Method for identifying weak link of trans-regional electro-hydrogen comprehensive energy system based on data-driven polynomial chaos expansion

By employing a data-driven polynomial chaotic expansion method, a weak link identification model for cross-regional integrated electric-hydrogen energy systems is constructed. This solves the problem of identifying weak links in cross-regional interconnected systems, achieves efficient and accurate system risk assessment, and improves the reliability of system operation.

CN121032176APending Publication Date: 2025-11-28CHONGQING UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202510930087.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-07
Publication Date
2025-11-28

AI Technical Summary

Technical Problem

Existing methods for identifying weaknesses in a single system cannot be effectively applied to inter-regional interconnected integrated electric-hydrogen energy systems. This leads to an increased burden of risk assessment calculations, a greater diversity of system operating states, and a greater complexity of energy flow distribution, making it difficult to accurately identify weaknesses.

Method used

A data-driven polynomial chaos expansion method is adopted to construct a hydrogen energy system that integrates hydrogen energy across regions and promotes mutual benefit. The optimal energy flow state of the system is calculated by using orthogonal polynomial basis of random variables and DPCE proxy model, and weak links are identified by ranking risk indicators.

Benefits of technology

It enables accurate and rapid identification of weak links in the integrated electric-hydrogen energy system, improves the efficiency of system risk assessment, and helps improve the level of system operation risk.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121032176A_ABST
    Figure CN121032176A_ABST
Patent Text Reader

Abstract

The invention discloses a data-driven polynomial chaos expansion-based weak link identification method for a trans-regional electricity-hydrogen comprehensive energy system. The method comprises the following steps of 1) constructing a hydrogen energy trans-regional complementary electricity-hydrogen comprehensive energy system; 2) establishing an optimal energy flow calculation model of the electro-hydrogen comprehensive energy system; 3) establishing a random variable orthogonal polynomial basis based on the multi-order moment of the system random factor data set; 4) establishing a DPCE proxy model of random output response; 5) solving a weighting coefficient of the model by using a least square method, and completing the construction of the DPCE proxy model; 6) calculating an optimal energy flow state value of the system through the DPCE proxy model; and 7) calculating system risk indexes, and identifying weak links of the electro-hydrogen comprehensive energy system based on risk index sorting. According to the method, the optimal energy flow calculation agent model of the system is constructed through the DPCE method, the problems of system operation state diversification and energy flow coupling complexity are effectively solved, and weak links of the electro-hydrogen comprehensive energy system are accurately identified.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application relates to the field of integrated energy system risk assessment and weak link identification, in particular to a cross-regional electricity-hydrogen integrated energy system weak link identification method based on data-driven polynomial chaos expansion. BACKGROUND

[0002] The cross-time and space flexible adjustment characteristics of hydrogen energy reconstruct the operation mode of the electricity-hydrogen integrated energy system, and change the weak link of the system. The calculation and analysis of the risk level of each part of the system can pinpoint the key factors affecting the energy supply of the system, providing an important decision basis for improving the energy supply reliability of the system. Therefore, the research on the weak link identification of the electricity-hydrogen integrated energy system helps to further tap the support of the hydrogen energy system to the power system, and provides a key theoretical basis for building a safe and reliable electricity-hydrogen integrated energy system.

[0003] With the continuous development of hydrogen energy systems, the electricity-hydrogen integrated energy system presents the characteristics of cross-regional interconnection and coordinated development, and its system size continues to increase. However, few studies have been conducted on the weak link identification of the electricity-hydrogen integrated energy system, which to some extent restricts the improvement of its safety and reliability level. In addition, the cross-regional transmission and cross-medium coupling of electricity and hydrogen energy enhance the complexity of the system, profoundly change the energy flow distribution of the system, and the strong randomness of the source, load and equipment failures of the system leads to the diversification of the system operation state, aggravating the calculation burden of the system risk assessment. The conventional single system weak link identification method and deterministic risk assessment method cannot effectively apply to the operation risk analysis of the electricity-hydrogen integrated energy system supported by hydrogen energy cross-regional complementation.

[0004] Therefore, it is urgent to establish a corresponding risk index to quantify the influence of hydrogen energy cross-regional complementary coordination on the operation of the electricity-hydrogen integrated energy system, and to tap the high-efficiency solving potential of the DPCE method in dealing with large-scale, multi-state probability evaluation problems, to realize the accurate and rapid identification of the weak link of the system, to provide an important decision basis for improving the operation risk level of the system, and to help the high-quality development of the modern energy system. SUMMARY

[0005] The purpose of the present application is to provide a cross-regional electricity-hydrogen integrated energy system weak link identification method based on data-driven polynomial chaos expansion, comprising the following steps:

[0006] 1) Construct a hydrogen energy cross-regional complementary coordination electricity-hydrogen integrated energy system, and obtain the parameters of the electricity-hydrogen integrated energy system;

[0007] 2) Establish an optimal energy flow calculation model of the hydrogen energy cross-regional complementary coordination electricity-hydrogen integrated energy system;

[0008] 3) Constructing random variable dataset based on system source load history random fluctuation data and equipment history random failure data, and establishing random variable orthogonal polynomial basis based on data multiple moments;

[0009] 4) Constructing standard orthogonal polynomials through full tensor product of random variable orthogonal polynomial basis, and establishing DPCE surrogate model of random output response based on weighted linear combination of each standard orthogonal polynomial;

[0010] 5) Selecting random variable data based on linear independence principle, and calculating system state response by optimal energy flow calculation model of electric hydrogen integrated energy system;

[0011] Taking random variable data and system state response as input and output, and solving weighted coefficients of DPCE surrogate model by least square method, thereby constructing DPCE surrogate model for calculating optimal energy flow state of system;

[0012] 6) Obtaining current electric hydrogen integrated energy system source load random power and equipment random failure state, and calculating system optimal energy flow state value by DPCE surrogate model;

[0013] 7) Calculating system risk index according to product of system state random probability and system state risk severity, and identifying weak link of electric hydrogen integrated energy system based on risk index sorting.

[0014] Further, in the hydrogen energy cross-regional complementary coordination electric hydrogen integrated energy system, the hydrogen energy in a single regional electric hydrogen integrated energy system realizes electric-hydrogen heterogeneous energy bidirectional conversion relying on the electric-hydrogen conversion equipment in the hydrogen refueling station; the electric energy between multiple regional electric hydrogen integrated energy systems is transmitted cross-region through the power tie line, and the hydrogen energy is transported cross-region through the hydrogen long tube trailer.

[0015] Further, the electric hydrogen integrated energy system parameters include: electric hydrogen system topology connection parameters, power line, wind and light unit, electrolyzer, hydrogen fuel cell, hydrogen storage tank, hydrogen refueling machine and hydrogen long tube trailer equipment parameters, wind and light unit processing and electric hydrogen load demand parameters.

[0016] Further, the objective function minC of the optimal energy flow calculation model of the hydrogen energy cross-regional complementary coordination electric hydrogen integrated energy system is as follows:

[0017]

[0018] In the formula, C BUY,m is the electricity purchase cost of the system in region m; C DG,m is the distributed power wind and light abandonment penalty cost of the system in region m; C SHED,m is the electric hydrogen load reduction penalty cost of the system in region m; C DR,m is the hydrogen load demand response cost of the system in region m; CFC,m Cost of hydrogen fuel cell power generation for regional m system; C HTT,m Cost of regional hydrogen long tube trailer operation for regional m system; C HTT,MR Cost of inter-regional hydrogen long tube trailer operation; N m Number of regions.

[0019] Further, the constraint conditions of the optimal energy flow calculation model of the hydrogen energy cross-region complementary coordinated electric-hydrogen integrated energy system include electric-hydrogen equipment operation constraints, electric-hydrogen power balance constraints, and hydrogen energy cross-region transportation constraints.

[0020] The electric-hydrogen equipment operation constraints are as follows:

[0021] P WE,it = k1N WE,i E WE,it I WE,it (2)

[0022]

[0023] Q FC,it = H FC,it / p H2 (7)

[0024] H HT,it = H HT,i,t-1 + (η in H HT,in,it - H HT,out,it / η out ) (8)

[0025] H DP,out,it = η DP H DP,in,it (9)

[0026] In the formula, N WE,i is the number of electrolytic cells in series and parallel connection, N FC,i is the number of hydrogen fuel cells in series and parallel connection; P WE,it , Q WE,it are the active power and reactive power consumed by the electrolytic cell for hydrogen production; I WE,it is the working current of the electrolytic cell, E WE,it is the working voltage of the electrolytic cell at node i at time t; E FC,it is the working voltage of the hydrogen fuel cell, P FC,it , Q FC,it are the active power and reactive power output by the hydrogen fuel cell; I FC,it is the working current of the hydrogen fuel cell; k1 and k2 are conversion coefficients, η f is the Faraday efficiency, p H2 is the standard density of hydrogen; H HT,i,t , HHT,i,t-1 η represents the hydrogen storage capacity of the hydrogen storage tank at time t and time t-1, respectively; in η out These represent the hydrogen filling efficiency and hydrogen discharging efficiency of the hydrogen storage tank, respectively; H DP,in,it The amount of hydrogen injected into the hydrogen refueling unit; H DP,out,it The amount of hydrogen output from the hydrogen refueling unit; η DP The hydrogenation efficiency of the hydrogenation unit is Δt; the time step is H. WE,it H is the hydrogen production rate of the electrolyzer; F is the Faraday constant; H HT,in,it H HT,out,it For the injection and output of hydrogen gas into the hydrogen storage tank; H FC,it Hydrogen consumption of hydrogen fuel cells;

[0027] The power balance constraints for hydrogen electrolysis are shown below:

[0028]

[0029] U i,t -U j,t ≤M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t (14)

[0030] U i,t -U j,t ≥-M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t (15)

[0031] P Grid,it +P WG,it +P PV,it +P FC,it =P WE,it +P load,it -L p,it i∈Ω s (16)

[0032] Q Grid,it +Q WG,it +Q PV,it =Q load,it -L q,it i∈Ω s (17)

[0033]

[0034] 0≤L p,it ≤P load,it (twenty three)

[0035] 0≤L q,it ≤Q load,it (twenty four)

[0036] 0≤L h,it ≤H load,it (25)

[0037] In the formula, P ij,t Q ij,t Let P be the active power and reactive power of branch ij at time t, respectively; Grid,jt Q Grid,jt P represents the active and reactive power of the hydrogen energy system interacting with the distribution network at time t, i.e., node j; load,jt Q load,jt These represent the active and reactive power of the electrical load at time t, respectively, at node j; L p,it L q,jt Let r be the active and reactive power reduction of the electrical load at time t at node j; ij x ij These are the resistance and reactance values ​​of branch ij, respectively; U i,t P is the square of the voltage value at time t at node i; WG,it Q WG,it P represents the actual active power and reactive power output of the wind turbine at time t, i and t, respectively, at the electro-hydrogen coupling node i; PV,it Q PV,it P represents the actual active power and reactive power output of the photovoltaic unit at time t, i and t, respectively, at the electro-hydrogen coupling node i; FC,it P WE,it H represents the active power output of the hydrogen fuel cell at time t, which is the active power consumed by the electrolyzer at time i, respectively; load,it The hydrogen load demand of hydrogen refueling station nodes; These represent the amount of hydrogen added and the amount of hydrogen released from the hydrogen storage tank of the c-th hydrogen long-tube trailer at the hydrogen refueling station at time t i, respectively. Let M and α represent the hydrogen filling and discharging amounts of the hydrogen storage tank of the k-th inter-regional hydrogen tube trailer at the r-th hydrogen refueling station at time t, respectively; M is a sufficiently large integer; α ij,t Let L be a binary variable, representing the switching state of surface branch ij at time t; h,it This refers to the reduction in hydrogen load at hydrogen refueling station nodes;

[0038] The cross-regional transportation constraints for hydrogen energy are as follows:

[0039]

[0040]

[0041] In the formula, A c,itis the position state variable of the cth hydrogen long tube trailer at time t whether it is located at the i th hydrogen sub-station or the road; c,0t is the position state variable of the cth hydrogen long tube trailer at time t whether it is located at the road; c,it c,0t is the position change state of the cth hydrogen long tube trailer at time t; π(j, i) is the traffic distance of the hydrogen long tube trailer transferred between two hydrogen sub-stations in a certain area, and v is the average driving speed of the hydrogen long tube trailer, represents rounding up. c,t c,t-1 and are the hydrogen storage amounts of the hydrogen storage tank of the cth hydrogen long tube trailer at time t-1, respectively; and are the hydrogen charging amount and hydrogen discharging amount of the hydrogen storage tank of the cth hydrogen long tube trailer at time t at the i th node hydrogen station, respectively; k,rt is the position state variable of the kth hydrogen long tube trailer at time t whether it is located at the rth on-site hydrogen station or the road; k,0t is the position state variable of the kth hydrogen long tube trailer at time t whether it is located at the road; RE is the set of on-site hydrogen stations in all regional hydrogen supply networks. k,rt k,0t is the position change state of the kth hydrogen long tube trailer at time t. π(p, r) is the traffic distance of the hydrogen long tube trailer transferred between two hydrogen stations. k,t k,t-1 and are the hydrogen storage amounts of the hydrogen storage tank of the kth regional hydrogen long tube trailer at time t and time t-1, respectively; and are the hydrogen charging amount and hydrogen discharging amount of the hydrogen storage tank of the kth regional hydrogen long tube trailer at time t at the rth hydrogen sub-station.

[0042] Further, in step 3), the step of establishing a random variable orthogonal polynomial basis based on data multiple moments includes:

[0043] 3.1) For any random variable ξ i in the random variable data set, define a one-dimensional orthogonal polynomial and a k-order origin moment μ k,i , that is:

[0044]

[0045]

[0046] In the formula, P i (l) (ξ i ) is a one-dimensional orthogonal polynomial; is a polynomial coefficient; d(ξ i ) is a probability distribution;​​​​

[0047] where the one-dimensional orthogonal polynomials satisfy the following constraints:

[0048]

[0049] where m, l = {0, 1, 2, …, H};

[0050] 3.2) Solve the arbitrary polynomial basis p (l)k,i and the moment matching equation of the k-th order origin moment μ i of the random variable ξ k,i to obtain the coefficients of the one-dimensional orthogonal polynomial basis;

[0051] where the moment matching equation is as follows:

[0052]

[0053] 3.3) Repeat steps 3.1) - step 3.2) to obtain the orthogonal polynomial basis of each random variable in the random variable data set, thereby constructing the random variable orthogonal polynomial basis based on the data multi-order moments.

[0054] Further, in step 4), the step of establishing the DPCE proxy model of the random output response includes:

[0055] 4.1) Calculate the full tensor of the multi-dimensional random variable orthogonal polynomial basis, that is:

[0056]

[0057] where α k,i is the order of the i-th one-dimensional polynomial;

[0058] 4.2) Based on the full tensor of the random variable orthogonal polynomial basis, establish the DPCE proxy model of the random output response, that is:

[0059]

[0060] where c k is the coefficient of the k-th expansion term φ k to be solved, M is the number of terms of the expansion Y, M = (H+N)! / (H!N!), the k-th standard orthogonal polynomial φ k (k = 1, 2, …, M) is the full tensor product of one-dimensional polynomials.

[0061] Further, in step 5), the step of the DPCE proxy includes:

[0062] 5.1) select a linearly independent input sample combination ξ from the random variable dataset, and calculate the system state response Y(ξ) through the optimal energy flow model;

[0063] 5.2) substitute the sample values of the random variables and the corresponding response function values into equations (45)-(46), and solve the weighted coefficients of the DPCE surrogate model by using the least square method;

[0064]

[0065] ΦC=Y (46)

[0066] The weighted coefficients of the DPCE surrogate model are as follows:

[0067]

[0068] In the formula, is a weighted coefficient vector.

[0069] Further, in step 6), the step of calculating the system optimal energy flow state value through the DPCE surrogate model comprises:

[0070] The wind and light unit output represented by continuous random variables and the system device failure represented by discrete random variables are obtained from the original dataset or the assumed probability distribution.

[0071] The obtained wind and light unit output and system device failure are substituted into the DPCE surrogate model to calculate the system state response; the system state response includes node voltage, line power and node electricity and hydrogen load reduction values.

[0072] Further, in step 7), the step of identifying the weak link of the electricity-hydrogen integrated energy system based on the risk index ranking comprises:

[0073] The corresponding node voltage out-of-limit risk index, line overload risk index, node electricity load reduction risk index and node hydrogen load reduction risk index are calculated through the system state response value.

[0074] The weak link of the electricity-hydrogen integrated energy system is obtained by ranking the risk indexes.

[0075] The technical effect of the present application is self-evident. The electricity-hydrogen integrated energy system weak link identification method based on DPCE is helpful to improve the efficiency of electricity-hydrogen integrated energy system weak link identification and realize accurate positioning of the system weak link.

[0076] The application applies DPCE to the risk assessment and weak link identification of an integrated electricity-hydrogen energy system. By using discrete random variables to represent system equipment failures and continuous random variables to represent wind and solar power output, a random factor dataset is constructed, and based on the data moments, an orthogonal polynomial is constructed to obtain a system risk assessment proxy model based on DPCE. This method can effectively deal with the weak link identification problem of system operation state diversification and complex energy flow coupling, thereby realizing the rapid calculation of the risk index of the integrated electricity-hydrogen energy system and accurately identifying the weak link of the integrated electricity-hydrogen energy system. BRIEF DESCRIPTION OF DRAWINGS

[0077] Figure 1 Flow chart of the weak link identification method of the cross-regional integrated electricity-hydrogen energy system based on DPCE;

[0078] Figure 2 Schematic diagram of the complementary and coordinated mechanism of the integrated electricity-hydrogen energy system;

[0079] Figure 3 Schematic diagram of the hydrogen energy cross-regional support mechanism;

[0080] Figure 4 Topology connection diagram of the integrated electricity-hydrogen energy system;

[0081] Figure 5 Relative error of the mean value of the system state quantity obtained by the DPCE method under different expansion orders. DETAILED DESCRIPTION

[0082] The application will be further described below in conjunction with the embodiments, but should not be understood as limiting the above-mentioned subject matter of the application to the following embodiments. According to ordinary technical knowledge and conventional means in the art, various substitutions and modifications can be made without departing from the above-mentioned technical idea of the application, and all should be included in the protection scope of the application.

[0083] Example 1:

[0084] The weak link identification method of the cross-regional integrated electricity-hydrogen energy system based on data-driven polynomial chaos expansion includes the following steps:

[0085] 1) Construct a hydrogen energy cross-regional complementary and coordinated integrated electricity-hydrogen energy system, and obtain the parameters of the integrated electricity-hydrogen energy system;

[0086] 2) Establish an optimal energy flow calculation model of the hydrogen energy cross-regional complementary and coordinated integrated electricity-hydrogen energy system;

[0087] 3) Based on the historical random fluctuation data of the system source and load and the historical random failure data of the equipment, construct a random variable dataset, and establish an orthogonal polynomial basis of the random variable based on the data moments;

[0088] 4) Constructing standard orthogonal polynomials by full tensor product of orthogonal polynomial bases of random variables, and establishing DPCE surrogate model of random output response based on weighted linear combination of each standard orthogonal polynomial;

[0089] 5) Selecting random variable data based on linear independence principle, and calculating system state response by optimal energy flow calculation model of integrated electricity-hydrogen energy system;

[0090] Taking random variable data and system state response as input and output, and solving weighted coefficients of DPCE surrogate model by least square method, thereby constructing DPCE surrogate model for calculating optimal energy flow state of system;

[0091] 6) Obtaining current random power of source and load and random fault state of equipment of integrated electricity-hydrogen energy system, and calculating optimal energy flow state value of system by DPCE surrogate model;

[0092] 7) Calculating system risk index according to product of system state random probability and risk severity of system state, and identifying weak link of integrated electricity-hydrogen energy system based on risk index ranking.

[0093] Embodiment 2:

[0094] The weak link identification method of cross-regional integrated electricity-hydrogen energy system based on data-driven polynomial chaos expansion has the same technical content as embodiment 1, and further, in the hydrogen energy cross-regional complementary coordination integrated electricity-hydrogen energy system, the hydrogen energy in a single regional integrated electricity-hydrogen energy system realizes electric-hydrogen heterogeneous energy bidirectional conversion relying on the electric-hydrogen conversion equipment in the hydrogen refueling station; the electric energy between multiple regional integrated electricity-hydrogen energy systems is transmitted across regions through power tie lines, and the hydrogen energy is transported across regions through hydrogen long tube trailers.

[0095] Embodiment 3:

[0096] The weak link identification method of cross-regional integrated electricity-hydrogen energy system based on data-driven polynomial chaos expansion has the same technical content as any one of embodiments 1-2, and further, the parameters of the integrated electricity-hydrogen energy system include: topological connection parameters of the electricity-hydrogen system, device parameters of power lines, wind and light units, electrolytic cells, hydrogen fuel cells, hydrogen storage tanks, hydrogen refueling machines and hydrogen long tube trailers, and wind and light unit processing and electricity-hydrogen load demand parameters.

[0097] Embodiment 4:

[0098] The weak link identification method of cross-regional integrated electricity-hydrogen energy system based on data-driven polynomial chaos expansion has the same technical content as any one of embodiments 1-3, and further, the objective function minC of the optimal energy flow calculation model of the hydrogen energy cross-regional complementary coordination integrated electricity-hydrogen energy system is as follows:

[0099]

[0100] wherein C BUY,m is the electricity purchase cost of the regional m system; C DG,m is the distributed generation wind and light abandonment penalty cost of the regional m system; C SHED,m is the electricity and hydrogen load reduction penalty cost of the regional m system; C DR,m is the hydrogen load demand response cost of the regional m system; C FC,m is the hydrogen fuel cell power generation cost of the regional m system; C HTT,m is the regional hydrogen long tube trailer operation cost of the regional m system; C HTT,MR is the inter-regional hydrogen long tube trailer operation cost; N m is the number of regions.

[0101] Embodiment 5:

[0102] The weak link identification method of the cross-regional electricity and hydrogen comprehensive energy system based on the data-driven polynomial chaos expansion, the technical content is the same as any one of embodiments 1-4, further, the constraint conditions of the optimal energy flow calculation model of the hydrogen energy cross-regional complementary coordination electricity and hydrogen comprehensive energy system include electricity and hydrogen equipment operation constraints, electricity and hydrogen power balance constraints and hydrogen energy cross-regional transportation constraints.

[0103] wherein the electricity and hydrogen equipment operation constraints are as follows:

[0104] P WE,it =k1N WE,i E WE,it I WE,it (2)

[0105]

[0106] Q FC,it =H FC,it / ρ H2 (7)

[0107] H HT,it =H HT,i,t-1 +(η in H HT,in,it -H HT,out,it / η out ) (8)

[0108] H DP,out,it =η DP H DP,in,it (9)

[0109] wherein N WE,i is the number of electrolytic cells in series and parallel connection, N FC,i is the number of hydrogen fuel cells in series and parallel connection; P WE,it , Q WE,it are the active power and reactive power consumed by the electrolytic cell for hydrogen production; IWE,it E is the working current of the electrolyzer WE,it Ei(t) is the working voltage of the electrolyzer at node i at time t FC,it P is the working voltage of the hydrogen fuel cell FC,it , Q FC,it P and Q are the active power and reactive power output by the hydrogen fuel cell FC,it I is the working current of the hydrogen fuel cell; k1, k2 are conversion coefficients f η is the Faraday efficiency H2 ρ is the standard density of hydrogen HT,i,t , H HT,i,t-1 Ht and Ht-1 are the hydrogen storage amounts of the hydrogen storage tank at time t and time t-1 respectively in , η out ηH and ηH are the hydrogen charging efficiency and hydrogen discharging efficiency of the hydrogen storage tank respectively DP,in,it Hin is the hydrogen injection amount of the hydrogen filling machine DP,out,it Hout is the hydrogen output amount of the hydrogen filling machine DP ηH is the hydrogen charging efficiency of the hydrogen filling machine WE,it H is the hydrogen production amount of the electrolyzer; F is the Faraday constant HT,in,it , H HT,out,it Hin and Hout are the hydrogen injection and output amounts of the hydrogen storage tank FC,it HFC is the hydrogen consumption amount of the hydrogen fuel cell

[0110] The power balance constraint of the hydrogen system is as follows:

[0111]

[0112] U i,t -U j,t ≤M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t ) (14)

[0113] U i,t -U j,t ≥-M(1-α ij,t )+2( rij P ij,t +x ij Q ij,t ) (15)

[0114] P Grid,it +P WG,it +P PV,it +P FC,it =P WE,it +P load,it -L p,it i∈Ω s (16)

[0115] Q Grid,it +Q WG,it +Q PV,it =Q load,it -L q,it i∈Ω s (17)

[0116]

[0117] 0≤L p,it ≤P load,it (23)

[0118] 0≤L q,it ≤Q load,it (24)

[0119] 0≤L h,it ≤H load,it (25)

[0120] In the formula, P ij,t , Q ij,t are active power and reactive power of branch ij at time t; P Grid,jt , Q Grid,jt are active power and reactive power of hydrogen energy system and power distribution network at time t; P load,jt , Q load,jt are active power and reactive power of electric load at time t; L p,it , L q,jt are active power and reactive power of electric load reduction at time t; r ij , x ij are resistance value and reactance value of branch ij; U i,t is square of voltage value of node i at time t; P WG,it , Q WG,it are active power and reactive power of wind turbine actual output at time t; P PV,it , Q PV,it are active power and reactive power of photovoltaic actual output at time t; P FC,it , P WE,it are active power and active power of electrolytic tank consumption of hydrogen fuel cell output at time t; H load,it is hydrogen load demand of hydrogen filling station node; are hydrogen filling amount and hydrogen discharging amount of hydrogen storage tank of the cth hydrogen long tube trailer at time t at node i of hydrogen filling station; are hydrogen filling amount and hydrogen discharging amount of hydrogen storage tank of the kth inter-regional hydrogen long tube trailer at time t at the rth hydrogen filling mother station; M is a large enough integer; α ij,tis the switch state of surface branch ij at time t; L h,it is the hydrogen load reduction amount of the hydrogen station node;

[0121] The hydrogen energy cross-region transportation constraints are as follows:

[0122]

[0123]

[0124] wherein, A c,it is the position state variable of the cth hydrogen long tube trailer at time t whether it is located at the ith hydrogen sub-mother station or road; A c,0t is the position state variable of the cth hydrogen long tube trailer at time t whether it is located at the road; h c,it , h c,0t is the position change state of the cth hydrogen long tube trailer at time t; π(j, i) is the transportation distance of the hydrogen long tube trailer between two hydrogen sub-mother stations in a region, and v is the average driving speed of the hydrogen long tube trailer, represents rounding up. H c,t , H c,t-1 are the hydrogen storage amounts of the hydrogen storage tank of the cth hydrogen long tube trailer at and time t-1, respectively; are the hydrogen charging amount and hydrogen discharging amount of the hydrogen storage tank of the cth hydrogen long tube trailer at time t at the node i hydrogen station, respectively; A k,rt is the position state variable of the kth hydrogen long tube trailer at time t whether it is located at the rth on-site hydrogen station or road; A k,0t is the position state variable of the kth hydrogen long tube trailer at time t whether it is located at the road; Ω RE is the set of on-site hydrogen stations in all regional hydrogen supply networks. h k,rt , h k,0t is the position change state of the kth hydrogen long tube trailer at time t. π(p, r) is the transportation distance of the hydrogen long tube trailer between two hydrogen stations. H k,t , H k,t-1 are the hydrogen storage amounts of the hydrogen storage tank of the kth regional hydrogen long tube trailer at time t and time t-1, respectively; are the hydrogen charging amount and hydrogen discharging amount of the hydrogen storage tank of the kth regional hydrogen long tube trailer at time t at the rth hydrogen mother station, respectively.

[0125] Example 6:

[0126] The weak link identification method of the cross-region electric-hydrogen integrated energy system based on the data-driven polynomial chaos expansion has the same technical content as any one of examples 1-5, and further, in step 3), the step of establishing the orthogonal polynomial basis of the random variable based on the data multiple moments comprises:

[0127] 3.1) For any random variable ξ in the random variable dataset i Define a one-dimensional orthogonal polynomial and k-th order origin moment μ k,i ,Right now:

[0128]

[0129] In the formula, P i (l) (ξ i () is a one-dimensional orthogonal polynomial; d(ξ) represents the polynomial coefficients. i ) represents a probability distribution;

[0130] Among them, one-dimensional orthogonal polynomial The following constraints must be met:

[0131]

[0132] In the formula, m,l={0,1,2,…,H};

[0133] 3.2) Solving for arbitrary polynomial basis p (l)k,i With random variable ξ i k-th order origin moment μ k,i The moment matching equations are used to obtain the coefficients of the one-dimensional orthogonal polynomial basis;

[0134] The moment matching equation is as follows:

[0135]

[0136] 3.3) Repeat steps 3.1)-3.2) to obtain the orthogonal polynomial basis for each random variable in the random variable dataset, thereby constructing an orthogonal polynomial basis for random variables based on the multi-moments of the data.

[0137] Example 7:

[0138] A method for identifying weak links in a cross-regional integrated hydrogen-electric energy system based on data-driven polynomial chaotic expansion, with technical content identical to any one of Examples 1-6, further comprising the following steps in step 4): Establishing a DPCE proxy model for random output response.

[0139] 4.1) Calculate the full tensor of the orthogonal polynomial basis of multidimensional random variables, i.e.:

[0140]

[0141] In the formula, α k,i Let be the order of the i-th one-dimensional polynomial;

[0142] 4.2) Based on the full tensor of orthogonal polynomial basis of random variables, the DPCE surrogate model of random output response is established, that is:

[0143]

[0144] where c k is the coefficient of the kth expansion term Φ k to be solved, M is the number of terms of the expansion Y, M=(H+N)! / (H!N!), the kth standard orthogonal polynomial Φ k (k=1, 2, …, M) is a full tensor product of one-dimensional polynomials.

[0145] Example 8:

[0146] The weak link identification method of the cross-regional electricity-hydrogen integrated energy system based on data-driven polynomial chaos expansion, the technical content is the same as any one of embodiments 1-7, and the step of establishing the DPCE surrogate model in step 5) includes:

[0147] 5.1) Select a linearly independent input sample combination ξ from the random variable data set, and calculate the system state response Y(ξ) through the optimal energy flow model;

[0148] 5.2) Substitute the sample values of the random variables and the corresponding response function values into equations (45)-(46), and solve the weighted coefficients of the DPCE surrogate model by the least squares method;

[0149]

[0150] ΦC=Y (46)

[0151] The weighted coefficients of the DPCE surrogate model are as follows:

[0152]

[0153] wherein, is the weighted coefficient vector.

[0154] Example 9:

[0155] The weak link identification method of the cross-regional electricity-hydrogen integrated energy system based on data-driven polynomial chaos expansion, the technical content is the same as any one of embodiments 1-8, further, in step 6), the step of calculating the optimal energy flow state value of the system through the DPCE surrogate model includes:

[0156] Obtain the wind and light unit output represented by continuous random variables and the system equipment failure represented by discrete random variables from the original data set or the assumed probability distribution;

[0157] The obtained wind and light output and system equipment failure are substituted into the DPCE agent model to calculate the system state response; the system state response includes node voltage, line power, and node electricity and hydrogen load reduction value.

[0158] Embodiment 10:

[0159] The weak link identification method of the cross-regional electric-hydrogen comprehensive energy system based on data-driven polynomial chaos expansion, the technical content of any one of embodiments 1-9, further, in step 7), the steps of identifying the weak link of the electric-hydrogen comprehensive energy system based on the risk index ranking include:

[0160] The corresponding node voltage out-of-limit risk index, line overload risk index, node electricity load reduction risk index and node hydrogen load reduction risk index are calculated through the system state response value;

[0161] The weak link of the electric-hydrogen comprehensive energy system is obtained by ranking the risk indexes.

[0162] Embodiment 11:

[0163] The weak link identification method of the cross-regional electric-hydrogen comprehensive energy system based on data-driven polynomial chaos expansion includes the following steps:

[0164] 1) Construct a hydrogen energy cross-regional complementary and coordinated electric-hydrogen comprehensive energy system, and obtain the topological parameters, equipment parameters, and operation parameters in the system;

[0165] 2) Establish an optimal energy flow calculation model of the hydrogen energy cross-regional complementary and coordinated electric-hydrogen comprehensive energy system;

[0166] 3) Based on the random fluctuation data of system source and load and the random failure data of equipment, a random variable data set is constructed, and a random variable orthogonal polynomial basis based on data multi-moment is established;

[0167] 4) The standard orthogonal polynomials are constructed by the full tensor product of the multi-dimensional random variable orthogonal polynomial basis, and the DPCE agent model of the random output response is established based on the weighted linear combination of each standard orthogonal polynomial;

[0168] 5) Based on the linear independence principle, the random variable data and the system optimal energy flow are selected to analyze the system operation state as the input and output of the DPCE agent model, respectively, and the weighted coefficients of the model are solved by using the least square method, to complete the construction of the DPCE agent model;

[0169] 6) Obtain the random power of the current electric-hydrogen comprehensive energy system source and load and the random failure state of the equipment, and calculate the system optimal energy flow state value through the DPCE agent model;

[0170] 7) Calculate the system risk index based on the product of the random probability of the system state and the severity of the system state risk, and identify the weak links of the integrated electric-hydrogen energy system based on the risk index ranking.

[0171] The aforementioned cross-regional complementary and mutually supportive integrated energy system for hydrogen energy is specifically as follows: within a single regional integrated energy system for hydrogen energy, hydrogen energy relies on the electro-hydrogen conversion equipment within the hydrogen refueling station to achieve bidirectional conversion and mutual support of heterogeneous electricity and hydrogen energy; between multiple regional integrated energy systems for hydrogen energy, electricity is transmitted across regions through power lines, and hydrogen energy is transported across regions through hydrogen long-tube trailers, achieving cross-regional complementary and mutually supportive electricity and hydrogen energy.

[0172] The parameters of the integrated hydrogen-electricity energy system for cross-regional complementarity and mutual support include: topology connection parameters of the hydrogen-electricity system, equipment parameters such as power lines, wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, hydrogen refueling machines and hydrogen tube trailers, and operating parameters such as the processing capacity of wind and solar turbines and the demand for hydrogen-electricity loads.

[0173] In step 2), the objective function of the optimal energy flow calculation model for the hydrogen-electric integrated energy system with cross-regional complementarity and mutual support is as follows:

[0174]

[0175] In the formula, C BUY,m The electricity purchase cost for the regional m system; C DG,m The cost of wind and solar power curtailment penalty for distributed power generation in the regional m system; C SHED,m To reduce the penalty cost of the hydrogen load reduction in the regional m system; C DR,m The hydrogen load demand response cost of the regional m system; C FC,m The cost of hydrogen fuel cell power generation for the regional m system; C HTT,m The operating cost of hydrogen tube trailers within the region of the m-system; C HTT,MR For the operating cost of inter-regional hydrogen tubular trailers; N m This refers to the number of regions.

[0176] In step 2), the constraints of the optimal energy flow calculation model for the hydrogen-electric integrated energy system with inter-regional complementarity and mutual support include the operating constraints of the hydrogen-electric equipment, the power balance constraints of the hydrogen-electric system, and the inter-regional transportation constraints of hydrogen energy, as shown below:

[0177] ① Operating constraints of hydrogen electrolysis equipment:

[0178] P WE,it =k1N WE,i E WE,it I WE,it

[0179]

[0180] QFC,it =H FC,it / ρ H2

[0181] In the formula, N WE,i N represents the number of electrolytic cells connected in series or parallel. FC,i P represents the number of cells connected in series and parallel in a hydrogen fuel cell system. WE,it The active power consumed by the electrolyzer for hydrogen production, I WE,it E is the operating current of the electrolytic cell. WE,it E represents the operating voltage of the electrolytic cell at time t, node i. FC,it P is the operating voltage of the hydrogen fuel cell. FC,it I represents the active power output of the hydrogen fuel cell. FC,it The operating current of the hydrogen fuel cell is k1, k2 are the conversion system, and η is the operating current of the hydrogen fuel cell. f For Faraday efficiency, ρ H2 The standard density of hydrogen (kg / Nm³) 3 ).

[0182] H HT,it =H HT,i,t-1 +(η in H HT,in,it -H HT,out,it / η out )

[0183] H DP,out,it =η DP H DP,in,it

[0184] In the formula, H HT,i,t H HT,i,t-1 η represents the hydrogen storage capacity of the hydrogen storage tank at time t and time t-1, respectively; in η out These represent the hydrogen filling efficiency and hydrogen discharging efficiency of the hydrogen storage tank, respectively; H DP,in,it The amount of hydrogen injected into the hydrogen refueling unit; H DP,out,it The amount of hydrogen output from the hydrogen refueling unit; η DP The hydrogen refueling efficiency of the hydrogen refueling machine.

[0185] ②Power balance constraint of hydrogen electricity:

[0186]

[0187] U i,t -U j,t ≤M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t )

[0188] U i,t -Uj,t ≥-M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t )

[0189] In the formula, P ij,t Q ij,t Let P be the active power and reactive power of branch ij at time t, respectively; Grid,jt Q Grid,jt P represents the active and reactive power of the hydrogen energy system interacting with the distribution network at time t, i.e., node j; load,jt Q load,jt These represent the active and reactive power of the electrical load at time t, respectively, at node j; L q,jt r represents the reactive power reduction of the electrical load at time t at node j; ij x ij These are the resistance and reactance values ​​of branch ij, respectively; U i,t The square of the voltage value at time t at node i;

[0190] P Grid,it +P WG,it +P PV,it +P FC,it =P WE,it +P load,it -L p,it i∈Ω s

[0191] Q Grid,it +Q WG,it +Q PV,it =Q load,it -L q,it i∈Ω s

[0192] In the formula, P WG,it Q WG,it P represents the actual active power and reactive power output of the wind turbine at time t, i and t, respectively, at the electro-hydrogen coupling node i; PV,it Q PV,it P represents the actual active power and reactive power output of the photovoltaic unit at time t, i and t, respectively, at the electro-hydrogen coupling node i; FC,it P WE,it These represent the active power output of the hydrogen fuel cell at time t, which is the active power consumed by the electrolyzer at time i, respectively.

[0193]

[0194] In the formula, H load,it The hydrogen load demand of hydrogen refueling station nodes;

[0195] 0≤Lp,it ≤P load,it

[0196] 0≤L q,it ≤Q load,it

[0197] 0≤L h,it ≤H load,it

[0198] ③ Hydrogen energy inter-regional transportation constraints:

[0199]

[0200]

[0201] In the formula, A c,it Let h be the position state variable indicating whether the c-th hydrogen tube trailer is located at the i-th hydrogen refueling station or road (r=0) at time t; c,it Let π(j,i) represent the positional change of the c-th hydrogen tube trailer at time t. π(j,i) represents the travel distance of the hydrogen tube trailer between two hydrogen refueling stations within a certain area, and v represents the average speed of the hydrogen tube trailer. Indicates rounding up. H c,t H c,t-1 Let be the hydrogen storage capacity of the hydrogen storage tank of the c-th hydrogen long-tube trailer at time t-1, and Hinc,it and Houtc,it be the hydrogen filling and discharging capacity of the hydrogen storage tank of the c-th hydrogen long-tube trailer at the hydrogen refueling station at time t-i, respectively.

[0202]

[0203] In the formula, A k,rt Let Ω be the positional state variable indicating whether the k-th hydrogen tube trailer is located at the r-th on-site hydrogen refueling station or road (r=0) at time t; RE This refers to the collection of on-site hydrogen refueling stations within all regional hydrogen supply networks. k,rt Let H represent the positional change of the k-th hydrogen-powered long-tube trailer at time t. π(p,r) represents the travel distance of the hydrogen-powered long-tube trailer between the two hydrogen refueling stations. k,t H k,t-1 The hydrogen storage capacity of the hydrogen storage tank of the k-th inter-regional hydrogen long-tube trailer at time t and time t-1 are respectively; Hink,rt and Houtk,rt are the hydrogen filling and discharging capacity of the hydrogen storage tank of the k-th inter-regional hydrogen long-tube trailer at the r-th hydrogen refueling mother station at time t, respectively.

[0204] In step 3), a random variable dataset is constructed based on the random fluctuation data of system source load and the random fault data of equipment, and an orthogonal polynomial basis for random variables based on the multi-moments of the data is established. The specific steps are as follows:

[0205] random variable ξ i One-dimensional orthogonal polynomial Defined as the sum of any polynomials from order 0 to 1:

[0206]

[0207] Among these requirements, the polynomial basis must satisfy the orthogonality condition, and the first coefficient of all polynomials must be 1.

[0208]

[0209]

[0210] In the formula, m,l={0,1,2,…,H}.

[0211] random variable ξ i k-th order origin moment μ k,i Defined as:

[0212]

[0213] Solving for arbitrary polynomial basis p (l)k,i (l=0,1,...,H) and random variable ξ i k-th order origin moment μ k,i Moment matching equations

[0214]

[0215] In step 4), an orthogonal polynomial is constructed by the full tensor product of the orthogonal polynomial basis of multidimensional random variables. A DPCE surrogate model for the random output response is then established based on a weighted linear combination of these orthogonal polynomials. The steps include:

[0216] The total tensor of a multidimensional random variable orthogonal polynomial basis is

[0217]

[0218] In the formula, α k,i Let be the order of the i-th one-dimensional polynomial.

[0219] The random response Y is represented by the H-order truncated multidimensional DPCE model as follows:

[0220]

[0221] In the formula, c k Let Φ be the expansion term of the k-th term to be solved. k The coefficients, M is the number of terms in the expansion Y, M = (H + N)! / (H! N!), and the k-th term is the orthogonal polynomial Φ. k (k = 1, 2, ..., M) is the total tensor product of a one-dimensional polynomial.

[0222] In step 5), random variable data and the system's optimal energy flow analysis system operating state are selected as the input and output of the DPCE surrogate model based on the principle of linear independence. The weighting coefficients of the model are solved using the least squares method to complete the construction of the DPCE surrogate model.

[0223] Select linearly independent input sample combinations ξ from the dataset and calculate the system state value Y(ξ) using the optimal energy flow model.

[0224] Substitute the sample values ​​of the random variable and the values ​​of the corresponding response function into the following formula.

[0225]

[0226] ΦC=Y

[0227] Solving the weighting coefficients of the DPCE surrogate model using the least squares method

[0228]

[0229] In step 6), the random power of the source load and the random fault state of the equipment in the current integrated electric-hydrogen energy system are obtained, and the optimal energy flow state value of the system is calculated using the DPCE proxy model as shown below:

[0230] From the original dataset or its hypothesized probability distribution, obtain the wind and solar turbine output represented by continuous random variables and the system equipment faults represented by discrete random variables, and then use these additional M... s Random input data from the system is substituted into the DPCE surrogate model to calculate the system state response value. The system state response value specifically includes node voltage, line power, and node electricity and hydrogen load reduction values.

[0231] In step 7), the system risk index is calculated based on the product of the random probability of the system state and the severity of the system state risk. The weak links of the integrated electric-hydrogen energy system are identified based on the risk index ranking as follows:

[0232] The corresponding node voltage over-limit risk indicators, line overload risk indicators, node electrical load reduction risk indicators, and node hydrogen load reduction risk indicators are calculated based on the system state response values. The weak links of the integrated electric-hydrogen energy system are then identified by ranking the risk indicators.

[0233] Example 12:

[0234] A method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion, with technical content identical to any one of Examples 1-11, further includes node voltage over-limit risk assessment indicators:

[0235] The risk function for the severity of voltage exceedance at node i is shown below:

[0236]

[0237] in,

[0238]

[0239] In the formula, U i ω is the actual value of the voltage at node i; i The upper and lower limits of the voltage at node i [U i,min U i,max The width of ]; γ is the proportional coefficient of the allowable voltage deviation range.

[0240] The Risk Index of Electrical Voltage Violation (RIEVV) for node i is shown in the following formula:

[0241]

[0242] Line overload risk assessment indicators:

[0243] The risk function for the severity of overload on line l is shown below:

[0244]

[0245] In the formula, P l P represents the transmission power value of line l. l,max This represents the maximum transmission power of line l.

[0246] The Risk Index of Electrical Line Overload (RIELO) is shown in the following formula:

[0247]

[0248] Risk assessment indicators for nodal load shedding:

[0249] The risk function for the severity of load shedding at node i is shown below:

[0250] Sev(L p,i ) = L p,i

[0251] In the formula, L p,i Let be the amount of electrical load reduction at node i.

[0252] The Risk Index of Electrical Load Loss (RIELL) for node i is shown in the following formula:

[0253]

[0254] Risk assessment indicators for node hydrogen load reduction:

[0255] The risk function for the severity of hydrogen load reduction at node i is shown below:

[0256] Sev(L h,i ) = L h,i

[0257] In the formula, L h,i Let be the amount of hydrogen load reduction at node i.

[0258] The Risk Index of Hydrogen Load Loss (RIHLL) for node i is shown in the following formula:

[0259]

[0260] Example 13:

[0261] See Figure 4 A simulation example of a data-driven polynomial chaotic expansion-based method for identifying weak links in a cross-regional integrated electric-hydrogen energy system is presented. A three-region interconnected integrated electric-hydrogen energy test system is constructed based on IEEE 15-node and IEEE 33-node power distribution networks and regional hydrogen supply networks to verify the accuracy and superiority of the proposed DPCE-based method for identifying weak links in a cross-regional integrated electric-hydrogen energy system. The following three methods are compared and analyzed:

[0262] Method 1: Sequential Monte Carlo simulation;

[0263] Method 2: The proposed data-driven second-order chaotic polynomial expansion method (second-order DPCE method)

[0264] Method 3: The proposed data-driven third-order chaotic polynomial expansion method (third-order DPCE method)

[0265] Using the calculation results of Method 1 as the benchmark, the effectiveness of the DPCE method proposed in this paper is compared and analyzed to verify the accuracy and speed of the proposed method. Figure 5 The mean relative errors of the node voltage, branch power, and hydrogen load reduction values ​​obtained by Method 2 and Method 3 are shown. Table 1 compares the calculation times of Method 1, Method 2, and Method 3.

[0266] Compared with Monte Carlo simulation, the maximum computational error of the second-order DPCE method exceeds 10%, while the computational error of the third-order DPCE method is less than 5%. The third-order DPCE method exhibits better fitting results by retaining more higher-order orthogonal polynomial bases.

[0267] In Table 1, T1 represents the time required to calculate the response values ​​of the M sets of random input variables, T2 represents the time required to construct the chaotic polynomial from the multiple moments of the original dataset, and T3 represents the time required to calculate the system's power supply reliability assessment index. total To calculate the total time required.

[0268] Table 1 Computation time for different solution methods

[0269]

[0270] As shown in Table 4.1, compared with Method 1, the DPCE method proposed in this paper takes less time (T) to calculate the total time. total = Approximately 2580s, far lower than the total calculation time T of the Monte Carlo simulation method. total =25980s, mainly because: the sample set of input random variables is selected based on the principle of linear independence, avoiding repeated and invalid optimal power flow calculations, and reducing the computational complexity when constructing the DPCE method. Therefore, the computation time T1 of the DPCE method is much smaller than the computation time T3 of the Monte Carlo simulation method. Furthermore, the DPCE method only focuses on the mapping relationship between input and output random variables. The response value of any random input can be directly calculated through the analytical expression of this mapping relationship, without needing to call the optimal power flow model again for system operation state analysis, thus significantly improving the solution efficiency. Therefore, the total computation time Ttotal of the DPCE method is much smaller than that of the Monte Carlo simulation method.

[0271] Since both the second-order and third-order DPCE methods use the same M sets of samples, their T1 time is the same. However, compared to the second-order expansion, the third-order expansion requires the computation of more orthogonal polynomial bases, therefore the T2 and T3 times of the third-order DPCE method are both longer than those of the second-order DPCE method. Considering both computational error and computation time, the third-order DPCE method is chosen for the weak link identification problem.

[0272] Based on the third-order DPCE method, this section identifies and analyzes the weak links of the following three simulation schemes (Case 1 to Case 3), and explores the impact of different energy cross-regional support mechanisms on the weak links of the integrated electric-hydrogen energy system.

[0273] Case 1: No exchange of electricity or hydrogen energy between regions;

[0274] Case 2: Only electrical energy exchange occurs between regions;

[0275] Case 3: The method presented in this paper, namely, the interaction of electric and hydrogen energy between regions.

[0276] The voltage over-limit risk of each node in the system was assessed in Cases 1 to 3. The 12 nodes with larger risk index values ​​were regarded as the weak links of the system, as shown in Table 2.

[0277] Table 2 Ranking of Voltage Exceedance Risk in Integrated Electric-Hydrogen Energy Systems (Case 1-3)

[0278]

[0279] As shown in Table 2, most nodes with high voltage exceedance risks are located at the ends of distribution network lines, such as nodes 13 to 18, 31 to 33, and 50 to 51. In Case 2, inter-regional energy interaction via power tie lines was considered. In fault scenarios, power tie lines can support the electrical load supply, thereby reducing the number of H2G mode activations at hydrogen refueling mother stations in each region. Therefore, the reduced electrical load generated by the hydrogen production demand at the hydrogen refueling mother station, to some extent, prevented voltage drops at nodes 15, 42, and their line-end nodes caused by excessive electrical load demand. In Case 3, inter-regional electrical and hydrogen energy interaction is achieved through power tie lines and hydrogen pipeline trailers. In some fault scenarios, hydrogen energy can be provided to other regions. Therefore, the hydrogen production at the hydrogen refueling mother station increases, and the electrolyzer, acting as an electrical load, causes voltage drops at nodes 15, 42, and their line-end nodes, increasing the risk of node voltage exceedances.

[0280] The overload risk of each line in the system was assessed in Cases 1 to 3. The eight lines with higher risk index values ​​were regarded as the weak links in the system, as shown in Table 3.

[0281] Table 3. Ranking of Overload Risk in Integrated Hydrogen-Electricity Energy Systems (Case 1-3)

[0282]

[0283] As shown in Table 3, the branches with higher overload risk values ​​are those connecting to hydrogen refueling mother stations, such as lines 6-26, 39-40, and 40-41, or lines at the beginning of the distribution network, such as lines 1-2, 2-3, 34-35, and 67-68. Because Cases 2 and 3 considered inter-regional energy interaction mechanisms, the production capacity of equipment such as electrolyzers and hydrogen fuel cells was increased to ensure a reliable supply of electricity and hydrogen loads under fault scenarios. This resulted in increased power flowing through the lines, thus increasing the overload risk of branches such as lines 6-26 and 39-40.

[0284] Assess the electrical load reduction risk of each node in the system in Cases 1 to 3, and identify the 10 nodes with higher risk index values ​​as the weak links in the system, as shown in Table 4.

[0285] As can be seen from the results in Table 4, Case 2 and Case 3 take into account the energy interaction between regions, effectively ensuring the supply demand of electrical load in the system under fault scenarios, thereby reducing the risk of reduction in electrical load at each node.

[0286] Table 4. Ranking of Electricity Load Reduction Risks in Integrated Electricity-Hydrogen Energy Systems (Case 1-3)

[0287]

[0288] Table 5 shows the hydrogen load reduction risk indicators for each hydrogen refueling station in Cases 1 through 3. Among them, hydrogen refueling stations with higher hydrogen load reduction risk indicators are considered as weak links.

[0289] Table 5. Ranking of Hydrogen Load Reduction Risks in Integrated Electric-Hydrogen Energy Systems (Case 1-3)

[0290]

[0291] As shown in Table 5, in Cases 1 and 2, the hydrogen load reduction risk is generally higher at hydrogen refueling substation nodes. This is because in the mother-daughter station hydrogen refueling mode, the substations do not have on-site hydrogen production capacity and can only obtain hydrogen from the mother station via hydrogen tube trailers. When the hydrogen production capacity of the mother station is insufficient, priority will be given to meeting the hydrogen load demand of the mother station to maintain the economic operation of the system. In Case 3, although some mother stations have high risk rankings, the overall hydrogen load reduction risk level of all hydrogen refueling stations in the system is better than in Cases 1 and 2. This is because in Case 3, the introduction of the inter-regional hydrogen energy exchange mechanism allows for more efficient utilization of the hydrogen production capacity of the mother station and the transportation capacity of the hydrogen tube trailers, thereby significantly reducing the risk of hydrogen load reduction at the nodes.

Claims

1. A method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion, characterized in that, Includes the following steps: 1) Construct a cross-regional complementary and mutually supportive integrated hydrogen-electric energy system and obtain the parameters of the integrated hydrogen-electric energy system; 2) Establish an optimal energy flow calculation model for a hydrogen-electric integrated energy system with cross-regional complementarity and mutual support; 3) Based on the historical random fluctuation data of system source load and the historical random fault data of equipment, construct a random variable dataset and establish an orthogonal polynomial basis of random variables based on the multi-moments of the data; 4) Construct orthogonal polynomials by the full tensor product of the orthogonal polynomial basis of random variables, and establish a DPCE surrogate model for random output response based on the weighted linear combination of each orthogonal polynomial. 5) Random variable data are selected based on the principle of linear independence, and the system state response is calculated using the optimal energy flow calculation model of the integrated electric-hydrogen energy system; Using random variable data and system state response as input and output, the weighting coefficients of the DPCE surrogate model are solved by the least squares method, thereby constructing a DPCE surrogate model for calculating the optimal energy flow state of the system. 6) Obtain the current random power of the source load and random fault status of the equipment in the integrated electric-hydrogen energy system, and calculate the optimal energy flow state value of the system through the DPCE proxy model; 7) Calculate the system risk index based on the product of the random probability of the system state and the severity of the system state risk, and identify the weak links of the integrated electric-hydrogen energy system based on the risk index ranking.

2. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on DPCE according to claim 1, characterized in that, In the aforementioned inter-regional complementary and mutually supportive integrated hydrogen energy system, hydrogen energy within a single regional integrated hydrogen energy system is converted into bidirectional heterogeneous energy (electricity and hydrogen) using hydrogen refueling station conversion equipment; electrical energy between multiple regional integrated hydrogen energy systems is transmitted across regions via power lines, and hydrogen energy is transported across regions via hydrogen pipeline trailers.

3. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on DPCE according to claim 1, characterized in that, The parameters of the integrated electric-hydrogen energy system include: the topology connection parameters of the electric-hydrogen system, the parameters of power lines, wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, hydrogen refueling machines and hydrogen tube trailers, and the parameters of wind and solar turbine processing and electric-hydrogen load requirements.

4. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on DPCE according to claim 1, characterized in that, The objective function minC of the optimal energy flow calculation model for the hydrogen energy cross-regional complementary and mutually supportive integrated energy system is shown below: In the formula, C BUY,m The electricity purchase cost for the regional m system; C DG,m The cost of wind and solar power curtailment penalty for distributed power generation in the regional m system; C SHED,m To reduce the penalty cost of the hydrogen load reduction in the regional m system; C DR,m The hydrogen load demand response cost of the regional m system; C FC,m The cost of hydrogen fuel cell power generation for the regional m system; C HTT,m The operating cost of hydrogen tube trailers within the region of the m-system; C HTT,MR For the operating cost of inter-regional hydrogen tubular trailers; N m This refers to the number of regions.

5. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on DPCE according to claim 1, characterized in that, The constraints of the optimal energy flow calculation model for the hydrogen-electric integrated energy system with cross-regional complementarity and mutual support include the operating constraints of hydrogen-electric equipment, the power balance constraints of hydrogen-electricity, and the cross-regional transportation constraints of hydrogen energy. The operating constraints of the hydrogen electrolysis equipment are as follows: P WE,it =k1N WE,i E WE,it IN WE,it (2) Q FC,it =H FC,it / ρ H2 (7) H HT,it =H HT,i,t-1 +(the in H HT,in,it -H HT,out,it / or out (8) H DP,out,it =the DP H DP,in,it (9) In the formula, N WE,i N represents the number of electrolytic cells connected in series or parallel. FC,i P represents the number of cells connected in series and parallel in a hydrogen fuel cell system. WE,it Q WE,it The active and reactive power consumed in hydrogen production by the electrolyzer; I WE,it E is the operating current of the electrolytic cell. WE,it E represents the operating voltage of the electrolytic cell at time t, node i. FC,it P is the operating voltage of the hydrogen fuel cell. FC,it Q FC,it The active power and reactive power output of the hydrogen fuel cell; I FC,it The operating current of the hydrogen fuel cell is k1, k2 are the conversion system, and η is the operating current of the hydrogen fuel cell. f For Faraday efficiency, ρ H2 H is the standard density of hydrogen gas. HT,i,t H HT,i,t-1 η represents the hydrogen storage capacity of the hydrogen storage tank at time t and time t-1, respectively; in η out These represent the hydrogen filling efficiency and hydrogen discharging efficiency of the hydrogen storage tank, respectively; H DP,in,it The amount of hydrogen injected into the hydrogen refueling unit; H DP,out,it The amount of hydrogen output from the hydrogen refueling unit; η DP The hydrogenation efficiency of the hydrogenation unit is Δt; the time step is H. WE,it H is the hydrogen production rate of the electrolyzer; F is the Faraday constant; H HT,in,it H HT,out,it For the injection and output of hydrogen gas into the hydrogen storage tank; H FC,it Hydrogen consumption of hydrogen fuel cells; The power balance constraints for hydrogen electrolysis are shown below: U i,t -U j,t ≤M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t ) (14) U i,t -U j,t ≥-M(1-α ij,t )+2(r ij P ij,t +x ij Q ij,t ) (15) P Grid,it +P WG,it +P PV,it +P FC,it =P WE,it +P load,it -L p,it i∈Ω s (16) Q Grid,it +Q WG,it +Q PV,it =Q load,it -L q,it i∈Ω s (17) 0≤L p,it ≤P load,it (23) 0≤L q,it ≤Q load,it (24) 0≤L h,it ≤H load,it (25) In the formula, P ij,t Q ij,t Let P be the active power and reactive power of branch ij at time t, respectively; Grid,jt Q Grid,jt P represents the active and reactive power of the hydrogen energy system interacting with the distribution network at time t, i.e., node j; load,jt Q load,jt These represent the active and reactive power of the electrical load at time t, respectively, at node j; L p,it L q,jt Let r be the active and reactive power reduction of the electrical load at time t at node j; ij x ij These are the resistance and reactance values ​​of branch ij, respectively; U i,t P is the square of the voltage value at time t at node i; WG,it Q WG,it P represents the actual active power and reactive power output of the wind turbine at time t, i and t, respectively, at the electro-hydrogen coupling node i; PV,it Q PV,it P represents the actual active power and reactive power output of the photovoltaic unit at time t, i and t, respectively, at the electro-hydrogen coupling node i; FC,it P WE,it H represents the active power output of the hydrogen fuel cell at time t, which is the active power consumed by the electrolyzer at time i, respectively; load,it The hydrogen load demand of hydrogen refueling station nodes; These represent the amount of hydrogen added and the amount of hydrogen released from the hydrogen storage tank of the c-th hydrogen long-tube trailer at the hydrogen refueling station at time t i, respectively. Let M and α represent the hydrogen filling and discharging amounts of the hydrogen storage tank of the k-th inter-regional hydrogen tube trailer at the r-th hydrogen refueling station at time t, respectively; M is a sufficiently large integer; α ij,t Let L be a binary variable, representing the switching state of surface branch ij at time t; h,it This refers to the reduction in hydrogen load at hydrogen refueling station nodes; The cross-regional transportation constraints for hydrogen energy are as follows: In the formula, A c,it Let A be the positional state variable indicating whether the c-th hydrogen tube trailer is located at the i-th hydrogen refueling station or road at time t; c,0t Let h be the position state variable indicating whether the c-th hydrogen-powered long-tube trailer is located on the road at time t; c,it h c,0t Let π(j,i) represent the positional change of the c-th hydrogen tube trailer at time t; π(j,i) represent the travel distance of the hydrogen tube trailer between two hydrogen refueling stations within a certain area; and v represents the average speed of the hydrogen tube trailer. Indicates rounding up. H c,t H c,t-1 Let be the amount of hydrogen stored in the hydrogen storage tank of the c-th hydrogen long-tube trailer at time t-1, respectively. Let A represent the amount of hydrogen added to and the amount of hydrogen released from the hydrogen storage tank of the c-th hydrogen tube trailer at the hydrogen refueling station at time t; k,rt Let A be the positional state variable indicating whether the k-th hydrogen tube trailer is located at the r-th on-site hydrogen refueling station or road at time t; k,0t Let Ω be the positional state variable indicating whether the k-th hydrogen-powered long-tube trailer is located on the road at time t; RE This refers to the collection of on-site hydrogen refueling stations within all regional hydrogen supply networks. k,rt h k,0t Let H represent the positional change of the k-th hydrogen-powered long-tube trailer at time t. π(p,r) represents the travel distance of the hydrogen-powered long-tube trailer between the two hydrogen refueling stations. k,t H k,t-1 The amount of hydrogen stored in the hydrogen storage tank of the k-th inter-regional hydrogen tube trailer at time t and time t-1, respectively; These represent the amount of hydrogen added and the amount of hydrogen released from the hydrogen storage tank of the k-th inter-regional hydrogen tube trailer at the r-th hydrogen refueling station at time t.

6. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion as described in claim 1, characterized in that, Step 3) involves establishing an orthogonal polynomial basis of random variables based on the multiple moments of the data. 3.1) For any random variable ξ in the random variable dataset i Define a one-dimensional orthogonal polynomial and k-th order origin moment μ k,i ,Right now: m k,i =∫ξ i k d(ξ i ) (39) In the formula, P i (l) (ξ i () is a one-dimensional orthogonal polynomial; d(ξ) represents the polynomial coefficients. i ) represents a probability distribution; Among them, one-dimensional orthogonal polynomial The following constraints must be met: In the formula, m,l={0,1,2,…,H}; 3.2) Solving for arbitrary polynomial basis p (l)κ,i With random variable ξ i k-th order origin moment μ k,i The moment matching equations are used to obtain the coefficients of the one-dimensional orthogonal polynomial basis; The moment matching equation is as follows: 3.3) Repeat steps 3.1)-3.2) to obtain the orthogonal polynomial basis for each random variable in the random variable dataset, thereby constructing an orthogonal polynomial basis for random variables based on the multi-moments of the data.

7. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion as described in claim 1, characterized in that, Step 4) involves establishing a DPCE proxy model for random output response, including the following steps: 4.1) Calculate the full tensor Φ of the orthogonal polynomial basis of multidimensional random variables. κ (ξ1,ξ2,...,ξ N ),Right now: In the formula, α κ,i Let be the order of the i-th one-dimensional polynomial; Based on the full tensor of the orthogonal polynomial basis of random variables, a DPCE surrogate model Y(ξ1,ξ2,...,ξ) for random output response is established. N ),Right now: In the formula, c κ Let Φ be the expansion term of the κ-th term to be solved. κ The coefficients, M is the number of terms in the expansion Y, M = (H+N)! / (H!N!), and the κ-th term is the orthogonal polynomial Φ. κ (κ=1,2,…,M) is the full tensor product of a one-dimensional polynomial.

8. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion as described in claim 1, characterized in that, Step 5) includes the following steps for the DPCE agent: 5.1) Select linearly independent input sample combinations ξ from the random variable dataset and calculate the system state response Y(ξ) using the optimal energy flow model; 5.2) Substitute the sample values ​​of the random variables and the values ​​of the corresponding response functions into equations (45)-(46) and use the least squares method to solve for the weighting coefficients of the DPCE surrogate model; ΦC=Y(46) The weighting coefficients of the DPCE proxy model are shown below: In the formula, This is the weighted coefficient vector.

9. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion as described in claim 1, characterized in that, Step 6), the steps of calculating the optimal energy flow state value of the system using the DPCE surrogate model, include: Obtain wind and solar turbine output represented by continuous random variables and system equipment faults represented by discrete random variables from the original dataset or hypothesized probability distribution; The obtained wind and solar turbine output and system equipment faults are substituted into the DPCE proxy model to calculate the system state response; the system state response includes node voltage, line power, and node electricity and hydrogen load reduction values.

10. The method for identifying weak links in a cross-regional integrated electric-hydrogen energy system based on data-driven polynomial chaotic expansion as described in claim 1, characterized in that, Step 7), which involves identifying the weak links in the integrated electric-hydrogen energy system based on risk indicator ranking, includes the following steps: The corresponding node voltage over-limit risk index, line overload risk index, node electrical load reduction risk index, and node hydrogen load reduction risk index are calculated based on the system status response values. Weak links in the integrated electric-hydrogen energy system were identified by ranking risk indicators.

Citation Information

Cited By

  • High-permeability distributed photovoltaic system operation safety analysis method and system

    CN121939501A