, the debonding probability and the bulging collapse critical value, a preset three-dimensional twin model is generated, the defects of insufficient modeling integrity and weak decision support are eliminated, and the failure mode detail simulation and real-time strategy optimization are supported.

[0011] Optionally, the potential failure mode of the hydrogen cylinder is predicted based on the preset knowledge graph according to the key weak area, comprising: Obtain historical failure cases; Analyze the historical failure cases to determine the historical damage mode; Analyze the preset knowledge graph to determine the node association; According to the node association, the coupling effect between the composite layer and the metal layer is determined; Obtain visual detection data, analyze the visual detection data, and determine the wear condition of the key weak area; According to the wear condition, the coupling effect and the historical damage mode, the potential failure mode of the hydrogen cylinder is predicted.

[0012] Through the scheme, the historical failure cases are obtained, and the prediction is enhanced in the aspect of empirical and relevance by avoiding reasoning in the air. The historical failure cases are analyzed to determine the historical damage mode, and the accuracy and consistency of the prediction are improved. The preset knowledge graph is analyzed to determine the node association, and the hidden failure association path is captured in the prediction process. According to the node association, the coupling effect between the composite layer and the metal layer is determined, and the cross-material failure transmission path is clear. The visual detection data is obtained, the visual detection data is analyzed, and the wear condition of the key weak area is determined, which reflects the contribution of the actual use trace to the failure. According to the wear condition, the coupling effect and the historical damage mode, the potential failure mode of the hydrogen cylinder is predicted, which is associated with the operable failure type and position.

[0013] Optionally, the transportation strategy is determined according to the simulation result, comprising: Obtain the vibration frequency spectrum data of the current transportation route; Analyze the vibration frequency spectrum data to determine the resonance frequency coincidence degree of the vibration frequency spectrum and the hydrogen embrittlement sensitivity; According to the resonance frequency coincidence degree, a detour instruction is constructed, and the detour instruction is determined as the transportation strategy.

[0014] By this scheme, the vibration spectrum data of the current transportation route is obtained, the potential vibration source causing hydrogen embrittlement of the metal part is identified, and the defect of weak decision support is eliminated. The vibration spectrum data is analyzed to determine the resonance frequency coincidence degree of the vibration spectrum and the hydrogen embrittlement sensitivity, in response to the deficiency that the vibration spectrum data cannot be dynamically associated with the hydrogen embrittlement sensitivity, the resonance risk is quantitatively identified, and an operable basis is provided for generating the bypass instruction. According to the resonance frequency coincidence degree, the bypass instruction is constructed, the bypass instruction is determined as the transportation strategy, the high-risk scene caused by vibration resonance is actively avoided, the problem that the high-risk scene cannot be actively avoided in the transportation process is eliminated, and the initiative of risk prevention and control is improved.

[0015] Optionally, the bypass instruction is constructed according to the resonance frequency coincidence degree, comprising: acquiring a transportation task of the hydrogen cylinder; analyzing the transportation task to determine a backup route; acquiring route information of the backup route; analyzing the route information to determine road flatness and slope data; analyzing the road flatness and the slope data to determine real-time influence on the hydrogen cylinder; constructing a bypass instruction according to the real-time influence and the resonance frequency coincidence degree.

[0016] By this scheme, the transportation task of the hydrogen cylinder is obtained to ensure that the risk assessment process is started based on actual logistics demand. The transportation task is analyzed to determine a backup route, which avoids providing feasible alternative path options to ensure that the transportation strategy has flexibility. The route information of the backup route is acquired to enrich the route data, so that the digital twin model accurately maps the influence of the external environment on the hydrogen cylinder. The route information is analyzed to determine the road flatness and slope data, and the potential effect of the road condition on the hydrogen cylinder is evaluated. The road flatness and slope data are analyzed to determine the real-time influence on the hydrogen cylinder, which provides a basis for decision-making to ensure that the risk prediction is real-time associated with the current route condition. The bypass instruction is constructed according to the real-time influence and the resonance frequency coincidence degree, the dynamic optimization of the transportation strategy is realized, and the high-risk scene is actively avoided.

[0017] Optionally, the analysis of the preset knowledge graph to determine the node association includes: based on the historical failure cases, analyzing the preset knowledge graph to determine associated nodes; acquiring stress distribution data of the hydrogen cylinder; matching the stress distribution data with the historical failure cases, and determining a load similarity according to the matching result; determining the association weight between any associated nodes according to the similarity; According to the correlation weight, the node correlation is determined.

[0018] Through the scheme, based on historical failure cases, the preset knowledge graph is analyzed, the associated nodes are determined, the defects of insufficient modeling integrity are eliminated, and the nodes such as hydrogen embrittlement sensitivity of metal parts and risk of composite layer are ensured to be included in the analysis. The stress distribution data of the hydrogen cylinder is obtained, and the defects of being unable to accurately capture the thermal stress failure are made up. The stress distribution data is matched with the historical failure cases, the load similarity is determined according to the matching result, and the defects of lacking accurate simulation of failure details are eliminated. According to the similarity, the correlation weight between any associated nodes is determined, the problem of missing group risk modeling is eliminated, and the group risk modeling is supported. According to the correlation weight, the node correlation is determined, the defects of weak decision support are responded, and the correlation basis is provided for the transportation strategy.

[0019] Optionally, after the bypass instruction is constructed according to the real-time influence and the resonance frequency coincidence degree, the method further includes: The transportation task is analyzed to determine a task time limit requirement; The bypass instruction is analyzed to determine a predicted delay time of the bypass path; The safety gain is determined according to the real-time influence; The bypass necessity is determined according to the predicted delay time and the safety gain based on the task time limit requirement.

[0020] Through the scheme, the transportation task is analyzed to determine the task time limit requirement, the evaluation is ensured to be carried out within the task time limit, and the decision deviation caused by ignoring the time limit constraint is avoided. The bypass instruction is analyzed to determine the predicted delay time of the bypass path, the time cost of the bypass strategy is quantified, and the basis for weighing the safety benefit and the time loss is provided. The safety gain is determined according to the real-time influence, the safety improvement degree of the bypass strategy is reflected, and the measurable risk relief benefit is provided for the decision. The bypass necessity is determined according to the predicted delay time and the safety gain based on the task time limit requirement, the dynamic optimization of the transportation strategy is realized, and the high-risk scene is actively avoided under the premise of meeting the time limit.

[0021] Optionally, after the blister collapse critical value is determined based on the thickness and the hydrogen permeation coefficient, the method further includes: Real-time temperature data of the plastic liner is obtained; The temperature data is analyzed to determine a permeation influence of temperature on the hydrogen permeation coefficient; The blister collapse critical value is updated according to the permeation influence.

[0022] By the scheme, real-time temperature data of the plastic liner is acquired, and correction of the bulge collapse critical value is ensured to be based on actual environmental conditions. The temperature data is analyzed to determine the permeation influence of temperature on the hydrogen permeation coefficient, and calculation deviation of the permeation rate caused by temperature fluctuation is eliminated. According to the permeation influence, the bulge collapse critical value is updated, and the prediction accuracy of the plastic liner bulge collapse is improved.

[0023] In a second aspect, the application provides a dynamic risk analysis system based on a knowledge graph and a digital twin model, the system comprising: a model acquisition module configured to acquire a preset three-dimensional twin model of a hydrogen cylinder; a region determination module configured to determine a key weak region of the hydrogen cylinder based on the preset three-dimensional twin model; a mode prediction module configured to predict a potential failure mode of the hydrogen cylinder based on a preset knowledge graph and the key weak region; a transportation determination module configured to simulate a risk evolution state of the hydrogen cylinder according to the failure mode, and determine a transportation strategy according to a simulation result.

[0024] Optionally, when the model acquisition module acquires the preset three-dimensional twin model of the hydrogen cylinder, the model acquisition module is configured to: acquire component information of the hydrogen cylinder; analyze the component information to determine component geometric parameters and component material properties; construct a stress-strain curve under a hydrogen environment according to the component geometric parameters and the component material properties; determine fiber orientation and matrix performance of a composite material according to the component material properties; determine a thickness of a plastic liner and a hydrogen permeation coefficient according to the component geometric parameters and the component material properties; generate the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness, and the hydrogen permeation coefficient.

[0025] Optionally, when the model acquisition module generates the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness, and the hydrogen permeation coefficient, the model acquisition module is configured to: determine hydrogen embrittlement sensitivity according to the stress-strain curve; acquire guided wave monitoring data under cyclic load; determine debonding probability of a composite layer according to the guided wave monitoring data under the cyclic load based on the fiber orientation and the matrix performance; acquire an acoustic emission signal of the plastic liner; analyze the acoustic emission signal to determine a bulge collapse critical value based on the thickness and the hydrogen permeation coefficient; According to the hydrogen embrittlement sensitivity, the debonding probability and the bulge collapse critical value, the preset three-dimensional twin model is generated.

[0026] Optionally, when the mode prediction module predicts the potential failure mode of the hydrogen cylinder based on the preset knowledge graph according to the key weak area, it is used for: acquiring historical failure cases; analyzing the historical failure cases to determine historical damage modes; analyzing the preset knowledge graph to determine node correlation; determining the coupling effect between the composite layer and the metal layer according to the node correlation; acquiring visual inspection data, analyzing the visual inspection data to determine the wear condition of the key weak area; According to the wear condition, the coupling effect and the historical damage mode, the potential failure mode of the hydrogen cylinder is predicted.

[0027] Optionally, when the transportation determination module determines the transportation strategy according to the simulation result, it is used for: acquiring vibration spectrum data of the current transportation route; analyzing the vibration spectrum data to determine the resonance frequency coincidence degree of the vibration spectrum and the hydrogen embrittlement sensitivity; According to the resonance frequency coincidence degree, the detour instruction is constructed, and the detour instruction is determined as the transportation strategy.

[0028] Optionally, when the transportation determination module constructs the detour instruction according to the resonance frequency coincidence degree, it is used for: acquiring the transportation task of the hydrogen cylinder; analyzing the transportation task to determine a backup route; acquiring route information of the backup route; analyzing the route information to determine road flatness and slope data; analyzing the road flatness and the slope data to determine the real-time influence on the hydrogen cylinder; According to the real-time influence and the resonance frequency coincidence degree, the detour instruction is constructed.

[0029] Optionally, when the mode prediction module analyzes the preset knowledge graph to determine the node correlation, it is used for: Based on the historical failure cases, the preset knowledge graph is analyzed to determine the associated nodes; acquiring stress distribution data of the hydrogen cylinder; matching the stress distribution data with historical failure cases, and determining the load similarity according to the matching result; According to the similarity, the correlation weight between any associated nodes is determined; According to the correlation weight, the node correlation is determined.

[0030] Optionally, the dynamic risk analysis system based on the knowledge graph and the digital twin model further comprises a necessity determination module configured to: analyze the transportation task to determine a task timeliness requirement; analyze the detour instruction to determine a predicted delay time of the detour path; determine a safety gain according to the real-time influence; determine a detour necessity based on the task timeliness requirement, the predicted delay time and the safety gain.

[0031] Optionally, the dynamic risk analysis system based on the knowledge graph and the digital twin model further comprises a critical value updating module configured to: obtain real-time temperature data of the plastic liner; analyze the temperature data to determine a permeation influence of the temperature on the hydrogen permeation coefficient; update the bulge collapse critical value according to the permeation influence. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0033] Figure 1 An application scenario schematic diagram provided by an embodiment of the present application; Figure 2 A flowchart of a dynamic risk analysis method based on a knowledge graph and a digital twin model provided by an embodiment of the present application; Figure 3 A structural schematic diagram of a dynamic risk analysis system based on a knowledge graph and a digital twin model provided by an embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme of the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0035] In addition, the term "and / or" in this document merely describes an association relationship of associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone. In addition, the character " / " in this document generally represents an "or" relationship between the front and rear associated objects unless otherwise specified.

[0036] The embodiments of the present application will be further described in detail below with reference to the accompanying drawings of the specification.

[0037] The existing hydrogen cylinder risk analysis method can only predict the approximate failure area based on the knowledge graph. This way will lead to a rough strategy when facing complex scenarios, which cannot support real-time decision-making and increase the risk of safety accidents.

[0038] Based on this, the present application provides a dynamic risk analysis method and system based on knowledge graph and digital twin model, acquires a preset three-dimensional twin model of the hydrogen cylinder, ensures a unified digital representation, avoids data fragmentation, and lays a foundation for risk evolution state, ensures accurate and timely input data. According to the preset three-dimensional twin model, the key weak area of the hydrogen cylinder is determined, the overhead of comprehensive scanning is reduced, the knowledge graph query and risk simulation are concentrated on the high probability failure point, and the overall analysis efficiency is improved. Based on the preset knowledge graph, according to the key weak area, the potential failure mode of the hydrogen cylinder is predicted, an operable input is provided for risk evolution state simulation, and the reliability and pertinence of risk identification are enhanced. According to the failure mode, the risk evolution state of the hydrogen cylinder is simulated, according to the simulation result, the transportation strategy is determined, the dynamic correlation of risk prevention and control and transportation decision is realized, and the initiative of safety management is improved.

[0039] Figure 1 An application scenario diagram is provided for the present application. When performing dynamic risk analysis, the method provided by the present application is applied.

[0040] Specifically, the method provided by the present application is applied in any server, the server interacts with a three-dimensional modeling tool, and a preset three-dimensional twin model of the hydrogen cylinder is constructed through the three-dimensional modeling tool. According to the preset three-dimensional twin model, the key weak area of the hydrogen cylinder is determined. Based on the preset knowledge graph, according to the key weak area, the potential failure mode of the hydrogen cylinder is predicted, an operable input is provided for risk evolution state simulation, and the reliability and pertinence of risk identification are enhanced. According to the failure mode, the risk evolution state of the hydrogen cylinder is simulated, according to the simulation result, the transportation strategy is determined, the dynamic correlation of risk prevention and control and transportation decision is realized, and the initiative of safety management is improved. The specific implementation mode can refer to the following embodiments.

[0041] Figure 2A flowchart of a dynamic risk analysis method based on a knowledge graph and a digital twin model is provided for an embodiment of the present application. The method of the embodiment can be applied to the server in the above scenarios. As shown in the figure, Figure 2 The method comprises the following steps: S201, obtaining a preset three-dimensional twin model of a hydrogen cylinder. The hydrogen cylinder can be a pressure container for storing and transporting hydrogen.

[0042] The preset three-dimensional twin model can be a pre-constructed virtual mapping model of the hydrogen cylinder. It is pre-stored in the server and called when used.

[0043] Specifically, based on the design data and historical operation data of the hydrogen cylinder, the geometric data of the hydrogen cylinder is imported through a three-dimensional modeling tool, and the material attribute parameters are input. Then, the model state is dynamically updated in combination with real-time sensor data, thereby constructing the preset three-dimensional twin model of the hydrogen cylinder.

[0044] S202, determining the key weak area of the hydrogen cylinder according to the preset three-dimensional twin model. The key weak area can be a physical part of the hydrogen cylinder that is prone to failure.

[0045] Specifically, the stress-strain simulation of the preset three-dimensional twin model is performed by using finite element analysis, the stress distribution under different working conditions is calculated, and the high stress area is identified. Then, based on the reliability engineering theory, the typical failure cases accumulated for a long time are stored in a structured manner, and a preset database is generated. Then, the historical failure records in the preset database are queried, and the similar structural features in the preset three-dimensional twin model are matched. Finally, the analysis results are summarized and analyzed according to the clustering algorithm, and the key weak area of the hydrogen cylinder is determined.

[0046] S203, predicting the potential failure mode of the hydrogen cylinder based on the preset knowledge graph and the key weak area. The preset knowledge graph can be a pre-constructed structured knowledge base that organizes entities and associated relationships in historical failure cases in a graph structure. It is pre-stored in the server and called when used.

[0047] The potential failure mode can be a type of failure that may occur in the hydrogen cylinder.

[0048] Specifically, based on the key weak area, the entity relationship modeling of the historical failure cases is performed according to the ontology and semantic network theory, thereby constructing the preset knowledge graph. Then, the related failure modes in the preset knowledge graph are searched. Then, the rule-based reasoning engine is applied, and the real-time environmental data collected by the environmental sensor is combined to predict the potential failure mode of the hydrogen cylinder.

[0049] S204, simulate the risk evolution state of the hydrogen cylinder according to the failure mode, and determine the transportation strategy according to the simulation result.

[0050] The risk evolution state can be a dynamic development process of the failure behavior.

[0051] The simulation result can be output data of the risk evolution state simulation.

[0052] The transportation strategy can be executable optimization instructions generated for the simulation result.

[0053] Specifically, the risk evolution state of the failure mode is simulated by using a computer simulation device: first, the failure position and size are determined according to the failure mode; then, the failure direction, rate and damage range are calculated in combination with real-time environmental data. Based on the simulation result, real-time transportation data collected by the vehicle-mounted sensor and positioning device are combined to generate a transportation strategy by using a heuristic algorithm.

[0054] Through the scheme, the preset three-dimensional twin model of the hydrogen cylinder is obtained, ensuring that the unified digital representation is based on, avoiding data fragmentation, and laying a foundation for the risk evolution state, ensuring that the input data is accurate and timely. According to the preset three-dimensional twin model, the key weak area of the hydrogen cylinder is determined, reducing the overhead of comprehensive scanning, ensuring that the knowledge graph query and risk simulation are concentrated on the high-probability failure point, and improving the overall analysis efficiency. Based on the preset knowledge graph, the potential failure mode of the hydrogen cylinder is predicted according to the key weak area, providing an operable input for the risk evolution state simulation, and enhancing the reliability and pertinence of risk identification. According to the failure mode, the risk evolution state of the hydrogen cylinder is simulated, and the transportation strategy is determined according to the simulation result, realizing the dynamic correlation of risk prevention and control and transportation decision, and improving the initiative of safety management.

[0055] In some embodiments, component information of the hydrogen cylinder is obtained; the component information is analyzed to determine component geometric parameters and component material properties; a stress-strain curve under a hydrogen environment is constructed according to the component geometric parameters and the component material properties; fiber orientation and matrix performance of a composite material are determined according to the component material properties; the thickness of a plastic liner and the hydrogen permeation coefficient are determined according to the component geometric parameters and the component material properties; and the preset three-dimensional twin model is generated according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient.

[0056] The component information can be descriptive data of each component of the hydrogen cylinder.

[0057] The component geometric parameters can be size, shape and position parameters of the hydrogen cylinder components.

[0058] The component material properties can be material characteristic parameters of the hydrogen cylinder components.

[0059] A stress-strain curve can be a curve that describes the deformation behavior of a material when subjected to force.

[0060] A composite material can be a material structure composed of fibers and a matrix.

[0061] A fiber orientation can be the arrangement direction of fibers in a composite material.

[0062] A matrix property can be a performance characteristic of a matrix material in a composite material.

[0063] A plastic liner can be a plastic layer inside a hydrogen cylinder.

[0064] A thickness can be a thickness dimension of a plastic liner.

[0065] A hydrogen permeation coefficient can be a parameter that describes the rate of hydrogen penetration through a material.

[0066] Specifically, by calling a data access interface, component information of the hydrogen cylinder is extracted. Then, regular expression matching is used to parse the component information to determine geometric parameters and component material properties. Subsequently, based on the geometric parameters and the component material properties, a material mechanics model constructed based on classical material mechanics theory is applied under the condition of simulating a hydrogen environment; further, a hydrogen environment influence factor extracted from a standard material manual is combined; then, a stress-strain curve is generated according to an interpolation method. Subsequently, the component material properties are parsed to identify attribute items of the composite material; then, a material test report is queried to extract fiber orientation and matrix properties. Subsequently, based on the geometric parameters and the component material properties, the material test report is queried to determine the thickness and the hydrogen permeation coefficient of the plastic liner through data calculation. Further, based on the geometric parameters, a three-dimensional geometric structure of the hydrogen cylinder is constructed according to a parameterized modeling device; finally, the stress-strain curve is applied to set the mechanical behavior, the fiber orientation and the matrix property are applied to set the behavior of the composite material layer, and the thickness and the hydrogen permeation coefficient are applied to set the behavior of the plastic liner, and the simulation engine is integrated to generate a preset three-dimensional twin model.

[0067] By the scheme, the component information of the hydrogen cylinder is acquired, and complete initial information support is ensured for the construction process of the preset three-dimensional twin model. The component information is analyzed, the component geometric parameters and component material properties are determined, accurate input parameters are provided for the construction of the stress-strain curve, data ambiguity is eliminated, and parameterized modeling is supported. According to the component geometric parameters and the component material properties, the stress-strain curve in the hydrogen environment is constructed, and the actual deformation characteristics in the hydrogen environment are reflected. According to the component material properties, the fiber orientation and matrix performance of the composite material are determined, and the failure behavior of the composite material is accurately characterized. According to the component geometric parameters and the component material properties, the thickness and hydrogen permeation coefficient of the plastic liner are determined, and the preset three-dimensional twin model is accurately simulated to ensure the physical and chemical behavior of the liner. According to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient, the preset three-dimensional twin model is generated, and failure simulation and strategy optimization are supported.

[0068] In some embodiments, according to the stress-strain curve, the hydrogen embrittlement sensitivity is determined; the guided wave monitoring data under the cyclic load is acquired; based on the fiber orientation and the matrix performance, the debonding probability of the composite layer is determined according to the guided wave monitoring data under the cyclic load; the acoustic emission signal of the plastic liner is acquired; based on the thickness and the hydrogen permeation coefficient, the acoustic emission signal is analyzed to determine the bulge collapse critical value; and according to the hydrogen embrittlement sensitivity, the debonding probability and the bulge collapse critical value, the preset three-dimensional twin model is generated.

[0069] The hydrogen embrittlement sensitivity can be a numerical index quantifying the susceptibility of the metal component to hydrogen embrittlement failure in the hydrogen environment.

[0070] The cyclic load can be a load condition of repeatedly applying mechanical stress.

[0071] The guided wave monitoring data can be signal data such as wave speed, attenuation coefficient and reflection characteristics of guided wave monitoring.

[0072] The composite layer can be a composite material layer structure in the hydrogen cylinder.

[0073] The debonding probability can be a probability value indicating the interlayer separation of the composite layer.

[0074] The acoustic emission signal can be acoustic signal data such as event count, amplitude and frequency spectrum.

[0075] The bulge collapse critical value can be a critical pressure threshold value for the plastic liner to occur bulge or collapse failure.

[0076] Specifically, according to the stress-strain curve, the mechanical parameters such as yield strength and elongation at break are extracted; then, the deformation characteristics of the stress-strain curve in the hydrogen environment are calculated to determine the hydrogen embrittlement sensitivity. Subsequently, according to the guided wave sensor of the hydrogen cylinder composite layer, the guided wave signal data is collected in real time under cyclic loading. Further, based on the fiber orientation and matrix performance, combined with the guided wave monitoring data under cyclic loading, the debonding probability of the composite layer is calculated by regression analysis. Then, according to the acoustic emission sensor on the surface of the plastic liner, the acoustic emission signals of the plastic liner are collected in real time under the running state of the hydrogen cylinder. Subsequently, the signal analysis algorithm is used to extract the features of the acoustic emission signals, and the critical stress threshold is calculated combined with the thickness and hydrogen permeation coefficient; according to the critical stress threshold, the critical value of the bulge collapse is determined. Finally, the hydrogen embrittlement sensitivity, debonding probability and bulge collapse critical value are integrated by the simulation engine to automatically generate a preset three-dimensional twin model.

[0077] According to the stress-strain curve, the hydrogen embrittlement sensitivity is determined to improve the prediction accuracy of hydrogen embrittlement failure. The guided wave monitoring data under cyclic loading is obtained to ensure the response to dynamic working condition changes. Based on the fiber orientation and matrix performance, the debonding probability of the composite layer is determined according to the guided wave monitoring data under cyclic loading to enhance the dynamic capture ability of the delamination failure. The acoustic emission signals of the plastic liner are obtained to reflect the actual degradation process of the plastic liner. Based on the thickness and hydrogen permeation coefficient, the acoustic emission signals are analyzed to determine the critical value of the bulge collapse to improve the prediction reliability of the hydrogen permeation leading to the instability of the liner. According to the hydrogen embrittlement sensitivity, the debonding probability and the bulge collapse critical value, a preset three-dimensional twin model is generated to eliminate the defects of insufficient modeling integrity and weak decision support, support failure mode detail simulation and real-time strategy optimization.

[0078] In some embodiments, historical failure cases are obtained; the historical failure cases are analyzed to determine historical damage modes; the preset knowledge graph is analyzed to determine node association; the coupling effect between the composite layer and the metal layer is determined according to the node association; the visual detection data is obtained, the visual detection data is analyzed, and the wear condition of the key weak area is determined; the potential failure mode of the hydrogen cylinder is predicted according to the wear condition, the coupling effect and the historical damage mode.

[0079] The historical failure cases can be failure event records of the hydrogen cylinder during historical transportation.

[0080] The historical damage mode can be a structured mode formed by failure feature classification.

[0081] The node association can be a quantitative indicator of the mutual influence strength between nodes.

[0082] The metal layer can be a metal component in the hydrogen cylinder.

[0083] The coupling effect can be a dynamic behavior description of the physical interaction between the composite layer and the metal layer.

[0084] The visual detection data can be hydrogen cylinder surface image data.

[0085] The wear condition can be a quantitative description of the surface damage state of the critical weak area.

[0086] Specifically, historical failure cases are retrieved from a pre-stored database. Then, the historical failure case data is parsed to extract failure characteristics such as failure type, location, and severity; further, the failure characteristics are integrated to determine the historical damage pattern. Subsequently, based on a pre-set knowledge graph, a graph traversal algorithm is executed to analyze the relationship between nodes, thereby determining the node correlation. Then, the node correlation is used to identify the correlation between the composite layer and the metal layer; further, according to the correlation, the physical interaction is deduced to determine the coupling effect between the composite layer and the metal layer. Then, visual detection data is collected through a visual sensor; further, the visual detection data is processed to identify the visual characteristics of the critical weak area; subsequently, the wear condition of the critical weak area is quantified according to the visual characteristics. Finally, the reasoning engine of the pre-set knowledge graph is applied to match the wear condition, the coupling effect, and the historical damage pattern; then, the failure evolution is simulated to predict the potential failure mode of the hydrogen cylinder.

[0087] Through the scheme, historical failure cases are obtained to avoid reasoning in the air, thereby enhancing the empirical nature and relevance of the prediction. By parsing the historical failure cases and determining the historical damage pattern, the accuracy and consistency of the prediction are improved. By analyzing the pre-set knowledge graph and determining the node correlation, it is ensured that the prediction process captures the hidden failure correlation path. According to the node correlation, the coupling effect between the composite layer and the metal layer is determined to clarify the cross-material failure transmission path. By obtaining the visual detection data and analyzing the visual detection data, the wear condition of the critical weak area is determined to reflect the contribution of actual use traces to failure. According to the wear condition, the coupling effect, and the historical damage pattern, the potential failure mode of the hydrogen cylinder is predicted, which is associated with the operable failure type and location.

[0088] In some embodiments, vibration spectrum data of the current transportation route is obtained; the vibration spectrum data is analyzed to determine the resonance frequency coincidence degree of the vibration spectrum and the hydrogen embrittlement sensitivity; and the detour instruction is constructed according to the resonance frequency coincidence degree, and the detour instruction is determined as the transportation strategy.

[0089] The current transportation route can be the path actually traveled by the hydrogen cylinder transportation vehicle.

[0090] The vibration spectrum data can be data representing the vibration characteristics of the transportation route.

[0091] The vibration spectrum can be a frequency distribution characteristic for describing the core frequency component of vibration.

[0092] The resonance frequency coincidence degree can be a quantitative indicator of the matching degree between the vibration spectrum and the resonance frequency of hydrogen embrittlement sensitivity.

[0093] The detour instruction can be an action command for avoiding a high-risk road section or selecting an alternative path.

[0094] Specifically, vibration spectrum data of the current transportation route is collected by a data collection device. Then, the vibration spectrum of the vibration spectrum data is compared with the resonance frequency of hydrogen embrittlement sensitivity; subsequently, the matching degree of the vibration spectrum and the hydrogen embrittlement sensitivity is calculated; further, the resonance frequency and the matching degree are integrated to determine the resonance frequency coincidence degree. Then, a preset risk threshold is set according to statistical analysis of historical failure cases; subsequently, it is judged whether the resonance frequency coincidence degree exceeds the preset risk threshold; if the resonance frequency coincidence degree exceeds the preset risk threshold, a detour instruction is generated; finally, the detour instruction is transmitted to the transportation management device, thereby generating a transportation strategy.

[0095] By the scheme, vibration spectrum data of the current transportation route is obtained, potential vibration sources causing hydrogen embrittlement of metal parts are identified, and the defect of weak decision support is eliminated. The vibration spectrum data is analyzed to determine the resonance frequency coincidence degree of the vibration spectrum and the hydrogen embrittlement sensitivity, the deficiency that the vibration spectrum data cannot be dynamically associated with hydrogen embrittlement sensitivity is responded to, the quantitative identification of resonance risk is realized, and an operable basis for generating a detour instruction is provided. According to the resonance frequency coincidence degree, a detour instruction is constructed, the detour instruction is determined as a transportation strategy, high-risk scenarios caused by vibration resonance are actively avoided, the problem that high-risk scenarios cannot be actively avoided during transportation is eliminated, and the initiative of risk prevention and control is improved.

[0096] In some embodiments, a transportation task of a hydrogen cylinder is obtained; the transportation task is analyzed to determine a backup route; route information of the backup route is obtained; the route information is parsed to determine road surface flatness and slope data; the road surface flatness and slope data are analyzed to determine real-time influence on the hydrogen cylinder; and a detour instruction is constructed according to the real-time influence and the resonance frequency coincidence degree.

[0097] The transportation task can be a transportation plan of the hydrogen cylinder from a starting point to an ending point.

[0098] The backup route can be a candidate alternative path for avoiding a high-risk area.

[0099] The route information can be attribute data of the backup route, such as road type, length, real-time traffic state, historical accident data, and environmental factors.

[0100] The road surface flatness can be a quantitative indicator of the roughness of the road surface.

[0101] The slope data can be the inclination of the road.

[0102] The real-time influence can be a dynamic effect of the road flatness and slope data on the hydrogen cylinder.

[0103] Specifically, the transportation task of the hydrogen cylinder is received through an application program interface of the transportation management device. Then, the transportation task is processed using a path planning algorithm, and a backup route is generated in combination with historical route data in a preset knowledge graph. Subsequently, route information of the backup route is queried through a map service API. Then, based on the route information, the road flatness is determined by analyzing historical vibration spectrum data, and the slope data is obtained from laser scanning data. Then, based on real-time sensor data, a physical rule model constructed based on the material characteristics of the hydrogen cylinder is applied to analyze the vibration amplitude caused by the road flatness and the pressure change caused by the slope data, thereby determining the real-time influence on the hydrogen cylinder. Then, the real-time influence is integrated with the resonance frequency coincidence degree; finally, a decision logic model constructed by a rule engine of the preset knowledge graph is used to determine whether the integrated value exceeds a preset risk threshold; if the preset risk threshold is exceeded, a detour instruction is generated Through the scheme, the transportation task of the hydrogen cylinder is obtained, ensuring that the risk assessment process is started based on actual logistics demand. The transportation task is analyzed to determine a backup route, avoiding providing viable alternative path options to ensure the flexibility of the transportation strategy. The route information of the backup route is obtained to enrich the route data, enabling the digital twin model to accurately map the influence of the external environment on the hydrogen cylinder. The route information is analyzed to determine the road flatness and slope data, and the potential effect of the road conditions on the hydrogen cylinder is evaluated. The road flatness and slope data are analyzed to determine the real-time influence on the hydrogen cylinder, providing a basis for decision-making to ensure that the risk prediction is real-time associated with the current route conditions. Based on the real-time influence and the resonance frequency coincidence degree, a detour instruction is constructed to dynamically optimize the transportation strategy, ensuring that high-risk scenarios are actively avoided.

[0104] In some embodiments, based on historical failure cases, the preset knowledge graph is analyzed to determine associated nodes; stress distribution data of the hydrogen cylinder is obtained; the stress distribution data is matched with the historical failure cases, and based on the matching result, a load similarity is determined; based on the similarity, an association weight between any associated nodes is determined; and based on the association weight, the node association is determined.

[0105] The associated nodes can be entity nodes and associated relationships related to the historical failure cases in the preset knowledge graph.

[0106] The stress distribution data can be stress value distribution data of each part of the hydrogen cylinder.

[0107] The matching result can be a similarity score output after matching the stress distribution data with the historical failure cases.

[0108] The load similarity can be a numerical indicator quantifying the similarity degree of the current stress load and the stress load of the historical failure cases.

[0109] The association weight can be a numerical weight representing the relationship strength between any associated nodes.

[0110] Specifically, according to the type of historical failure cases, the graph traversal algorithm is used to match the nodes stored in the preset knowledge graph, and the entity nodes and association relationships are extracted to determine the associated nodes. Then, the application programming interface of the digital twin model constructed by the virtual mapping theory of physical entities reads the stress distribution data of the hydrogen cylinder in real time. Subsequently, the stress feature vector of the historical failure case is extracted, and the similarity calculation algorithm is used to match the stress distribution data with the historical failure case; then, according to the matching result, the load similarity is determined. Further, based on the load similarity, the weight update rule is applied to adjust the association weight between any associated nodes in the preset knowledge graph. Finally, the association weight is processed according to the weighted average algorithm to determine the node association.

[0111] Through the scheme, based on historical failure cases, the preset knowledge graph is analyzed to determine associated nodes, eliminate the defect of insufficient modeling integrity, and ensure that nodes such as metal component hydrogen embrittlement sensitivity and composite layer risk are included in the analysis. The stress distribution data of the hydrogen cylinder is obtained to make up for the defect of being unable to accurately capture thermal stress failure. The stress distribution data is matched with the historical failure case, the load similarity is determined according to the matching result, and the defect of lacking accurate simulation of failure details is eliminated. According to the similarity, the association weight between any associated nodes is determined to eliminate the problem of missing group risk modeling, and to support group risk modeling. According to the association weight, the node association is determined to respond to the defect of weak decision support, and to provide association basis for transportation strategy.

[0112] In some embodiments, the transportation task is analyzed to determine the task time requirement; the detour instruction is analyzed to determine the predicted delay time of the detour path; the safety gain is determined according to the real-time influence; and the necessity of detour is determined based on the task time requirement, the predicted delay time and the safety gain.

[0113] The task time requirement can be a time constraint requirement of the transportation task.

[0114] The detour path can be an alternative transportation route recommended to avoid risks.

[0115] The predicted delay time can be a predicted delay time relative to the original planned path.

[0116] The safety gain can be the degree of risk reduction or the quantitative value of safety improvement.

[0117] The necessity of detour can be a decision conclusion of whether to execute the detour instruction.

[0118] Specifically, the database of the transportation task is accessed to extract the planned completion time or deadline of the transportation task as the task time limit requirement. Then, the travel time of the new path is estimated by time calculation according to the path coordinates in the detour instruction and the road condition parameters; subsequently, the expected delay time of the detour path is determined by comparing with the reference time of the original path. Based on the real-time influence, the safety gain is calculated by the safety evaluation algorithm. Finally, the task time limit requirement, the expected delay time and the safety gain are integrated, and the decision algorithm is used to evaluate the necessity of detour; if the necessity of detour is "necessary", the detour instruction is determined as the transportation strategy.

[0119] Through the scheme, the transportation task is analyzed to determine the task time limit requirement, ensuring that the evaluation is carried out within the task time frame, avoiding decision deviation caused by ignoring the time limit constraint. The detour instruction is analyzed to determine the expected delay time of the detour path, quantifying the time cost of the detour strategy, providing a basis for weighing the safety benefit and time loss. According to the real-time influence, the safety gain is determined to reflect the safety improvement degree of the detour strategy, providing a measurable risk mitigation benefit for decision-making. Based on the task time limit requirement, the necessity of detour is determined according to the expected delay time and the safety gain, realizing the dynamic optimization of the transportation strategy, and ensuring to actively avoid high-risk scenarios under the premise of meeting the time limit.

[0120] In some embodiments, real-time temperature data of the plastic liner is obtained; the temperature data is analyzed to determine the permeation influence of temperature on the hydrogen permeation coefficient; and the bulge collapse critical value is updated according to the permeation influence.

[0121] The real-time temperature data can be the current temperature value of the surface of the plastic liner of the hydrogen cylinder.

[0122] The permeation influence can be the quantitative effect of temperature change on the hydrogen permeation coefficient.

[0123] Specifically, the real-time temperature data of the plastic liner is obtained through the temperature sensor on the surface of the plastic liner of the hydrogen cylinder; then, the hydrogen permeation coefficient-temperature relationship model of the pre-stored plastic liner material based on experimental data is called, the real-time temperature data is input into the hydrogen permeation coefficient-temperature relationship model, and the permeation influence of the hydrogen permeation coefficient at the current temperature is calculated. Further, the hydrogen permeation coefficient is adjusted according to the permeation influence; then, the thickness of the plastic liner and the acoustic emission signal are combined; further, the physical parameters mapped by the digital twin model and the historical failure cases are integrated based on the pre-set knowledge graph to perform theoretical reasoning, thereby constructing a bulge collapse critical value calculation model; finally, the adjusted hydrogen permeation coefficient, the thickness and the acoustic emission signal are input into the bulge collapse critical value calculation model to update the bulge collapse critical value.

[0124] By the scheme, real-time temperature data of the plastic liner is acquired, and correction of the bulge collapse critical value is ensured to be based on actual environmental conditions. Temperature data is analyzed to determine the permeation influence of temperature on the hydrogen permeation coefficient, and calculation deviation of the permeation rate caused by temperature fluctuation is eliminated. According to the permeation influence, the bulge collapse critical value is updated, and the prediction accuracy of the plastic liner bulge collapse is improved.

[0125] Figure 3 A structural schematic diagram of a dynamic risk analysis system based on a knowledge graph and a digital twin model provided for an embodiment of the present application is shown in Figure 3 The dynamic risk analysis system based on the knowledge graph and the digital twin model 300 of the embodiment includes a model acquisition module 301, a region determination module 302, a mode prediction module 303, and a transportation determination module 304.

[0126] The model acquisition module 301 is configured to acquire a preset three-dimensional twin model of a hydrogen cylinder. The region determination module 302 is configured to determine a key weak region of the hydrogen cylinder according to the preset three-dimensional twin model. The mode prediction module 303 is configured to predict a potential failure mode of the hydrogen cylinder based on a preset knowledge graph and the key weak region. The transportation determination module 304 is configured to simulate a risk evolution state of the hydrogen cylinder according to the failure mode, and determine a transportation strategy according to a simulation result.

[0127] Optionally, when the model acquisition module 301 acquires the preset three-dimensional twin model of the hydrogen cylinder, the model acquisition module 301 is configured to: acquire component information of the hydrogen cylinder, analyze the component information to determine component geometric parameters and component material properties; construct a stress-strain curve under a hydrogen environment according to the component geometric parameters and the component material properties; determine fiber orientation and matrix performance of a composite material according to the component material properties; determine a thickness of a plastic liner and a hydrogen permeation coefficient according to the component geometric parameters and the component material properties; generate the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness, and the hydrogen permeation coefficient.

[0128] Optionally, when the model acquisition module 301 generates the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness, and the hydrogen permeation coefficient, the model acquisition module 301 is configured to: determine hydrogen embrittlement sensitivity according to the stress-strain curve; acquire guided wave monitoring data under cyclic load; determine a debonding probability of the composite layer according to the guided wave monitoring data under the cyclic load based on the fiber orientation and the matrix property; acquire an acoustic emission signal of the plastic liner; analyze the acoustic emission signal to determine a bulge collapse critical value based on the thickness and the hydrogen permeation coefficient; generate the preset three-dimensional twin model according to the hydrogen embrittlement sensitivity, the debonding probability, and the bulge collapse critical value.

[0129] Optionally, when the mode prediction module 303 predicts the potential failure mode of the hydrogen cylinder based on the preset knowledge graph according to the key weak area, it is used for: acquire historical failure cases; analyze the historical failure cases to determine historical damage modes; analyze the preset knowledge graph to determine node relevance; determine the coupling effect between the composite layer and the metal layer according to the node relevance; acquire visual inspection data, analyze the visual inspection data, and determine the wear condition of the key weak area; predict the potential failure mode of the hydrogen cylinder according to the wear condition, the coupling effect, and the historical damage mode.

[0130] Optionally, when the transportation determination module 304 determines a transportation strategy according to the simulation result, it is used for: acquire vibration spectrum data of a current transportation route; analyze the vibration spectrum data to determine a resonance frequency overlap degree of a vibration spectrum and the hydrogen embrittlement sensitivity; construct a detour instruction according to the resonance frequency overlap degree, and determine the detour instruction as a transportation strategy.

[0131] Optionally, when the transportation determination module 304 constructs a detour instruction according to the resonance frequency overlap degree, it is used for: acquire a transportation task of the hydrogen cylinder; analyze the transportation task to determine a backup route; acquire route information of the backup route, analyze the route information to determine road flatness and slope data; analyze the road flatness and the slope data to determine a real-time impact on the hydrogen cylinder; construct a detour instruction according to the real-time impact and the resonance frequency overlap degree.

[0132] Optionally, when the mode prediction module 303 analyzes the preset knowledge graph to determine node relevance, it is used for: Based on the historical failure cases, the preset knowledge graph is analyzed to determine the associated nodes; Obtain stress distribution data of the hydrogen cylinder; Match the stress distribution data with the historical failure cases, and determine the load similarity according to the matching result; According to the similarity, determine the association weight between any associated nodes; According to the association weight, determine the node association.

[0133] Optionally, the dynamic risk analysis system based on knowledge graph and digital twin model further includes a necessity determination module 305, configured to: Analyze the transportation task to determine the task time requirement; Analyze the detour instruction to determine the expected delay time of the detour path; According to the real-time influence, determine the safety gain; Based on the task time requirement, determine the necessity of detour according to the expected delay time and the safety gain.

[0134] Optionally, the dynamic risk analysis system based on knowledge graph and digital twin model further includes a critical value updating module 306, configured to: Obtain real-time temperature data of the plastic liner; Analyze the temperature data to determine the permeation influence of temperature on the hydrogen permeation coefficient; According to the permeation influence, update the bulge collapse critical value.

[0135] The system of the embodiment can be used to execute the method of any of the above embodiments, and has similar implementation principles and technical effects, which will not be described here.

Claims

1. A dynamic risk analysis method based on a knowledge graph and a digital twin model, characterized in that, The method comprises the following steps: acquiring a preset three-dimensional twin model of a hydrogen cylinder; determining a key weak area of the hydrogen cylinder according to the preset three-dimensional twin model; predicting a potential failure mode of the hydrogen cylinder according to the key weak area based on a preset knowledge graph; determining a transportation strategy according to the failure mode by simulating a risk evolution state of the hydrogen cylinder.

2. The method of claim 1, wherein, The method of acquiring the preset three-dimensional twin model of the hydrogen cylinder comprises the following steps: acquiring component information of the hydrogen cylinder; analyzing the component information to determine component geometric parameters and component material properties; constructing a stress-strain curve under a hydrogen environment according to the component geometric parameters and the component material properties; determining fiber orientation and matrix performance of a composite material according to the component material properties; determining the thickness of a plastic liner and the hydrogen permeation coefficient according to the component geometric parameters and the component material properties; generating the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient.

3. The method of claim 2, wherein, The method of generating the preset three-dimensional twin model according to the stress-strain curve, the fiber orientation, the matrix performance, the thickness and the hydrogen permeation coefficient comprises the following steps: determining hydrogen embrittlement sensitivity according to the stress-strain curve; acquiring guided wave monitoring data under cyclic loading; determining the debonding probability of the composite layer according to the guided wave monitoring data under cyclic loading based on the fiber orientation and the matrix performance; acquiring acoustic emission signals of the plastic liner; analyzing the acoustic emission signals to determine the critical value of bulge collapse based on the thickness and the hydrogen permeation coefficient; generating the preset three-dimensional twin model according to the hydrogen embrittlement sensitivity, the debonding probability and the critical value of bulge collapse.

4. The method of claim 1, wherein, The method of predicting the potential failure mode of the hydrogen cylinder according to the key weak area based on the preset knowledge graph comprises the following steps: acquiring historical failure cases; analyzing the historical failure cases to determine historical damage modes; analyzing the preset knowledge graph to determine node correlation; determining the coupling effect between the composite layer and the metal layer according to the node correlation; acquiring visual inspection data, analyzing the visual inspection data to determine the wear condition of the key weak area; predicting the potential failure mode of the hydrogen cylinder according to the wear condition, the coupling effect and the historical damage mode.

5. The method of claim 3, wherein, The method of determining the transportation strategy according to the simulation result comprises the following steps: acquiring vibration spectrum data of the current transportation route; analyzing the vibration spectrum data to determine the resonance frequency coincidence degree of the vibration spectrum and the hydrogen embrittlement sensitivity; constructing a detour instruction according to the resonance frequency coincidence degree, and determining the detour instruction as the transportation strategy.

6. The method of claim 5, wherein, The method of constructing the detour instruction according to the resonance frequency coincidence degree comprises the following steps: acquiring a transportation task of the hydrogen cylinder; analyzing the transportation task to determine a standby route; acquiring route information of the standby route; analyzing the route information to determine road flatness and slope data; analyzing the road flatness and the slope data to determine the real-time influence on the hydrogen cylinder; constructing a detour instruction according to the real-time influence and the resonance frequency coincidence degree.

7. The method of claim 4, wherein, The analysis of the preset knowledge graph determines the node relevance, including: Based on the historical failure cases, analyze the preset knowledge graph to determine the associated nodes; Obtain the stress distribution data of the hydrogen cylinder; Match the stress distribution data with the historical failure cases, and determine the load similarity according to the matching result; According to the similarity, determine the association weight between any associated nodes; According to the association weight, determine the node relevance.

8. The method of claim 6, wherein, After constructing the bypass instruction according to the real-time influence and the resonance frequency coincidence degree, it further includes: Analyze the transportation task to determine the task time limit requirement; Analyze the bypass instruction to determine the expected delay time of the bypass path; According to the real-time influence, determine the safety gain; Based on the task time limit requirement, according to the expected delay time and the safety gain, determine the necessity of bypass.

9. The method of claim 3, wherein, After analyzing the acoustic emission signal based on the thickness and the hydrogen permeation coefficient to determine the bulge collapse critical value, it further includes: Obtain the real-time temperature data of the plastic liner; Analyze the temperature data to determine the permeation influence of temperature on the hydrogen permeation coefficient; According to the permeation influence, update the bulge collapse critical value. 10.A dynamic risk analysis system based on a knowledge graph and a digital twin model, characterized in that, Applied to the method of any one of claims 1-9, including: Model acquisition module, for acquiring the preset three-dimensional twin model of hydrogen cylinder; Region determination module, for determining the key weak area of the hydrogen cylinder according to the preset three-dimensional twin model; Pattern prediction module, for predicting the potential failure mode of the hydrogen cylinder based on the preset knowledge graph according to the key weak area; Transportation determination module, for simulating the risk evolution state of the hydrogen cylinder according to the failure mode, and determining the transportation strategy according to the simulation result.

Images