Electricity-gas coupling system reliability evaluation method considering cascading failure buffer time

By establishing a buffer time evaluation model and martingale theory based on fluid dynamics, combined with Bayes theorem, considering the buffer time and repair time when the chain failure of the electric-gas coupled system propagates across the system, the electric-gas in the existing technology is solved. The coupling system chain fault evaluation method failed to consider the problem of time scale mismatch, achieving more accurate reliability evaluation.

CN118886261APending Publication Date: 2024-11-01TAIYUAN UNIVERSITY OF TECHNOLOGY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410939970.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-15
Publication Date
2024-11-01

AI Technical Summary

Technical Problem

The existing method of chain fault reliability evaluation of electrical-gas coupling systems fails to fully consider the problem of mismatch between the time scale of the power and natural gas systems, resulting in deviations from the actual physical process, making it difficult to accurately evaluate the chain fault of electrical-gas coupling systems. Dynamic process.

Method used

By establishing a buffer time evaluation model based on fluid dynamics equations, using finite element software to simulate buffer time under different operating conditions, and combining martingale theory and Bayes theorem, a chain fault model for electrical-gas coupling system is constructed to consider chain faults. Buffer time and repair time when propagating across systems, improving reliability evaluation methods.

Benefits of technology

This method can more accurately analyze the reliability of chain faults of the electrical-gas coupling system, consider the time intervals and repair situations between faults, provide more practical evaluation results, and improve the accuracy and practical value of the evaluation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118886261A_ABST
    Figure CN118886261A_ABST
Patent Text Reader

Abstract

The invention discloses an electricity-gas coupling system reliability evaluation method considering cascading failure buffer time, and relates to the field of cascading failure reliability evaluation of electricity-gas coupling systems. The method comprises the following steps: firstly, proposing a concept of buffer time, and describing a cascading failure process as an intermittent progressive increase process; and then the buffer time is solved by utilizing finite element software. And then, based on the application of a yoke theory in traditional reliability evaluation, buffer time and repair time are incorporated into the yoke theory, and the cascading failure process of the electric-gas coupling system is described by using the improved yoke theory. And finally, based on a plurality of theorems and an improved yoke expression, establishing an electricity-gas coupling system cascading failure reliability evaluation model, and calculating the influence of the model on the cascading failure occurrence probability from the angles of buffer time and repair time. The electricity-gas coupling system cascading failure reliability evaluation model constructed by the method is closer to engineering practice, and reference value can be provided for improving the electricity-gas coupling system cascading failure reliability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of reliability assessment of cascading failures of electric-pneumatic coupling systems, and in particular to a reliability assessment method of electric-pneumatic coupling systems taking into account a buffer time of cascading failures. Background Art

[0002] As a new type of energy system, the electric-gas coupling system plays an important role in promoting the efficient use of clean energy and reducing carbon emissions. However, the deep coupling between energy sources has led to an increasing number of scenarios in which faults propagate across systems, which not only increases the probability of cascading failures, but also expands the scope of impact of cascading failures due to the dependencies between systems, making the reliable operation of the electric-gas coupling system face huge challenges. On the other hand, due to the different transmission speeds of electricity and natural gas, there are differences in the time scales for studying cascading failures in power systems and natural gas systems, which makes it difficult to analyze the dynamic process of cascading failures in electric-gas coupling systems. At present, the reliability assessment of cascading failures in electric-gas coupling systems usually follows traditional methods. However, traditional methods fail to fully consider the characteristics of the mismatched time scales of the two subsystems in the electric-gas coupling system, resulting in a certain deviation between the model description and the actual physical process. Therefore, establishing an accurate and effective reliability assessment model for cascading failures in electric-gas coupling systems is the key to supporting the safe and reliable operation of electric-gas coupling systems. Summary of the invention

[0003] In order to solve the problem that the existing reliability assessment method for cascading failures of an electric-pneumatic coupling system fails to consider the time scale mismatch between two subsystems, the present invention provides a reliability assessment method for an electric-pneumatic coupling system considering the buffer time of cascading failures.

[0004] Since the time scale of the dynamic process of the natural gas system is much smaller than that of the power system, when a fault propagates across the system, the process of the cascading failure of the electric-gas coupling system has obvious intermittency. The existing reliability assessment of cascading failures in the electric-gas coupling system usually follows the method in the traditional power system, that is, the cascading failures are equivalent to simultaneous failures. However, the intermittent nature of the cascading failures in the electric-gas coupling system indicates that the system can take effective measures to prevent its further development during the cross-system propagation of the fault, so the simultaneous failure model is not applicable to the electric-gas coupling system. Based on this, the present invention takes into account the buffer time when the cascading failure propagates across the system, which is used to evaluate the reliability of the cascading failures of the electric-gas coupling system.

[0005] The present invention is implemented by the following technical solution: a reliability assessment method for an electric-pneumatic coupling system considering the buffer time of cascading failures, comprising the following steps:

[0006] S1: Based on the fluid dynamics equation, a buffer time evaluation model for the electric-pneumatic coupling system is established; the simulation data of the buffer time is obtained using finite element software; the details are as follows:

[0007] S1-1: To calculate the buffer time, it is necessary to establish an air pressure analysis model to obtain the law of air pressure change relative to time; according to the relevant theories of fluid dynamics, the dynamic expression of air pressure is obtained by solving the continuity equation, momentum conservation equation and state equation:

[0008]

[0009] In the formula, ρ and v are the density and flow velocity of the gas, t is the time, x is any position in the pipeline, p is the gas pressure, λ is the friction coefficient, D is the diameter of the pipeline, and c is the isothermal sound velocity of the gas;

[0010] Arrange formulas (1) to (3), eliminate density ρ and gas flow velocity v in the formula, and obtain the following formula:

[0011]

[0012] Since the variation of the gas pressure in the pipeline with time and position is described by a set of partial differential equations, the characteristic line method is used to solve the variation law of the gas pressure with time; using the idea of ​​difference, the time and position of the pipeline are divided into several segments, each with a length of Δt and Δx, respectively. The points on the pipeline are set as A, B, C, ..., at t = 0, if a cascading failure occurs from the natural gas system to the power system, the pipeline gas pressure will drop; since the gas pressure and mass flow rate cannot change suddenly, the gas pressure and mass flow rate of each point A, B, C, ... on the pipeline at t = 0 are the steady-state values ​​before the failure occurs, which are recorded as p A 0 ,p B 0 ,p C 0 ,… and G A 0 ,G B 0 ,G C 0 ,…; After that, affected by the fault, the air pressure and mass flow rate in the pipeline change dynamically. After Δt time, the air pressure and mass flow rate at the same position but at different times become p A Δt ,p B Δt ,p C Δt ,… and G A Δt ,G B Δt ,G CΔt ,…; To obtain the air pressure and mass flow rate at each point after Δt time, use the air pressure p at points A and C at time t = 0 A 0 、p C 0 , solve for the pressure p at point B after Δt time B Δt ; At the same time, the mass flow rate G of points A and C at time t = 0 is used A 0 , G C 0 , solve for the mass flow rate G at point B after Δt time B Δt ;

[0013] Based on the information of points A, B and C at the time of fault onset, the differential forms of the partial differential terms and linear terms in formulas (4) and (5) are given as follows:

[0014]

[0015] Substituting formula (6) and (7) into formula (4) and (5) respectively, the differential expression is:

[0016]

[0017] The characteristic line method shows that if the values ​​of the differentiated physical quantities Δt and Δx are extremely small, the relationship between the two is approximated as a linear relationship. It can be seen that Δt and Δx satisfy the following relationship:

[0018] Δx=aΔt (10)

[0019] In the formula, a is a constant;

[0020] Substituting formula (10) into formula (8) and (9) respectively, we get:

[0021]

[0022] When the remaining parameters are known, use MATLAB loop iteration to find p B Δt and G B Δt Numerical solution at each point, and set to use only p B Δt The value of is used for subsequent analysis;

[0023] The above analysis shows how to use the difference method to solve the air pressure at a certain point along the pipeline after Δt time; Formulas (11) and (12) are used to calculate the air pressure p at each node along the pipeline after Δt time. A Δt ,pB Δt ,p C Δt ,…; Average pipeline pressure p after Δt time Δt for:

[0024]

[0025] Where b is the number of discrete segments of the pipeline;

[0026] S1-2: The calculation process of the buffer time of cascading faults propagating from the natural gas system to the power system is as follows: Δt and safety threshold p 1 , if p Δt >p 1 , it means that after Δt time, the pipeline is not faulty; then, continue to use the air pressure p A Δt ,p B Δt ,p C Δt ,…calculate the air pressure p after 2Δt A 2Δt ,p B 2Δt ,p C 2Δt ,…; and so on, after nΔt time, n is a natural number, if p nΔt With p 1 The absolute value of the difference is less than an infinitesimal quantity, i.e. |p nΔt -p 1 |≤ε, it means that after nΔt time, the pipeline gas pressure is infinitely close to the fault threshold; therefore, the buffer time of the cascading fault propagation from the natural gas system to the power system is:

[0027] T b =nΔt (14)

[0028] S1-3: Calculation process of buffer time of cascading failures propagating from power system to natural gas system: The buffer time of cascading failures is calculated based on the gas pressure p Δt and safety threshold p 2 Then calculate, after nΔt time, if p nΔt With p 1 The absolute value of the difference is less than an infinitesimal quantity, i.e. |p nΔt -p 2 |≤ε, indicating that after nΔt time, the pipeline gas pressure approaches the fault threshold infinitely; at this time, the differential calculation process is stopped and the buffer time is obtained.

[0029] S2: Derive the martingale theory with buffer time and repair time, and establish a cascading failure model for the electric-pneumatic coupling system; the details are as follows:

[0030] When a cascading failure occurs in the electric-pneumatic coupling system, the number of failures at the next moment is not less than the number of failures at the previous moment. Therefore, the number of failures in the electric-pneumatic coupling system meets the definition of a lower martingale. On this basis, according to the Doob-Meyer decomposition theorem, the change in the number of failures in the electric-pneumatic coupling system is uniquely decomposed into the following two parts:

[0031]

[0032] In the formula is the domain flow (Filtration), which contains the fault observation information before time t; non-decreasing predictable random process Observe information for a given fault When , the mathematical expectation of the change in the number of failures dN(t); dM(t) is a zero-mean martingale;

[0033] In the process of cascading failure propagation, one failure may cause multiple lower-level failures. b Less than the mean time to repair a fault T r When T is reached, the component is repaired in the fault buffer stage, and it is considered that the component has not failed, so it will not cause subsequent chain failures. b Greater than T r When the component fails to be repaired within the buffer time, the failure of the component will continue to cause changes in the operating conditions in the system, causing more components to participate in the chain failure; the subsequent faults at all levels will repeat the judgment process; using the indicative function I(T r -T b <0) indicates the result of a single component failure, I:T r -T b <0; then the number of faulty components in the system at time T is n T The expression is:

[0034]

[0035] Where N is the total number of system components, T b,j is the buffer time of the jth component in the gas grid; when a cascading failure occurs from the natural gas system to the power system, T r,j is the repair time of the jth component; when a cascading failure occurs from the power system to the natural gas system, T r,j is the repair time of the component that caused the jth component failure;

[0036] When the fault information of the known cascading fault before time T When , the change in the number of faults in the future period of time Δt is the fault rate ρ. Using the idea of ​​limit, the differential form of the fault rate is obtained as follows:

[0037]

[0038] Where N(t+Δt) and N(t) are the number of faulty components in the system at time t+Δt and t, respectively;

[0039] Substituting formula (17) into formula (15), the martingale expression of the number of faults is converted to:

[0040]

[0041] Using the buffer time T b and repair time T r The total number of failures n T The ratio of the time T from the beginning of the cascading failure to the present is used to calculate the historical failure rate of the system, which is equivalent to the future failure rate of the system. Thus, we have:

[0042]

[0043] Combining formula (18) and formula (19), we get the improved martingale expression considering buffer time and repair time as follows:

[0044]

[0045] S3: Based on the simulation data and fault model, the Bayesian theorem is used to calculate the probability of cascading failures in the electric-pneumatic coupling system, as follows:

[0046] Bayes' theorem defines the expression for the probability density function of a binary random variable:

[0047] p(x,y)=cp 0 (x,y)Λ (21)

[0048] Where (x, y) is a binary random variable; p 0 (x, y) is the prior probability density function of the binary random variable; Λ is the likelihood function; p(x, y) is the probability density function of the random variable; c is a constant;

[0049] S3-1: Constructing a priori probability density function p 0 (T b ,T r ):

[0050] Assume the buffer time T b Following the gamma distribution, the repair time is T r They obey exponential distribution, and their probability density function expressions are shown in formula (22):

[0051]

[0052] Where Г(·) is the gamma function; γ is the shape parameter of the gamma distribution; β is the scale parameter of the gamma distribution; μ is the repair rate;

[0053] T b and T r As two independent random variables, we get (T b ,T r )’s prior probability density function:

[0054]

[0055] In the formula, c 1 、c 2 are all constants.

[0056] S3-2: Construct likelihood function Λ:

[0057] According to the Radon-Nikodym theorem, for a random process with continuous time parameters and discrete state space, the likelihood function is expressed as follows:

[0058]

[0059] Where ρ is the failure rate; n s is the number of faulty components at time s;

[0060] Substituting the failure rate ρ into formula (24), we can obtain:

[0061]

[0062] Let α T =∑ j=1 N I(T b -T r <0), the specific form of the likelihood function is as follows:

[0063]

[0064] S3-3: The probability of a cascading failure P(T b ,T r )

[0065] Based on the determined prior probability density function p 0 (T b ,T r ) and likelihood function Λ, substitute formula (23) and (26) into formula (21), and the probability density function of cascading failure is obtained as:

[0066]

[0067] Therefore, the probability of cascading failures P(T b ,T r )for:

[0068]

[0069] Compared with the prior art, the present invention has the following beneficial effects: the reliability assessment method of an electric-pneumatic coupling system considering the buffer time of cascading failures provided by the present invention is applicable to the probability assessment of cascading failures of an electric-pneumatic coupling system in actual work, and has significant practical value in theory and application, as follows:

[0070] 1) In terms of theory, the present invention analyzes the mechanism of cross-system propagation of cascading failures, equates the cascading failures in the electric-pneumatic coupling system to an intermittent process, and uses the martingale theory to establish a cross-system propagation model of cascading failures. The intermittent process indicates that there is a certain time interval between failures. The present invention defines this time interval as the buffer time, and uses finite element software to simulate the buffer time under different working conditions. Different buffer times are used as the input of the likelihood function in the Bayesian theorem, and the probability density function of the cascading failures of the electric-pneumatic coupling system is obtained through theoretical deduction, which is used for the reliability assessment of the system. Compared with the traditional reliability assessment method, the present invention takes into account the time interval between failures, and regards whether the failures within the interval can be repaired as an important influencing factor on whether the cascading failures continue to occur, so that the assessment results are more in line with reality.

[0071] 2) In terms of application, the reliability assessment method proposed in the present invention can evaluate the reliability of cascading failures of different electric-pneumatic coupling systems according to actual conditions. This method can not only consider the specific buffer time in different systems to obtain the probability density of cascading failures that conforms to the actual situation of the system, but also select different prior probability density functions according to needs to calculate the probability of cascading failures. In addition, this method also obtains some general conclusions that can reduce the probability of cascading failures in the system, providing a reference for the reliable operation of the system. The cascading failure reliability assessment established by the present invention is a general model, and its extensibility further expands its scope of application. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 Schematic diagram of buffer time solution using characteristic line method.

[0073] Figure 2 Schematic diagram of the cascading fault buffering process propagating from the natural gas system to the power system.

[0074] Figure 3 Schematic diagram of the cascading fault buffering process propagating from the power system to the natural gas system.

[0075] Figure 4 is the probability distribution of buffer time and repair time.

[0076] Figure 5 Schematic diagram of the finite element model.

[0077] Figure 6 This is a schematic diagram of transient analysis.

[0078] Figure 7 This is a diagram showing the change in air pressure over time and location.

[0079] Figure 8 This is a diagram showing the change in average air pressure over time.

[0080] Fig. 9 Schematic diagram of the topology of the electric-gas coupling system.

[0081] Fig.10 This is the equivalent schematic diagram of the electric-gas coupling system.

[0082] Fig.11 Schematic diagram of simulation data for pipes with different diameters.

[0083] Fig.12 Schematic diagram of the average air pressure changing over time in this embodiment.

[0084] Fig.13 Schematic diagram of the probability density function of cascading failures. DETAILED DESCRIPTION

[0085] The present invention is further described below in conjunction with specific embodiments.

[0086] This embodiment is based on the improved IEEE 24-node power system and the Belgian 20-node natural gas system. Fig. 9 The example of the electric-gas coupling system shown. The IEEE 24-node power system has a total of 9 generator sets, of which buses 2 and 16 are gas-fired generator sets, connected to the Brugge and Liège nodes of the natural gas system respectively. The remaining 7 generator sets are coal-fired generator sets. In the Belgian 20-node natural gas system, Gent and Wanze are regarded as equivalent electric load nodes, connected to buses 4 and 20 in the power system respectively. Based on this electric-gas coupling system, the average buffer time is studied. and mean time to repair Impact on reliability of cascading failures.

[0087] The above-mentioned electric-pneumatic coupling system reliability assessment method considering the buffer time of cascading failures is used to calculate the present embodiment, including the following steps:

[0088] S1: Based on the fluid dynamics equation, a buffer time evaluation model for the electric-pneumatic coupling system is established; the simulation data of the buffer time is obtained using finite element software; the details are as follows:

[0089] To calculate the buffer time, it is necessary to establish an air pressure analysis model to obtain the law of air pressure change relative to time; according to the relevant theories of fluid dynamics, the dynamic expression of air pressure is obtained by solving the continuity equation, momentum conservation equation and state equation:

[0090]

[0091] In the formula, ρ and v are the density and flow velocity of the gas, t is the time, x is any position in the pipeline, p is the gas pressure, λ is the friction coefficient, D is the diameter of the pipeline, and c is the isothermal sound velocity of the gas;

[0092] Arrange formulas (1) to (3), eliminate density ρ and gas flow velocity v in the formula, and obtain the following formula:

[0093]

[0094]

[0095] Since the change of air pressure in the pipeline with time and position is described by a set of partial differential equations, the characteristic line method is used to solve the change law of air pressure with time; using the idea of ​​difference, the time and position of the pipeline are divided into several equal sections, each with a length of Δt and Δx, such as Figure 1 Assume that each point on the pipeline is A, B, C, ..., at t = 0, if a chain failure occurs from the natural gas system to the power system, the pipeline gas pressure will drop; since the gas pressure and mass flow rate cannot change suddenly, the gas pressure and mass flow rate of each point A, B, C, ... on the pipeline at t = 0 are the steady-state values ​​before the failure occurs, which are recorded as p A 0 ,p B 0 ,p C 0 ,… and G A 0 ,G B 0 ,G C 0 ,…; After that, affected by the fault, the air pressure and mass flow rate in the pipeline change dynamically. After Δt time, the air pressure and mass flow rate at the same position but at different times become p A Δt ,p B Δt ,p C Δt ,… and G AΔt ,G B Δt ,G C Δt ,…; To obtain the air pressure and mass flow rate at each point after Δt time, use the air pressure p at points A and C at time t = 0 A 0 、p C 0 , solve for the pressure p at point B after Δt time B Δt ; At the same time, the mass flow rate G of points A and C at time t = 0 is used A 0 , G C 0 , solve for the mass flow rate G at point B after Δt time B Δt ;

[0096] Based on the information of points A, B and C at the time of fault onset, the differential forms of the partial differential terms and linear terms in formulas (4) and (5) are given as follows:

[0097]

[0098] Substituting formula (6) and (7) into formula (4) and (5) respectively, the differential expression is:

[0099]

[0100]

[0101] The characteristic line method shows that if the values ​​of the differentiated physical quantities Δt and Δx are extremely small, the relationship between the two is approximated as a linear relationship. It can be seen that Δt and Δx satisfy the following relationship:

[0102] Δx=aΔt (10)

[0103] In the formula, a is a constant;

[0104] Substituting formula (10) into formula (8) and (9) respectively, we get:

[0105]

[0106] When the remaining parameters are known, use MATLAB loop iteration to find p B Δt and G B Δt Numerical solution at each point, and set to use only p B Δt The value of is used for subsequent analysis;

[0107] The above analysis shows how to use the difference method to solve the air pressure at a certain point along the pipeline after Δt time; Formulas (11) and (12) are used to calculate the air pressure p at each node along the pipeline after Δt time. A Δt ,p B Δt ,p C Δt ,…;like Figure 2 As shown, the average pipeline pressure p after Δt time Δt for:

[0108]

[0109] Where b is the number of discrete segments of the pipeline;

[0110] S1-2: The calculation process of the buffer time of cascading faults propagating from the natural gas system to the power system is as follows: Δt and safety threshold p 1 , if p Δt >p 1 , it means that after Δt time, the pipeline is not faulty; then, continue to use the air pressure p A Δt ,p B Δt ,p C Δt ,…calculate the air pressure p after 2Δt A 2Δt ,p B 2Δt ,p C 2Δt ,…; and so on, after nΔt time, n is a natural number, if p nΔt With p 1 The absolute value of the difference is less than an infinitesimal quantity, i.e. |p nΔt -p 1 |≤ε, it means that after nΔt time, the pipeline gas pressure is infinitely close to the fault threshold; therefore, the buffer time of the cascading fault propagation from the natural gas system to the power system is:

[0111] T b =nΔt (14)

[0112] S1-3: Calculation process of buffer time of cascading failures propagating from power system to natural gas system: The buffer time of cascading failures is calculated based on the gas pressure p Δt and safety threshold p 2 Then calculate, after nΔt time, if p nΔt With p 1 The absolute value of the difference is less than an infinitesimal quantity, i.e. |p nΔt -p2 |≤ε, indicating that after nΔt time, the pipeline gas pressure approaches the fault threshold infinitely; at this time, the differential calculation process is stopped and the buffer time is obtained.

[0113] S1-4: The finite element software ANSYS is used to simulate the pressure change process when a cascading failure occurs in the natural gas system to obtain experimental data on the buffer time.

[0114] S1-4-1: Use the Design Modeler built into ANSYS software to draw the natural gas pipeline model demonstrated in this section, such as Figure 5 shown.

[0115] S1-4-2: Further process the model, define the positive boundary of the y-axis as the gas inlet Inlet, the negative boundary of the y-axis as the gas outlet Outlet, and the outermost layer in the middle of the inlet and outlet as the pipe wall Wall. Use the software's built-in expansion method for meshing, and the parameter settings are shown in Table 1.

[0116] Table 1 Mesh partitioning parameter settings

[0117]

[0118] S1-4-3: Use ANSYS to perform transient simulation of air pressure. The changes of air pressure over time and position in the full time range of two types of cascading failures are shown in the following figure. Figure 6 shown.

[0119] S1-4-4: Extract the relationship between air pressure and time, such as Figure 7 Determine the upper and lower limits of the pipeline pressure and read the specific value of the buffer time from the diagram, as shown in Figure 8 shown.

[0120] S2: Derive the martingale theory with buffer time and repair time, and establish a cascading failure model for the electric-pneumatic coupling system; the details are as follows:

[0121] When a cascading failure occurs in the electric-pneumatic coupling system, the number of failures at the next moment is not less than the number of failures at the previous moment. Therefore, the number of failures in the electric-pneumatic coupling system meets the definition of a lower martingale. On this basis, according to the Doob-Meyer decomposition theorem, the change in the number of failures in the electric-pneumatic coupling system is uniquely decomposed into the following two parts:

[0122]

[0123] In the formula is the domain flow (Filtration), which contains the fault observation information before time t; non-decreasing predictable random process Observe information for a given fault When , the mathematical expectation of the change in the number of failures dN(t); dM(t) is a zero-mean martingale;

[0124] In the process of cascading failure propagation, one failure may cause multiple lower-level failures. b Less than the mean time to repair a fault T r When T is reached, the component is repaired in the fault buffer stage, and it is considered that the component has not failed, so it will not cause subsequent chain failures. b Greater than T r When the component fails to be repaired within the buffer time, the failure of the component will continue to cause changes in the operating conditions in the system, causing more components to participate in the chain failure; the subsequent faults at all levels will repeat the judgment process; using the indicative function I(T r -T b <0) indicates the result of a single component failure, I:T r -T b <0; then the number of faulty components in the system at time T is n T The expression is:

[0125]

[0126] Where N is the total number of system components, T b,j is the buffer time of the jth component in the gas grid; when a cascading failure occurs from the natural gas system to the power system, T r,j is the repair time of the jth component; when a cascading failure occurs from the power system to the natural gas system, T r,j is the repair time of the component that caused the jth component failure;

[0127] When the fault information of the known cascading fault before time T When , the change in the number of faults in the future period of time Δt is the fault rate ρ. Using the idea of ​​limit, the differential form of the fault rate is obtained as follows:

[0128]

[0129] Where N(t+Δt) and N(t) are the number of faulty components in the system at time t+Δt and t, respectively;

[0130] Substituting formula (17) into formula (15), the martingale expression of the number of faults is converted to:

[0131]

[0132] Using the buffer time T b and repair time T r The total number of failures n TThe ratio of the time T from the beginning of the cascading failure to the present is used to calculate the historical failure rate of the system, which is equivalent to the future failure rate of the system. Thus, we have:

[0133]

[0134] Combining formula (18) and formula (19), we get the improved martingale expression considering buffer time and repair time as follows:

[0135]

[0136] S3: Based on the simulation data and fault model, the Bayesian theorem is used to calculate the probability of cascading failures in the electric-pneumatic coupling system, as follows:

[0137] Bayes' theorem defines the expression for the probability density function of a binary random variable:

[0138] p(x,y)=cp 0 (x,y)Λ (21)

[0139] Where (x, y) is a binary random variable; p 0 (x, y) is the prior probability density function of the binary random variable; Λ is the likelihood function; p(x, y) is the probability density function of the random variable; c is a constant;

[0140] S3-1: Constructing a priori probability density function p 0 (T b ,T r ):

[0141] Assume the buffer time T b Following the gamma distribution, the repair time is T r They obey exponential distribution. The probability density function expressions of the two are shown in formula (22). The changing rules are as follows: Figure 4 As shown:

[0142]

[0143] Where Г(·) is the gamma function; γ is the shape parameter of the gamma distribution; β is the scale parameter of the gamma distribution; μ is the repair rate;

[0144] T b and T r As two independent random variables, we get (T b ,T r )’s prior probability density function:

[0145]

[0146] In the formula, c 1 、c2 are all constants.

[0147] S3-2: Construct likelihood function Λ:

[0148] According to the Radon-Nikodym theorem, for a random process with continuous time parameters and discrete state space, the likelihood function is expressed as follows:

[0149]

[0150] Where ρ is the failure rate; n s is the number of faulty components at time s;

[0151] Substituting the failure rate ρ into formula (24), we can obtain:

[0152]

[0153] Let α T =∑ j=1 N I(T b -T r <0), the specific form of the likelihood function is as follows:

[0154]

[0155] S3-3: The probability of a cascading failure P(T b ,T r )

[0156] Based on the determined prior probability density function p 0 (T b ,T r ) and likelihood function Λ, substitute formula (23) and (26) into formula (21), and the probability density function of cascading failure is obtained as:

[0157]

[0158] Therefore, the probability of cascading failures P(T b ,T r )for:

[0159]

[0160] This embodiment provides the following assumptions from the perspective of reliability assessment of the electric-pneumatic coupling system:

[0161] 1) The system is equivalent to three parts, namely the power system part, the natural gas system part and the coupling part, such as Fig.10As shown. It should be noted that the coupling part is actually a group of natural gas pipelines. According to the analysis of the present invention, the buffer time is related to the gas pressure change process caused by the failure of the components in the power system or the natural gas system. The coupling part in the system needs to be equivalent to the components related to the gas pressure to characterize this characteristic. Therefore, the natural gas pipeline is used to represent the coupling part in the system.

[0162] 2) Currently, natural gas pipeline failure models are divided into two categories: one is the pipeline leakage rate model, and the other is the gas pressure threshold model. Considering that the probability of failure of the pipeline due to its own factors is very small, this embodiment assumes that the pipeline operation is 100% reliable without considering the influence of external factors. Therefore, the reliability assessment in this section adopts the gas pressure threshold model to reflect the transition process of the system from normal working state to fault state.

[0163] 3) The change process of air pressure over time is different for different pipe diameters, which leads to different buffering times. Therefore, this embodiment studies the distribution law of buffering time with different pipe diameters.

[0164] Based on the above assumptions, the finite element software is used to simulate the pressure change process of the first type of cascading failure and the second type of cascading failure. The basic input data of the simulation are shown in Table 2. Taking the pressure simulation of the first type of cascading failure as an example, Fig.11 (a) shows. The results show that the air pressure near the power system is higher than that near the natural gas system, the difference is about 750Pa, and the air pressure changes monotonically along the pipeline. The second type of cascading failure uses the same simulation process, and the results are shown in Fig.11 As shown in (b), the conclusion is exactly opposite to that of the second type of cascading failure.

[0165] Table 2 Input parameters of buffer time simulation

[0166]

[0167] S1: Buffer time evaluation: According to engineering practice, this embodiment selects four commonly used pipe diameters for simulation, namely DN315.5, DN395.5, DN590.1 and DN890.0, and it is considered that the air pressure change of any length of pipe is proportional to the air pressure change obtained by 100m simulation.

[0168] Fig.11 The pressure variation of pipelines with different diameters based on finite element software simulation is shown. Taking the DN315.5 pipeline as an example, Fig.11 The data in (a) are explained as follows:

[0169] 1) The upper half plane is a cascading failure scenario from the power system to the natural gas system. Before the failure occurs, the initial air pressure is 5.56Mpa. At Position = 0, the air pressure starts from 5.56Mpa and rises to 6.62Mpa within 199ms, an increase of about 35%. At the same time, at the end of the simulation, the air pressure difference between Position = 0 and Position = 100m is 1.952kPa.

[0170] 2) The lower half plane is the cascading failure scenario of the natural gas system propagating to the power system. At Position = 0, the gas pressure starts from 5.56Mpa and drops to 2.5Mpa within 819ms, a drop of about 55%. Also at the last moment of the simulation, the pressure difference between the two ends of the pipeline is 1.582kPa.

[0171] For the other three pipe diameters, similar conclusions can be obtained by using the same analysis method. Fig.11 As shown in (b)-(d).

[0172] At a certain moment, the air pressure at any position in the pipeline is different. For the convenience of analysis, the average air pressure is used in the following study to represent the air pressure level of the pipeline. The average air pressure at each point at the same moment is taken as Fig.12 The average air pressure changes over time. Fig.12 Based on the data in , combined with the upper and lower limits of pipeline gas pressure, the average buffer time of four diameters of pipelines under two types of cascading failures is listed respectively, as shown in Table 3.

[0173] Table 3 Average buffering time

[0174]

[0175] In this embodiment, S2 is as follows: Cascading failure probability density function: Determine the likelihood function to construct a cascading failure probability density function The key step. Given the average buffer time and mean time to repair In the case of j and t j Using alpha T =∑ j =1 N I(T b -T r <0)Restore and Finally, the probability density functions of the reliability of the two types of cascading failures obtained by formula (27) are as follows: Fig.13 As shown in (a) and (b).

[0176] S3 is as follows: Cascading failure probability assessment: Assume that the pipeline length is 100 km. This embodiment sets four scenarios to quantify the impact of average buffer time and average repair time on the probability of cascading failure in the electric-gas coupling system. The four scenarios are as follows:

[0177] Scenario 1: Consider a cascading failure from the power system to the natural gas system. The average buffer time is 600s, and different mean repair times are selected to evaluate the probability of cascading failures.

[0178] Scenario 2: Consider a cascading failure from the power system to the natural gas system. The average repair time is taken as 600s, and different average buffer times are selected to evaluate the probability of cascading failures.

[0179] Scenario 3: Same time setting as Scenario 1, but studying the cascading failures from the gas system to the electric power system.

[0180] Scenario 4: Same time setting as Scenario 2, but studying the cascading failures from the gas system to the electric power system.

[0181] Table 4 Evaluation results of scenario 1 and scenario 3

[0182]

[0183] Table 5 Evaluation results of scenario 2 and scenario 4

[0184]

[0185] From the analysis of Tables 4 and 5, it can be found that in scenario 1, when the mean repair time is less than the mean buffer time, the probability of cascading failures is relatively low. As the mean repair time increases, the probability of cascading failures gradually increases from 0.1199 to 0.3581. For scenario 2, as the mean buffer time increases, the probability of cascading failures gradually decreases from 0.7170 to 0.0235. Similarly, in scenario 3, as the mean repair time increases, the probability of cascading failures gradually increases from 0.1398 to 0.4635, and in scenario 4, as the mean buffer time increases, the probability of cascading failures gradually decreases from 0.6125 to 0.0103. It can be seen that the mean repair time and the mean buffer time have an important influence on the probability of cascading failures, which provides a reference for the optimization and improvement of the system. Reasonably increasing the average buffer time of the system or shortening the mean repair time can help improve the system's fault resistance, reduce the occurrence of cascading failures, and ensure the safe operation of the electric-gas coupling system.

[0186] The scope of protection claimed by the present invention is not limited to the above specific embodiments. For those skilled in the art, the present invention may have various deformations and changes. Any modifications, improvements and equivalent substitutions made within the concept and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A reliability assessment method for electric-pneumatic coupling system considering buffer time of cascading failures. Features: The steps include: S1: Based on the fluid dynamics equation, a buffer time evaluation model for the electric-pneumatic coupling system is established; the simulation data of the buffer time is obtained using finite element software; S2: Derive the martingale theory with buffer time and repair time, and establish a cascading failure model for the electric-pneumatic coupling system; S3: Based on simulation data and fault models, the Bayesian theorem is used to calculate the probability of cascading failures in the electric-pneumatic coupling system.

2. According to claim 1, a reliability assessment method for an electric-pneumatic coupling system considering a cascading failure buffer time, Features: Step S1 is specifically as follows: S1-1: To calculate the buffer time, it is necessary to establish an air pressure analysis model to obtain the law of air pressure change relative to time; according to the relevant theories of fluid dynamics, the dynamic expression of air pressure is obtained by solving the continuity equation, momentum conservation equation and state equation: In the formula, ρ and v are the density and flow velocity of the gas, t is the time, x is any position in the pipeline, p is the gas pressure, λ is the friction coefficient, D is the diameter of the pipeline, and c is the isothermal sound velocity of the gas; Arrange formulas (1) to (3), eliminate density ρ and gas flow velocity v in the formula, and obtain the following formula: Since the variation of the gas pressure in the pipeline with time and position is described by a set of partial differential equations, the characteristic line method is used to solve the variation law of the gas pressure with time; using the idea of ​​difference, the time and position of the pipeline are divided into several segments, each with a length of Δt and Δx, respectively. The points on the pipeline are set as A, B, C... At t = 0, if a cascading failure occurs from the natural gas system to the power system, the pipeline gas pressure will drop; since the gas pressure and mass flow rate cannot change suddenly, the gas pressure and mass flow rate of each point A, B, C,... on the pipeline at t = 0 are the steady-state values ​​before the failure occurs, which are recorded as p A 0 ,p B 0 ,p C 0 ,… and G A 0 ,G B 0 ,G C 0 ,…; after that, affected by the fault, the air pressure and mass flow rate in the pipeline change dynamically; after Δt time, the air pressure and mass flow rate at the same position but at different times become p A Δt ,p B Δt ,p C Δt ,… and G A Δt ,G B Δt ,G C Δt ,…; To obtain the air pressure and mass flow rate at each point after Δt time, use the air pressure p at points A and C at time t = 0 A 0 、p C 0 , solve for the pressure p at point B after Δt time B Δt ; At the same time, the mass flow rate G of points A and C at time t = 0 is used A 0 , G C 0 , solve for the mass flow rate G at point B after Δt time B Δt ; Based on the information of points A, B and C at the time of fault onset, the differential forms of the partial differential terms and linear terms in formulas (4) and (5) are given as follows: Substituting formula (6) and (7) into formula (4) and (5) respectively, the differential expression is: The characteristic line method shows that if the values ​​of the differentiated physical quantities Δt and Δx are extremely small, the relationship between the two is approximated as a linear relationship. It can be seen that Δt and Δx satisfy the following relationship: Δx=aΔt (10) In the formula, a is a constant; Substituting formula (10) into formula (8) and (9) respectively, we get: When the remaining parameters are known, use MATLAB loop iteration to find p B Δt and G B Δt Numerical solution at each point, and set to use only p B Δt The value of is used for subsequent analysis; Formulas (11) and (12) are used to calculate the pressure p at each node along the pipeline after Δt time. A Δt ,p B Δt ,p C Δt ,…; Average pipeline pressure p after Δt time Δt for: Where b is the number of discrete segments of the pipeline; S1-2: The calculation process of the buffer time of cascading faults propagating from the natural gas system to the power system is as follows: Δt and safety threshold p 1 , if p Δt > 1 , it means that after Δt time, the pipeline is not faulty; then, continue to use the air pressure p A Δt ,p B Δt ,p C Δt ,…calculate the air pressure p after 2Δt A 2Δt ,p B 2Δt ,p C 2Δt ,…; and so on, after nΔt time, n is a natural number, if p nΔt With p 1 The absolute value of the difference is less than an infinitesimal quantity, i.e. |p nΔt -p 1 |≤ε, it means that after nΔt, the pipeline gas pressure is infinitely close to the fault threshold; therefore, the buffer time of the cascading fault propagation from the natural gas system to the power system is: T b =nΔt (14) S1-3: Calculation process of buffer time of cascading failures propagating from power system to natural gas system: The buffer time of cascading failures is calculated based on the gas pressure p Δt and safety threshold p 2 Then calculate, after nΔt time, if p nΔt With p 1 The absolute value of the difference is less than an infinitesimal quantity, i.e. |p nΔt -p 2 |≤ε, indicating that after nΔt time, the pipeline gas pressure approaches the fault threshold infinitely; at this time, the differential calculation process is stopped and the buffer time is obtained.

3. A reliability assessment method for an electric-pneumatic coupling system considering a cascading failure buffer time according to claim 2, Features: Step S2 is specifically as follows: When a cascading failure occurs in the electric-pneumatic coupling system, the number of failures at the next moment is not less than the number of failures at the previous moment. Therefore, the number of failures in the electric-pneumatic coupling system meets the definition of a lower martingale. On this basis, according to the Doob-Meyer decomposition theorem, the change in the number of failures in the electric-pneumatic coupling system is uniquely decomposed into the following two parts: In the formula is the domain flow, which contains the fault observation information before time t; Non-decreasing predictable random process Observe information for a given fault When , the mathematical expectation of the change in the number of failures dN(t); dM(t) is a zero-mean martingale; When the average buffer time T b Less than the mean time to repair a fault T r When the component is repaired in the fault buffer stage, it is considered that the component has not failed, so it will not cause subsequent cascading failures; On the contrary, when T b Greater than T r When the component fails to be repaired within the buffer time, the failure of the component will continue to cause changes in the operating conditions in the system, causing more components to participate in the chain failure; the subsequent faults at all levels will repeat the judgment process; using the indicative function I(T r -T b <0) indicates the result of a single component failure, I:T r -T b <0; then the number of faulty components in the system at time T is n T The expression is: Where N is the total number of system components, T b,j is the buffer time of the jth component in the gas grid; when a cascading failure occurs from the natural gas system to the power system, T r,j is the repair time of the jth component; When a cascading failure occurs from the power system to the natural gas system, T r,j is the repair time of the component that caused the jth component failure; When the fault information of the known cascading fault before time T When , the change in the number of faults in the future period of time Δt is the fault rate ρ. Using the idea of ​​limit, the differential form of the fault rate is obtained as follows: Where N(t+Δt) and N(t) are the number of faulty components in the system at time t+Δt and t, respectively; Substituting formula (17) into formula (15), the martingale expression of the number of faults is converted to: Using the buffer time T b and repair time T r The total number of failures n T The ratio of the time T from the beginning of the cascading failure to the present is used to calculate the historical failure rate of the system, which is equivalent to the future failure rate of the system. Thus, we have: Combining formula (18) and formula (19), we get the improved martingale expression considering buffer time and repair time as follows:

4. A reliability assessment method for an electric-pneumatic coupling system considering a cascading failure buffer time according to claim 3, Features: Step S3 is as follows: Bayes' theorem defines the expression for the probability density function of a binary random variable: p(x,y) = cp 0 (x,y)Λ (21) Where (x, y) is a binary random variable; p 0 (x, y) is the prior probability density function of the binary random variable; Λ is the likelihood function; p(x, y) is the probability density function of the random variable; c is a constant; S3-1: Constructing a priori probability density function p 0 (T b ,T r ): Assume the buffer time T b Following the gamma distribution, the repair time is T r They obey exponential distribution, and their probability density function expressions are shown in formula (22): Where Г(·) is the gamma function; γ is the shape parameter of the gamma distribution; β is the scale parameter of the gamma distribution; μ is the repair rate; T b and T r As two independent random variables, we get (T b ,T r )’s prior probability density function: In the formula, c 1 、c 2 are all constants; S3-2: Construct likelihood function Λ: According to the Radon-Nikodym theorem, for a random process with continuous time parameters and discrete state space, the likelihood function is expressed as follows: Where ρ is the failure rate; n s is the number of faulty components at time s; Substituting the failure rate ρ into formula (24), we can obtain: Let α T =∑ j=1 N I(T b -T r <0), the specific form of the likelihood function is as follows: S3-3: The probability of a cascading failure P(T b ,T r ) Based on the determined prior probability density function p 0 (T b ,T r ) and likelihood function Λ, substitute formula (23) and (26) into formula (21), and the probability density function of cascading failure is obtained as: Therefore, the probability of cascading failures P(T b ,T r )for:

5. A reliability assessment method for an electric-pneumatic coupling system considering a cascading failure buffer time according to claim 2, Features: Step S1 also includes the following steps: S1-4: Use the finite element software ANSYS to simulate the pressure change process when a cascading failure occurs in the natural gas system and obtain experimental data on the buffer time; S1-4-1: Use the Design Modeler built into ANSYS software to draw the natural gas pipeline model demonstrated in this section; S1-4-2: Further process the model, define the boundary in the positive direction of the y-axis as the gas inlet Inlet, the boundary in the negative direction of the y-axis as the gas outlet Outlet, and the outermost layer in the middle of the inlet and outlet as the pipe wall Wall; use the software's built-in expansion method to perform meshing and set parameters, including expansion options, boundary layer transition ratio, maximum number of boundary layers, boundary growth rate, mesh size and feature size; S1-4-3: Use ANSYS to perform transient simulation of air pressure and obtain the change process of air pressure with time and position in the full time range for two types of cascading failures; S1-4-4: Extract the relationship between air pressure and time, determine the upper and lower limits of the air pressure in the corresponding section of the pipeline, and read the specific value of the buffer time from the graph.