Hydrogen refueling station-power distribution network double-layer low-carbon collaborative planning method considering carbon emission flow

Abstract

The invention discloses a hydrogen refueling station-power distribution network double-layer low-carbon collaborative planning method considering carbon emission flow. The method comprises the following steps: constructing a carbon emission flow model of a power distribution network and a hydrogen refueling station; introducing a wind power plant construction decision as renewable energy supplement, calculating the carbon emission intensity corresponding to the input power of the hydrogen refueling station, and realizing quantification of the carbon emission responsibility of the hydrogen refueling station; establishing a traffic flow distribution model, and obtaining road and runoff distribution in the traffic network; establishing a hydrogen refueling station positioning model, screening candidate stations meeting hydrogen refueling requirements, and evaluating the operation risk of the hydrogen refueling station; establishing a hydrogen refueling station-power distribution network bilevel planning model considering the carbon emission constraint; and performing linearization processing on the nonlinear target and the constraint, performing iterative solution on the hydrogen refueling station-power distribution network double-layer planning model until a power load between two iterations of the hydrogen refueling station meets a convergence requirement, and outputting a final wind power plant, power distribution network construction and hydrogen refueling station collaborative planning scheme. According to the method, the low-carbon property and the system adaptability of the hydrogen energy traffic infrastructure layout are improved.

Description

Technical Field

[0001] The present invention relates to the field of low-carbon economic dispatching of power systems, and in particular to a double-layer low-carbon collaborative planning method for hydrogen refueling stations and distribution networks that considers carbon emission flows. Background Art

[0002] As a zero-carbon emission clean energy carrier, hydrogen energy has received widespread attention worldwide in recent years, especially in the transportation sector, showing great potential to replace traditional fossil fuels. As the core infrastructure of the hydrogen energy transportation system, hydrogen refueling stations not only undertake the function of hydrogen fuel refueling, but also are one of the important paths to achieve deep decarbonization in the transportation sector. The rational planning and construction of hydrogen refueling station networks can not only improve the operating convenience and coverage of hydrogen fuel cell vehicles (HPVs), but also effectively promote the coordinated development of the entire hydrogen energy industry chain, thereby accelerating the low-carbon transformation of the transportation energy structure. However, when studying the site selection and layout optimization of hydrogen refueling stations, there are still many shortcomings in low-carbon site selection planning.

[0003] Existing research on the optimization of hydrogen refueling station site selection often focuses only on factors such as traffic convenience, construction cost, hydrogen production method, and hydrogen supply capacity. For example: Reference [1]: Wang Shuzheng, Shan Tingting, Zhao Yang, et al. Highway hydrogen refueling station layout planning considering different hydrogen production methods [J]. Electric Power Automation Equipment, 2024, 44(04): 9-17. DOI: 10.16081 / j.epae.202310023. A comprehensive evaluation of different hydrogen production facilities and locations is conducted, and after selecting the optimal hydrogen production method, the layout and capacity of hydrogen refueling stations on the road section are planned and configured. However, taking highways as the application scenario, the traffic flow path is relatively simple and the network structure is relatively regular. It is difficult to fully reflect the practical problems faced by the layout of hydrogen refueling stations in complex urban or regional transportation networks, such as multi-source traffic demand, multi-path flow, and dynamic coupling of supply and demand.

[0004] Reference [2]: Zhao Yuanfa, Si Yang, Ma Linrui, et al. Site selection and sizing of hydrogen production and refueling stations under the framework of transportation network-power grid coupling [J]. Acta Energiae Solaris Sinica, 2024, 45(08): 54-62. DOI: 10.19912 / j.0254-0096.tynxb.2023-0619. A multi-objective evaluation model for site selection and sizing of hydrogen refueling stations is constructed from the transportation network, power grid and economic levels to facilitate the development of new power systems with new energy grid connection and growing electric hydrogen load. However, it is overlooked that hydrogen refueling stations also belong to the electricity consumption link. In the planning process, it is necessary to fully consider their carbon emissions to select sites. That is, the process of hydrogen production by electrolysis of water is highly dependent on power grid power supply, which will indirectly generate carbon emissions. The temporal and spatial differences in the carbon intensity of electricity will cause the cleanliness of hydrogen energy to drift.

[0005] Carbon Emission Flow (CEF) technology is the core method for quantifying the dynamic carbon footprint of power systems. By coupling power flow and carbon flow, it can accurately trace the carbon emission responsibility of the entire process of electricity production, transmission and consumption, and provide technical support for the connection between low-carbon goals and power system operation. Reference [3]: Zhang Xiaoyan, Wang Luyu, Huang Lei, et al. Integrated energy optimization scheduling of parks considering extended carbon emission flow and carbon trading bargaining model [J]. Automation of Power Systems, 2023, 47(09): 34-46. An extended carbon emission flow model that takes energy storage and carbon capture equipment into consideration is proposed, so that integrated energy parks can be priced through carbon emission flow models. However, existing research on the construction of carbon emission flow models for deep coupling networks between hydrogen stations and distribution networks still has significant deficiencies. That is, as a high-energy load, the dynamic correlation between the strong time-varying hydrogen production behavior of hydrogen stations and the carbon intensity fluctuations of the power grid has not been effectively solved. In addition, the spatiotemporal nonlinear characteristics of carbon emission flows caused by the increase in renewable energy penetration in the distribution network make the quantification of carbon responsibility face the problem of multi-scale interaction. Current research lacks a systematic modeling framework, making it difficult to support the dynamic carbon flow allocation mechanism of such coupled networks, limiting the application potential of hydrogen refueling stations in the low-carbon coordinated regulation of power systems. Summary of the Invention

[0006] In order to study the power source of hydrogen refueling stations and quantify their carbon emission responsibility, the present invention proposes a two-layer low-carbon collaborative planning method for hydrogen refueling stations and distribution networks that takes into account carbon emission flows. By constructing a carbon emission flow correlation model between the distribution network and the hydrogen refueling station, and dynamically coupling the transportation network equilibrium theory with the layout requirements of the hydrogen refueling station, a low-carbon and high-efficiency site selection planning of the hydrogen energy transportation system is achieved; at the same time, an integrated two-layer optimization model of the electricity-hydrogen-transportation system is constructed to achieve collaborative planning of the distribution network and the hydrogen refueling station under carbon emission constraints, thereby solving the problem that the existing methods fail to comprehensively consider the heterogeneity of electricity carbon sources and the dynamics of the transportation network, and improving the low-carbon nature and system adaptability of the layout of hydrogen energy transportation infrastructure.

[0007] The technical solution adopted by the present invention is:

[0008] A two-tier low-carbon collaborative planning method for hydrogen refueling stations and distribution networks considering carbon emission flows includes the following steps:

[0009] Step 1: Based on the proportional sharing principle, a carbon emission flow model of the distribution network and hydrogen refueling station is constructed;

[0010] Step 2: Introduce wind farm construction decisions as a renewable energy supplement, calculate the carbon emission intensity corresponding to the hydrogen station's input electricity, and quantify the carbon emission responsibility of the hydrogen station;

[0011] Step 3: Establish a traffic flow distribution model to obtain the path flow distribution in the traffic network;

[0012] Step 4: Establish a hydrogen refueling station positioning model to screen candidate sites that meet hydrogen refueling needs and assess the operational risks of hydrogen refueling stations;

[0013] Step 5: Establish a two-layer planning model for hydrogen refueling stations and distribution networks that takes carbon emission constraints into account. The upper-layer model transmits the unit output and power flow operation results to the lower-layer model.

[0014] The lower model first calculates the carbon emission flow based on the output of the upper model, and then plans the site selection of hydrogen refueling stations based on the carbon emission intensity of the nodes;

[0015] The lower-level model then transmits the hydrogen station's power load to the upper-level model, optimizing the site selection for hydrogen stations and wind farms. This cycle continuously optimizes the upper-level model's renewable energy output and the lower-level model's load, reducing the system's overall carbon emissions.

[0016] Step 6: Linearize the nonlinear objectives and constraints in steps 1 to 5, and use the solver to iteratively solve the hydrogen station-distribution network two-level planning model until the power load between two iterations of the hydrogen station meets the convergence requirements, and output the final coordinated planning plan for wind farm, distribution network construction and hydrogen station.

[0017] In step 1, the carbon emission intensity of the grid node is proportionally shared with the carbon flow contribution of its input power, and a carbon emission flow model of the distribution network and hydrogen refueling station is constructed to quantitatively assess the actual responsibility of power consumption units such as hydrogen refueling stations in the grid carbon emissions. The specific method for constructing the carbon emission flow model of the distribution network and hydrogen refueling station is as follows:

[0018]

[0019] In the above formula, They represent the incoming and outgoing terminal sets of node i respectively; P′ mj is the contribution of the mth incoming line to the jth outgoing line; P m,i is the active power flowing through line mi; P i G is the output power of the i-th generator; P j is the active power on the jth outgoing line; P′ Gj is the contribution of the generator to the jth outgoing line; g is the incoming line node; P g is the power flow at the incoming line; R j is the carbon flow rate of the jth outlet; ρ i , ρ i G 、 are the carbon emission intensities of node i, generator at node i, and line mi respectively; R m,i 、 are the carbon flow rates of the generator on line mi and node i respectively; is the comprehensive carbon emission intensity of node j.

[0020] In step 2, after considering the carbon emission flow intensity of the hydrogen refueling station, and taking into account that new energy construction can provide high-quality low-carbon electricity to reduce overall carbon emissions, the construction decision of the new energy plant is introduced when selecting the site of the hydrogen refueling station. To facilitate the model calculation, the wind farm is used to replace the overall new energy farm, as follows:

[0021]

[0022] In formula (5), is the carbon emission intensity of node i; P i W is the output power of wind power; is the carbon emission intensity of wind power; is the carbon flow rate of the branch containing the wind farm; R i is the carbon flow rate of node i; P i is the active power of node i.

[0023] In step 3, a traffic flow distribution model under traffic network equilibrium is constructed:

[0024] Based on the Bureau of Public Roads (BPR) function, two setting conditions are proposed to reduce the difference between the BPR function and actual traffic congestion. Based on the two setting conditions, a traffic flow distribution model in UE mode is established. The specific construction method is as follows:

[0025] Setting condition ①: Road traffic flow should always be kept within its maximum carrying capacity to ensure smooth and safe traffic;

[0026] Setting condition ②: When the traffic flow on the road reaches its maximum capacity, the vehicle's speed will slow down, resulting in a delay in travel time;

[0027]

[0028] In formula (6), x a,t,s is the traffic volume on road a; t a,t,s is the actual travel time of road a; is the free travel time of road a; c a is the traffic capacity of road a; ε a,t,s is the delay time.

[0029] Based on the above setting conditions, the traffic flow distribution model under social optimality is established as follows:

[0030]

[0031] x a,t,s ≤c a [ε a,t,s ](11);

[0032]

[0033] In the above formula, is the traffic flow on path k; is the road and path association matrix; is the OD demand of HPV at time t; Q OD is the total OD demand in a day; ρ HPV is the HPV penetration rate; τ t,s is the hydrogenation rate of HPV at each time period under scenario s; γ a,t,s , ε a,t,s is the dual variable of the relevant constraints; a is the road index; s is the scenario index; k is the path index; t is the time index; OD represents the origin-destination (OD); [·] is an auxiliary variable introduced to convert the corresponding formula into a traffic assignment model under user equilibrium through subsequent Lagrangian relaxation and strong dual theory.

[0034] In step 4, based on the flow-refueling location model (FRLM), the analysis of the path expansion in the network is omitted. Two driving rules for HPV are proposed to establish a hydrogen station location model. The specific construction method is as follows:

[0035] Based on the dynamic driving behavior constraints of HPV in the traffic network, the following two driving rules are proposed:

[0036] Rule ①: On the same route, the distance between two adjacent hydrogen refueling stations must be less than the HPV's range.

[0037] Rule ②: HPV should store a certain amount of hydrogen at the starting point and end point of any journey

[0038] Rule 1 is to prevent the HPV from stopping due to fuel exhaustion during long-distance driving, ensuring the continuity of cross-region driving. Rule 2 is to ensure that the HPV has sufficient initial hydrogen at the time of departure and retains a safe fuel reserve before reaching the destination to travel to another hydrogen refueling station or the final destination.

[0039] After each hydrogen refueling, the battery state of the HPV is W=1, and the candidate location of the hydrogen refueling station is the midpoint of each road. Assuming that the cruising range of the HPV after full hydrogen storage is R (unit: km), to reach state W OEnter the traffic network and leave state W without consuming all the fuel D Leave the transportation network.

[0040] The planning of hydrogen refueling stations in this invention is mainly to meet the HPV hydrogenation demand during peak hours. Therefore, the hydrogen refueling station positioning model can be expressed as the following operating constraints:

[0041] 1) HPV can reach the first hydrogen refueling station node:

[0042]

[0043] 2) The battery status of HPV when it reaches the destination is higher than w D , to satisfy rule ②:

[0044]

[0045] 3) HPV can reach all candidate hydrogen refueling station nodes:

[0046] W a,k,s ≥0,a∈Ω k ,k∈Ω OD (15);

[0047] 4) Between two candidate hydrogen refueling stations, the HPV will not stop moving due to energy exhaustion:

[0048]

[0049] 5) HPV battery status at the candidate hydrogen station node:

[0050]

[0051] In the above formula: a, b, k are road index and path index respectively; s represents the scene; y a The construction status of the hydrogen refueling station is 1 or 0 to indicate whether the hydrogen refueling station is being built or not; is the starting road set on path k; is the set of roads close to the end point on path k; Ω OD is a set of paths; are the previous and subsequent road sets on path k respectively; Ω k is the set of roads on path k; r a 、r b is the distance between roads; W a,k,s is the HPV battery status of road a on path k under scenario s.

[0052] In step 4, after completing the site selection of the hydrogen refueling station by comprehensively considering the traffic flow, the safe operation of the hydrogen refueling station also needs to be taken into consideration. To this end, the TNT equivalent method is used as a safe operation indicator to measure the energy generated by the explosion of the hydrogen storage tank, thereby realizing the assessment of the operating risk of the hydrogen refueling station.

[0053] The explosion energy of hydrogen storage tank is:

[0054]

[0055] In formula (19), E a,t,s is the explosion energy of the hydrogen storage tank; is the mass of hydrogen in the hydrogen storage tank; α is the explosion coefficient; E A is the TNT equivalent coefficient of the explosion; E f is the calorific value of hydrogen; E TNT It is the explosion heat generated when TNT explodes;

[0056] Set the hydrogen storage tank explosion energy warning value E according to the explosion energy war and the limit value E max When the hydrogen storage tank is operating between the two, it is in a dangerous operating state. Therefore, the safety constraint of the hydrogen refueling station is set as:

[0057] 0≤E a,t,s ≤E war +I a,t,s M (20);

[0058] E war +(I a,t,s -1)M≤E a,t,s ≤E max (twenty one);

[0059] In the above formula: I a,t,s It is a binary variable, and its value is 1 or 0, indicating whether the hydrogen station is operating in the warning range. war ,E max ]; M takes the maximum value.

[0060] In step 5, a two-level planning model of hydrogen refueling station and distribution network taking into account carbon emission constraints is established:

[0061] The upper-level model is for distribution network planning: with the goal of minimizing total cost, it optimizes distribution network construction and renewable energy layout, reduces the overall carbon intensity of the grid, and dynamically updates the carbon emission intensity of each node in the grid based on power flow calculations and carbon emission flow models, providing a low-carbon power environment for the lower-level model.

[0062] The lower-level model is for hydrogen refueling station planning: based on the node carbon intensity data output by the upper-level model, with the goal of jointly minimizing the operation and maintenance costs and carbon emission costs of hydrogen refueling stations, the site selection and hydrogen production behavior of hydrogen refueling stations are optimized to ensure that they match the real-time electricity carbon intensity and meet carbon emission limits and safe operation requirements.

[0063] The specific construction method is as follows:

[0064] 1. Objective function of the upper model:

[0065] The upper model is mainly used to realize the site selection and planning of power grid expansion and new energy, so the objective function is set to minimize the total cost, including the investment cost F S,In and operating costs F S,Op :

[0066] min f S =F S,In +F S,Op +F L,E (twenty two);

[0067] In formula (22), f S is the total cost of the upper model; F L,E represents the cost other than electricity purchase in the lower model;

[0068] Investment cost F S,In It also includes the cost of line expansion and the construction cost of new energy stations:

[0069]

[0070] In the above formula, γ is the equipment annual equivalent coefficient; θ is the discount rate; Y is the planning period; w1 is the cost of distribution line expansion; is the number of line expansions; x i It is a binary variable, representing the construction status of new energy. When the value is 1 or 0, it means that new energy stations are being built or not. W is the unit capacity investment cost of new energy equipment; P i W,rated represents the wind power construction capacity of node i; Ω DN is the set of distribution network nodes; Ω W is the candidate location set of the wind farm station; m is the distribution network line index; i is the distribution network node index;

[0071] Operating costs mainly include electricity purchase costs and equipment maintenance costs:

[0072]

[0073] In formula (25), F Op is the operating cost of the upper model; βM is the equipment maintenance factor; D s is the probability corresponding to the scene; The power of purchased electricity; is the time-of-use electricity price; Ω s is a typical scene set; Ω T is the time set; △t is the time interval.

[0074] 2. Constraints of the upper model:

[0075] The safe operation of the distribution network is the primary requirement for the upper-level model. Specifically, load changes in the lower-level model will affect the magnitude and direction of the power flow, and thus the carbon emission flow. After the hydrogen refueling station is connected to the distribution network as a load, it will be affected by the HPV traffic flow. The fluctuation of its electricity demand will affect the system power flow. Using the Distflow model to calculate the power flow and voltage distribution of the distribution network, the following constraints are imposed:

[0076] 1) Power balance constraints:

[0077]

[0078] In the above formula, n is the distribution network node; s(i) is the set of head-end nodes corresponding to the line with end node i; e(i) is the set of end nodes corresponding to the line with end node i; is the wind power active power; P m,i,t,s , Q m,i,t,s are the active and reactive power flowing through line mi respectively; P i,n,t,s , Q i,n,t,s are the active and reactive power flowing through line ni respectively; Divided into basic active load and reactive load; is the load of the hydrogen refueling station.

[0079] 2) Voltage drop constraints after expansion:

[0080] When the existing line cannot meet the planned power flow, it is necessary to consider expanding the line capacity. The new distribution line is built in parallel with the original line and uses the same parameters as the original line. The voltage drop constraint of the expanded line is:

[0081]

[0082] In formula (28):

[0083] U i,t,s is the node voltage of node i of line mi at time t under scenario s; U m,t,s is the node voltage of node m of line mi at time t under scenario s; New variables introduced for linearization; R m,iis the resistance of circuit mi; ​​X m,i is the reactance of line mi; U1 is the voltage at node 1; D DN Indicates the maximum number of distribution network lines that can be expanded; u represents the segment index of the number of power grid line expansions.

[0084] Linearize the voltage drop constraint:

[0085]

[0086] In the above formula:

[0087] D DN Indicates the maximum number of distribution network lines that can be expanded; Replace the introduced binary variable Introducing new variables Linearization is performed; U1 is the voltage at node 1; U i,t,s is the voltage at node i at time t; is the upper limit of node i voltage; is the upper limit of the voltage at node m;

[0088] 3) Node voltage constraints:

[0089]

[0090] In formula (33): are the upper and lower limits of the voltage at the distribution network node i, respectively.

[0091] 4) Distribution network carbon emission cap:

[0092]

[0093] In formula (34):

[0094] Provide a reference for carbon emission intensity for the upper-level power grid; is the total carbon emissions of the distribution network; D S is the probability of scenario s occurring; is the electricity purchase power at time t under scenario s; ρ W is the carbon emission intensity of the wind farm; is the wind power of node i at time t under scenario s.

[0095] 5) Constraints on new energy equipment:

[0096]

[0097] Where,

[0098] is the predicted per-unit output value of the wind turbine; N is the maximum capacity that can be built for a single wind farm; w is the number of wind farms to be built; M W,rated is the upper limit of the total capacity construction of the wind farm; P i W,rated is the wind power plant capacity at node i; X i It is a 0-1 variable, indicating the construction status of the wind farm at node i.

[0099] 3. Objective function of the lower model:

[0100] The lower-level model is for hydrogen refueling station planning, addressing the site selection and carbon emission issues of hydrogen refueling stations. The lower-level model first needs to track the carbon footprint and calculate the node carbon intensity based on the results of the upper-level model.

[0101] When the active power P m,i When is positive, the carbon emission intensity of line mi is equal to the carbon emission intensity of node m:

[0102]

[0103] In formula (39): is the carbon emission intensity of node m on line mi; is the carbon emission intensity of node i on line mi. The objective function of the lower model consists of three parts:

[0104] 1)F L,In : The investment cost of planning and constructing a hydrogen refueling station and its equipment, as well as its safety investment cost;

[0105] 2)F L,Op : The operating costs of hydrogen refueling stations include penalty costs for dangerous operations, carbon emission costs, maintenance costs of hydrogen refueling station equipment, and the cost of purchasing electricity from the main grid;

[0106] 3) Costs other than electricity purchase in the upper model:

[0107] min f L =F L,In +F L,Op +F S,E (40);

[0108]

[0109] Where y a It is a 0-1 variable, indicating the construction status of the hydrogen refueling station;

[0110] γ L is the equal-year coefficient of the lower model;

[0111] w2 is the risk penalty coefficient;

[0112] w3 is the unit carbon emission cost;

[0113] The planned capacity of electrolyzers, hydrogen storage tanks, and safety equipment;

[0114] c HRS 、c e 、c tank 、c safe are the construction costs of hydrogen refueling stations, electrolyzers, hydrogen storage tanks, and safety equipment cost coefficients;

[0115] E quo A cap on carbon emissions for hydrogen refueling stations;

[0116] F S,E This is the cost other than electricity purchase in the upper model;

[0117] Ω A is the road set; I a,t,s It is a 0-1 variable, indicating whether the hydrogen storage tank is operating within the warning range; E a,t,s is the explosion energy of the hydrogen storage tank; β M is the equipment maintenance factor.

[0118] 4. Constraints of the lower model:

[0119] 1) Carbon emission constraints of hydrogen refueling stations:

[0120]

[0121] In formula (43): is the power consumption of the electrolytic cell; is the carbon emission intensity of the road where the hydrogen refueling station is located.

[0122] 2) Electrolyzer operation constraints:

[0123]

[0124] In formula (44): e The working efficiency of the electrolyzer; the quality of hydrogen produced for the electrolyser; is the electrolytic cell capacity; It is the upper limit of the hydrogen production capacity of the electrolyzer.

[0125] 3) Hydrogen storage tank constraints:

[0126]

[0127] In formula (45): is the mass of hydrogen in the hydrogen storage tank at time t+1; is the mass of hydrogen in the hydrogen storage tank at time t; The hydrogen demand of HPV; The upper limit of the hydrogen storage tank capacity.

[0128] 3) Hydrogenation machine operation constraints:

[0129]

[0130] In formula (46): Hydrogen demand of HPV; J a is the number of hydrogen refueling machines built; H ave is the average hydrogenation mass of the hydrogenator; δ HPV The mass of hydrogen refilled for each HPV refill; The maximum number of hydrogen refueling machines that can be built in a hydrogen refueling station; x a,t,s For traffic that requires hydrogen refueling.

[0131] The power load of the lower model needs to be transferred to the upper model for update iteration, which can be calculated by the following formula:

[0132]

[0133] In formula (47): The power consumed by the hydrogen refueling station at node i in the distribution network; Ω A(i) is the set of roads associated with distribution network node i.

[0134] The proposed two-tier planning model for hydrogen refueling stations and distribution networks, which takes carbon emission constraints into account, meets stringent carbon emission requirements by optimizing the distribution system grid, constructing new energy units, and planning hydrogen refueling station sites. Therefore, the planning model solution process is as follows:

[0135] Step 1: Input the basic system parameters and the initial load of the hydrogen station into the upper model;

[0136] Step 2: The upper model constructs new energy and distribution network lines based on hydrogen station load changes and carbon emission requirements. Step 3: The power flow data and unit output data obtained by the upper model are passed to the carbon emission flow model of the lower model to calculate the carbon emission intensity of each node.

[0137] Step 4: The lower model further guides the site selection and construction of hydrogen refueling stations and hydrogen production behavior based on the carbon emission intensity of the nodes, while meeting the requirements of safety and low carbon.

[0138] Step 5: Determine whether the power load of the hydrogen refueling station between two iterations meets the requirements of the following formula. If so, output the planning scheme. If not, transmit the power load of the hydrogen refueling station to the upper model and iterate and solve from step 2.

[0139]

[0140] In formula (48): is the tolerance, z is the number of iterations; The power consumed by the hydrogen refueling station at the zth iteration; The electric power consumed by the hydrogen refueling station during the z-1th iteration.

[0141] In step 6, some constraints have been linearized in the above steps. In this step, the constraints of the traffic flow distribution model and the hydrogen station positioning model are linearized. The specific construction method is as follows:

[0142] Since the traffic flow distribution model established in step 3 cannot meet the constraints when the traffic flow is too large, the Lagrange relaxation principle and strong duality theory are used to eliminate the duality gap and finally transform it into a linearized traffic distribution model under user equilibrium;

[0143]

[0144] In formula (49), is the traffic flow on path k; is the road and path association matrix; is the OD demand of HPV at time t; Q OD is the total OD demand in a day; ρ HPV is the HPV penetration rate; τ t,s is the hydrogenation rate of HPV at each time period under scenario s; γ a,t,s , ε a,t,s is the dual variable of the relevant constraint.

[0145] The nonlinear operation constraints in the hydrogen station positioning model can be linearized using the large M method, and the auxiliary variable ζ is introduced. a,k,s , so that the MINLP problem is converted to a MILP problem:

[0146]

[0147] In formula (50), a, b, and k are the road index and path index respectively; s represents the scene; y a The construction status of the hydrogen refueling station is 1 or 0 to indicate whether the hydrogen refueling station is being built or not; is the starting road set on path k; is the set of roads close to the end point on path k; Ω OD is a set of paths; are the previous and subsequent road sets on path k respectively; Ω k is the set of roads on path k; R represents the distance before and after HPV enters and leaves the traffic network; ζ a,k,s It is an auxiliary variable introduced by the large-M linearization method.

[0148] The present invention provides a two-layer low-carbon collaborative planning method for hydrogen refueling stations and distribution networks that considers carbon emission flows. The technical effects are as follows:

[0149] 1) The introduction of this carbon emission flow model tracks the carbon potential of nodes in the power grid and quantifies the carbon emission responsibility of the electricity sources involved in hydrogen production at hydrogen refueling stations. This also enables hydrogen refueling stations to increase their awareness of carbon emissions, thereby optimizing their hydrogen production practices and reducing carbon emissions. Compared to traditional carbon emission calculation methods, the carbon emission flow model more comprehensively considers the characteristics of the system grid and power source location to rationally and effectively divide carbon emission responsibilities among hydrogen refueling stations.

[0150] 2) Compared with traditional models, the low-carbon planning model of hydrogen refueling stations and distribution networks proposed in this invention, which is based on carbon emission flows and takes carbon emission constraints into consideration, can achieve significant carbon emission reductions and safety improvements with only a slight increase in cost investment.

[0151] 3) The traffic flow distribution model of the present invention closely combines the traffic flow distribution in the traffic network with the electricity-hydrogen-traffic coupling system, taking into account both traffic congestion and user equilibrium under the driver's autonomous path selection. It also uses the carbon emission flow model to directly affect the node carbon emission intensity on the traffic flow and the location of hydrogen refueling stations, thus achieving the coordinated optimization of traffic demand, energy supply and carbon emissions. This model can not only dynamically reflect the impact of traffic flow on hydrogen refueling demand and grid load, but also optimize hydrogen refueling behavior and wind power layout through low-carbon oriented constraints, thereby improving the low-carbon and economic efficiency of the system in multiple scenarios and time periods. At the same time, it incorporates hydrogen storage safety into unified planning, effectively reducing dangerous operating time and carbon emissions, and achieving a deep integration of energy utilization, traffic operation and environmental friendliness, which is superior to the traditional single traffic distribution method.

[0152] 4) The biggest advantage of the hydrogen station-distribution network two-layer planning model that takes into account carbon emission constraints in the present invention is that it introduces carbon emission flow theory into the coordinated planning of hydrogen stations and distribution networks, realizing accurate allocation of carbon emission responsibilities and low-carbon guidance from power sources to transportation terminals. The model is centered on a two-layer optimization framework. The upper layer reduces the carbon emission intensity of nodes through new energy and distribution network line expansion, and the lower layer optimizes the site selection, hydrogen production behavior and operational safety of hydrogen stations based on the carbon potential of nodes and traffic flow demand, thereby significantly reducing system carbon emissions while meeting the hydrogen storage safety of hydrogen stations. Unlike traditional models that only focus on cost or traffic demand, this model can collaboratively consider traffic congestion, power grid trends, wind power consumption and carbon emission costs in multiple scenarios and time periods, effectively avoiding high-carbon and high-risk nodes, while balancing investment and operating expenses, and achieving significant carbon emission reductions and safety improvements at a relatively small additional cost. In addition, the model also reveals the sensitivity of carbon prices, cruising range, etc. to system investment and emissions, has strong practical guiding significance, and can provide a systematic solution for the integration of future low-carbon transportation and energy networks.

[0153] 5) By constructing a carbon emission flow model and a two-layer dynamic optimization framework, the present invention solves the problems of unclear carbon emission responsibility, insufficient clean energy coordination, and safety risk management in hydrogen station planning. It provides a systematic technical path for improving renewable energy absorption capacity and system safety, as well as low-carbon coordinated planning of electricity-hydrogen-transportation systems. It has important practical value for deep decarbonization in the transportation field and the construction of new power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0154] The present invention will be further described below with reference to the accompanying drawings and examples:

[0155] Figure 1 Schematic diagram of the electricity-hydrogen-transportation coupling system model.

[0156] Figure 2 Schematic diagram of the two-tier planning of hydrogen refueling stations and distribution networks taking into account carbon emission constraints.

[0157] Figure 3 Wind power output situation.

[0158] Figure 4 This is the power load diagram for each scenario.

[0159] Figure 5 Meet the hydrogen refueling needs for various scenarios.

[0160] Figure 6 This is the coupling topology diagram of the distribution network and transportation network.

[0161] Figure 7 Plan the results for the case study.

[0162] Figure 8 This is a carbon emission intensity map of transportation network roads.

[0163] Figure 9 This is a diagram showing the changes in hydrogen production behavior at a hydrogen refueling station. DETAILED DESCRIPTION

[0164] A two-layer low-carbon collaborative planning method for hydrogen refueling stations and distribution networks considering carbon emission flows includes the following steps:

[0165] Step S1: Based on the principle of proportional sharing, a carbon emission flow model of the distribution network and hydrogen refueling station is constructed. By introducing wind farm construction decisions as a renewable energy supplement, the carbon emission intensity corresponding to the input electricity of the hydrogen refueling station is calculated to quantify the carbon emission responsibility of the hydrogen refueling station and provide dynamic data support for subsequent low-carbon site selection.

[0166] Step S2: Based on the network equilibrium theory, simulate the driver's travel in the traffic network, establish a traffic flow distribution model under network equilibrium, obtain the path flow distribution in the traffic network, combine the HPV's range and hydrogen refueling behavior constraints, deduce the hydrogen demand under different traffic scenarios, and generate the demand-side input data of the electric-hydrogen-transportation system.

[0167] Step S3: Improve the hydrogen station positioning model of the flow capture site selection model. Based on the HPV range and path continuity rules, screen candidate sites that meet the hydrogen refueling needs. At the same time, introduce hydrogen storage safety constraints to ensure that the site selection plan meets the hydrogen storage safety threshold and form a preliminary hydrogen station layout plan.

[0168] Step S4: Establish a two-tiered planning model for hydrogen refueling stations and distribution networks that takes carbon emission constraints into account. The upper-tier model is for distribution network planning: With the goal of minimizing total cost, it optimizes distribution network construction and the layout of new energy sources to reduce the overall carbon intensity of the grid. Based on power flow calculations and carbon emission flow models, it dynamically updates the carbon emission intensity of each node in the grid, providing a low-carbon power environment for the lower-tier model. The lower-tier model is for hydrogen refueling station planning: Based on the node carbon intensity data output by the upper-tier model, with the goal of jointly minimizing hydrogen refueling station operation and maintenance costs and carbon emission costs, it optimizes hydrogen refueling station site selection and hydrogen production behavior to ensure that they match the real-time carbon intensity of electricity and meet carbon emission limits and safe operation requirements.

[0169] Step S5: Linearize the nonlinear objectives and constraints in steps S1 to S4, and use the solver to iteratively solve the two-layer model until the power load between two iterations of the hydrogen refueling station meets the convergence requirements, and output the final wind farm, distribution network construction and hydrogen refueling station coordinated planning plan.

[0170] In step S1, the process of constructing the carbon emission flow model is as follows:

[0171] Based on the principle of proportional sharing, the carbon emission intensity of the grid node is shared proportionally with the carbon flow contribution of its input power, realizing a quantitative assessment of the actual responsibility of power consumption units such as hydrogen refueling stations in the grid carbon emissions. The specific construction method is as follows:

[0172]

[0173] Where, They represent the incoming and outgoing terminal sets of node i respectively; P′ mj is the contribution of the mth incoming line to the jth outgoing line; P m,i is the active power flowing through line mi; P i G is the output power of the i-th generator; P j is the active power on the jth outgoing line; P′ Gjis the contribution of the generator to the jth outgoing line; g is the incoming line node; P g is the power flow at the incoming line; R j is the carbon flow rate of the jth outlet; ρ i , ρ i G 、 are the carbon emission intensities of node i, generator at node i, and line mi respectively; R m,i 、 are the carbon flow rates of the generator on line mi and node i respectively.

[0174] After considering the carbon emission flow intensity of hydrogen refueling stations, and taking into account that new energy construction can provide high-quality low-carbon electricity to reduce overall carbon emissions, the construction decision of new energy plants is introduced when selecting the site of hydrogen refueling stations. To facilitate model calculations, wind farms are used to replace the entire new energy field. The specific construction method is as follows:

[0175]

[0176] Where, P i W is the output power of wind power; is the carbon emission intensity of wind power; is the carbon flow rate of the branch containing the wind farm.

[0177] In step S2, a traffic flow distribution model is constructed based on the Bureau of Public Roads (BPR) function. Two setting conditions are proposed to reduce the difference between the BPR function and actual traffic congestion. The traffic flow distribution model in the UE mode is established based on the two setting conditions. The specific construction method is as follows:

[0178] Setting condition ①: Road traffic flow should always be kept within its maximum carrying capacity to ensure smooth and safe traffic.

[0179] Setting condition ②: When the traffic flow on the road reaches its maximum capacity, the vehicle's speed will slow down, resulting in a delay in travel time:

[0180]

[0181] Where x a,t,s is the traffic volume on road a; t a,t,s is the actual travel time of road a; is the free travel time of road a; c a is the traffic capacity of road a; ε a,t,s is the delay time.

[0182] Based on the above assumptions, the traffic assignment model under social optimality is established as follows:

[0183]

[0184] x a,t,s ≤c a [ε a,t,s ];

[0185]

[0186] Where, is the traffic flow on path k; is the road and path association matrix; is the OD demand of HPV at time t; Q OD is the total OD demand in a day; ρ HPV is the HPV penetration rate; τ t,s is the hydrogenation rate of HPV at each time period under scenario s; γ a,t,s , ε a,t,s is the dual variable of the relevant constraint.

[0187] In step S3, based on the flow-refueling location model (FRLM), but omitting the step of continuously expanding the path analysis in the network, two driving rules of HPV are proposed to establish a hydrogen station positioning model. The specific construction method is as follows:

[0188] Based on the dynamic driving behavior constraints of HPV in the traffic network, the following two driving rules are proposed:

[0189] Rule ①: On the same route, the distance between two adjacent hydrogen refueling stations must be less than the HPV's range.

[0190] Rule ②: HPV should store a certain amount of hydrogen at the starting point and end point of any journey

[0191] Rule 1 is to prevent the HPV from stopping due to fuel exhaustion during long-distance driving, ensuring the continuity of cross-region driving. Rule 2 is to ensure that the HPV has sufficient initial hydrogen at the time of departure and retains a safe fuel reserve before reaching the destination to travel to another hydrogen refueling station or the final destination.

[0192] After each hydrogen refueling, the battery state of the HPV is W=1, and the candidate location of the hydrogen refueling station is the midpoint of each road. Assuming that the range of the HPV after being fully filled with hydrogen is R (km), to reach the state W O Enter the traffic network and leave the state W without consuming all the fuel DLeaving the transportation network. The planning of hydrogen refueling stations in this invention is mainly to meet the HPV hydrogenation demand during peak hours, so the hydrogen refueling station positioning model can be expressed as the following operation constraints:

[0193] 1) HPV can reach the first hydrogen refueling station node:

[0194]

[0195] 2) The battery status of HPV when it reaches the destination is higher than w D , to satisfy Rule 2:

[0196]

[0197] 3) HPV can reach all candidate hydrogen refueling station nodes:

[0198] W a,k,s ≥0,a∈Ω k ,k∈Ω OD ;

[0199] 4) Between two candidate hydrogen refueling stations, the HPV will not stop moving due to energy exhaustion:

[0200]

[0201] 5) HPV battery status at the candidate hydrogen station node:

[0202]

[0203] Where: a, b, k are road index and path index respectively; s represents the scene; y a The construction status of the hydrogen refueling station is 1 or 0 to indicate whether the hydrogen refueling station is being built or not; is the starting road set on path k; is the set of roads close to the end point on path k; Ω OD is a set of paths; are the previous and subsequent road sets on path k respectively; Ω k is the set of roads on path k; r a 、r b is the distance between roads.

[0204] After the site selection of the hydrogen refueling station is completed by comprehensively considering the traffic flow, the safe operation of the hydrogen refueling station must also be taken into consideration. To this end, the TNT equivalent method is used as a safety operation indicator to measure the energy generated by the explosion of the hydrogen storage tank, so as to evaluate the operating risk of the hydrogen refueling station. The explosion energy of the hydrogen storage tank is:

[0205]

[0206] Set the hydrogen storage tank explosion energy warning value E according to the explosion energy war and the limit value E max When the hydrogen storage tank is operating between the two, it is in a dangerous operating state. Therefore, the safety constraint of the hydrogen refueling station is set as:

[0207] 0E a,t,s ≤E war +I a,t,s M;

[0208] E war +(I a,t,s -1)M≤E a,t,s ≤E max ;

[0209] Where, E a,t,s is the explosion energy of the hydrogen storage tank; is the mass of hydrogen in the hydrogen storage tank; α is the explosion coefficient; E A is the TNT equivalent coefficient of the explosion; E f is the calorific value of hydrogen; E TNT It is the explosion heat generated when TNT explodes; a,t,s It is a binary variable, and its value is 1 or 0, indicating whether the hydrogen station is operating in the warning range. war ,E max ]; M takes the maximum value.

[0210] In step S4, a two-layer planning model of hydrogen refueling station and distribution network taking into account carbon emission constraints is established based on steps S1 to S3. The upper model plans and constructs the distribution network and new energy, and then transmits the unit output and flow operation results to the lower model. The lower model first calculates the carbon emission flow based on the upper output results, and then plans the site selection of the hydrogen refueling station based on the node carbon emission intensity. The lower model transmits the power load of the hydrogen refueling station to the upper model to optimize the site selection of the new energy station. This cycle continuously optimizes the new energy output of the upper model and the load optimization of the lower model to achieve a reduction in the overall carbon emissions of the system. The specific construction method is as follows:

[0211] 1. Objective function of the upper model:

[0212] The upper model is mainly used to realize the site selection and planning of power grid expansion and new energy, so the objective function is set to minimize the total cost, including the investment cost F S,In and operating costs F S,Op :

[0213] min f S =F S,In +F S,Op +F L,E ;

[0214] The investment cost includes the cost of line expansion and the construction cost of new energy stations:

[0215]

[0216] Where, F In is the annual investment cost of the upper model; γ is the annual equivalent coefficient of the equipment; θ is the discount rate; Y is the planning period; w1 is the cost of distribution line expansion; is the number of line expansions; x i It is a binary variable, representing the construction status of new energy. When the value is 1 or 0, it means that new energy stations are being built or not. W is the unit capacity investment cost of new energy equipment; P i W,rated represents the wind power construction capacity of node i; Ω DN is the set of distribution network nodes; Ω W is the candidate location set of the wind farm station; F L,E is the cost excluding electricity purchase in the lower model.

[0217] Operating costs mainly include electricity purchase costs and equipment maintenance costs:

[0218]

[0219] Where, F Op is the operating cost of the upper model; β M is the equipment maintenance factor; D s is the probability corresponding to the scene; The power of purchased electricity; is the time-of-use electricity price; Ω s is a typical scene set; Ω T For time collection.

[0220] 2. Constraints of the upper model:

[0221] The safe operation of the distribution network is the primary requirement for the upper-level model. Specifically, load changes in the lower-level model will affect the magnitude and direction of the power flow, and thus the carbon emission flow. After the hydrogen refueling station is connected to the distribution network as a load, it will be affected by the HPV traffic flow. The fluctuation of its electricity demand will affect the system power flow. Using the Distflow model to calculate the power flow and voltage distribution of the distribution network, the following constraints are imposed:

[0222] 1) Power balance constraints:

[0223]

[0224] Where n is the distribution network node; D DNIndicates the maximum number of distribution network lines that can be expanded; s(i) is the set of head-end nodes corresponding to the line with end node i; e(i) is the set of end nodes corresponding to the line with head node i; is the wind power active power; P m,i,t,s , Q m,i,t,s are the active and reactive power flowing through line mi respectively; P i,n,t,s , Q i,n,t,s are the active and reactive power flowing through line ni respectively; Divided into basic active load and reactive load; is the load of the hydrogen refueling station.

[0225] 2) Voltage drop constraints after expansion:

[0226] When the existing line cannot meet the planned power flow, it is necessary to consider expanding the line capacity. The new distribution line is built in parallel with the original line and uses the same parameters as the original line. The voltage drop constraint of the expanded line is:

[0227]

[0228] Linearize the voltage drop constraint:

[0229]

[0230] Where D DN Indicates the maximum number of distribution network lines that can be expanded; Replace the introduced binary variable Introducing new variables Linearization is performed; U1 is the voltage at node 1; U i,t,s is the voltage at node i at time t; is the upper limit of node i voltage; is the upper limit of the voltage at node m.

[0231] 3) Node voltage constraints:

[0232]

[0233] Where, are the upper and lower limits of the voltage at the distribution network node i, respectively.

[0234] 4) Distribution network carbon emission cap:

[0235]

[0236] Where, Provide a reference for carbon emission intensity for the upper-level power grid; is the total carbon emissions of the distribution network.

[0237] 5) Constraints on new energy equipment:

[0238]

[0239] Where, is the predicted per-unit output value of the wind turbine; N is the maximum capacity that can be built for a single wind farm; k is the number of wind farms to be built; M W,rated It is the upper limit of the total capacity construction of the wind farm.

[0240] 3. Objective function of the lower model:

[0241] The lower-level model is for hydrogen refueling station planning, addressing the site selection and carbon emission issues of hydrogen refueling stations. The lower-level model first needs to track the carbon footprint and calculate the node carbon intensity based on the results of the upper-level model.

[0242] When the active power P m,i When is positive, the carbon emission intensity of line mi is equal to the carbon emission intensity of node m:

[0243]

[0244] The objective function of the lower model consists of three parts:

[0245] 1)F L,In : Investment costs for planning and construction of hydrogen refueling stations and their equipment, as well as their safety investment costs; 2) F L,Op : The operating costs of hydrogen refueling stations include penalty costs for dangerous operations, carbon emission costs, maintenance costs of hydrogen refueling station equipment, and the cost of purchasing electricity from the main grid; 3) Costs other than electricity purchases in the upper model:

[0246] min f L =F L,In +F L,Op +F S,E ;

[0247]

[0248] Where, γ L is the equal annual value coefficient of the lower model; w2 is the risk penalty coefficient; w3 is the unit carbon emission cost; The planned capacity of electrolyzers, hydrogen storage tanks and safety equipment; c HRS 、c e 、c tank 、c safe are the construction costs of hydrogen refueling stations, electrolyzers, hydrogen storage tanks, and safety equipment cost coefficients; E quo F is the upper limit of carbon emissions for hydrogen refueling stations; S,E This is the cost excluding electricity purchase in the upper model.

[0249] 4. Constraints of the lower model:

[0250] 1) Carbon emission constraints of hydrogen refueling stations:

[0251]

[0252] Where, is the power consumption of the electrolytic cell.

[0253] 2) Electrolyzer operation constraints:

[0254]

[0255] Where λ e The working efficiency of the electrolyzer; the quality of hydrogen produced for the electrolyser; The upper limit of the electrolyzer's hydrogen production capacity. 3) Hydrogen storage tank constraints:

[0256]

[0257] Where, is the mass of hydrogen in the hydrogen storage tank at time t+1; is the mass of hydrogen in the hydrogen storage tank at time t; The hydrogen demand of HPV; The upper limit of the hydrogen storage tank capacity.

[0258] 4) Hydrogenation machine operation constraints:

[0259]

[0260] Where H ave is the average hydrogenation mass of the hydrogenator; δ HPV The mass of hydrogen refilled for each HPV refill; The maximum number of hydrogen refueling machines that can be built in a hydrogen refueling station; x a,t,s For traffic that requires hydrogen refueling.

[0261] The power load of the lower model needs to be transferred to the upper model for update iteration, which can be calculated by the following formula:

[0262]

[0263] The low-carbon planning model for hydrogen refueling stations and distribution networks established in this paper, which takes into account carbon emission constraints, meets stringent carbon emission requirements by optimizing the distribution system grid, constructing new energy units, and planning the site selection of hydrogen refueling stations. Therefore, the planning model solution process is as follows:

[0264] Step 1: Input the basic system parameters and the initial load of the hydrogen station into the upper model;

[0265] Step 2: The upper model constructs new energy and distribution network lines based on hydrogen station load changes and carbon emission requirements. Step 3: The power flow data and unit output data obtained by the upper model are passed to the carbon emission flow model of the lower model to calculate the carbon emission intensity of each node.

[0266] Step 4: The lower model further guides the site selection and construction of hydrogen refueling stations and hydrogen production behavior based on the carbon emission intensity of the nodes, while meeting the requirements of safety and low carbon.

[0267] Step 5: Determine whether the power load of the hydrogen refueling station between two iterations meets the requirements of the following formula. If so, output the planning scheme. If not, transmit the power load of the hydrogen refueling station to the upper model and iterate and solve from step 2.

[0268]

[0269] Where, is the tolerance, and z is the number of iterations.

[0270] In step S5, some constraints have been linearized in the above steps. In this step, the constraints of the traffic flow distribution model and the hydrogen station positioning model are linearized. The specific construction method is as follows:

[0271] Since the constraints of the model established in step S2 cannot be met when the traffic flow is too large, the Lagrange relaxation principle and strong duality theory are used to eliminate the duality gap and finally convert it into a linearized traffic assignment model under user equilibrium:

[0272]

[0273] Where, is the traffic flow on path k; is the road and path association matrix; is the OD demand of HPV at time t; Q OD is the total OD demand in a day; ρ HPV is the HPV penetration rate; τ t,s is the hydrogenation rate of HPV at each time period under scenario s; γ a,t,s , ε a,t,s is the dual variable of the relevant constraint.

[0274] The nonlinear operation constraints in the hydrogen station positioning model can be linearized using the large M method, and the auxiliary variable ζ is introduced. a,k,s , so that the MINLP problem is converted to a MILP problem:

[0275]

[0276] Where a, b, and k are road index and path index respectively; s represents the scene; y a The construction status of the hydrogen refueling station is 1 or 0 to indicate whether the hydrogen refueling station is being built or not; is the starting road set on path k; is the set of roads close to the end point on path k; Ω OD is a set of paths; are the previous and subsequent road sets on path k respectively; Ω k is the set of roads on path k; R represents the distance of HPV before and after entering and leaving the transportation network.

[0277] Example:

[0278] This paper uses a modified IEEE 33-node distribution network and a 12-node transportation network as the basic calculation background. The planning period is set to 20 years, where a is the number of roads, the discount rate is 12%, and the number of expansions per road is 10. The transportation network road parameters and the traffic demand settings for the OD pairs are shown in Table 1. The time-of-day electricity prices are shown in Table 2. The transportation network parameters are shown in Table 3.

[0279] Table 1 OD traffic demand

[0280]

[0281] Table 2 Time-of-day electricity prices

[0282]

[0283] Table 3 Transportation network parameters

[0284]

[0285] Set S1, S2, and S3 to represent summer, winter, and transition seasons, with corresponding days of 100, 93, and 172. Hydrogen refueling demand is divided into weekdays and weekends, as follows: Figure 5 As shown in the figure, since load scenarios and hydrogenation loads can be combined, 6 typical scenarios are formed. Assuming that a total of five wind farms are built, the wind power output is generated using the Monte Carlo method to generate 1000 scenarios and then K-means clustering is used to divide them into 3 scenarios as shown in the figure. Figure 1 As shown in the figure, the power load in the corresponding season is as follows Figure 4 The candidate wind power sites are Figure 6 The blue dotted box in the

[0286] Based on the set parameters, the results of hydrogen station, wind farm site selection and road expansion and upgrade are as follows: Figure 7 shown.

[0287] Table 4 Line expansion results

[0288]

[0289] The specific line expansion results are shown in Table 4. Figure 7 The amount of expansion is indicated by the thickness of its line. Figure 7 It can be found that wind farms are mainly deployed in the terminal area of ​​the distribution network. On the one hand, this is because this layout can increase the voltage level at the end of the line and prevent voltage instability; on the other hand, clean energy can be transmitted to the upstream node through reverse flow, thereby effectively reducing the carbon emission intensity of the node where the hydrogen station is located.

[0290] Depend on Figure 8 It can be seen that the carbon emission intensity of road areas where hydrogen refueling stations and wind farms are deployed has shown a significant downward trend compared to other conventional roads. From a spatial distribution perspective, road network units with hydrogen refueling stations as core nodes and wind farms as clean energy supplies can generate electricity through renewable energy while also supplying hydrogen to hydrogen stations through water electrolysis and providing energy for HPVs. The clean electricity injected into the grid by wind farms reduces the proportion of traditional fossil energy power supply and the substitution effect of zero-emission vehicles supported by hydrogen refueling stations on fuel vehicles. This differentiated emission characteristic not only verifies the emission reduction efficiency of the coordinated layout of clean energy infrastructure, but also provides an empirical basis for regional low-carbon transportation network planning.

[0291] In order to verify the effectiveness of the hydrogen station safety constraints and low-carbon collaborative planning in the model of the present invention, the following scenarios are set for comparative analysis.

[0292] Scenario 1 is the baseline scenario: no carbon emission constraints are considered, nor are hydrogen refueling station safety constraints;

[0293] Scenario 2 is a low-carbon planning scenario: it simultaneously considers carbon emission constraints, hydrogen station safety constraints, and 30% distribution network carbon emission reduction constraints.

[0294] Table 5 Planning results for different scenarios

[0295]

[0296] As shown in Table 5, Scenario 1 considers neither carbon emission constraints nor hydrogen refueling station safety constraints, so there are no wind power construction capacity or safety investment costs. In this case, directly purchasing electricity from the upstream grid would be more economical, but it would also result in 332.9 tons of carbon emissions per day and 71 hours of dangerous operating time.

[0297] Comparing the planning results of a baseline scenario with a low-carbon planning scenario demonstrates the comprehensive advantages of the proposed two-tier low-carbon collaborative planning model in terms of emission reduction efficiency, safety, and economic efficiency. The baseline scenario, which targets the lowest total cost and does not consider carbon emission and safety constraints, relies on traditional grid power supply, resulting in system-wide average daily carbon emissions of 332.9 tons, 71 hours of hazardous operating time for hydrogen refueling stations, and no wind farms. In contrast, the low-carbon planning scenario, after simultaneously imposing carbon emission and safety constraints and a 30% carbon reduction target for the distribution network, significantly reduces the grid's carbon intensity by optimizing wind farm siting and expanding the distribution network, reducing system-wide average daily carbon emissions to 233 tons, a 30% decrease. Furthermore, the hazardous operating time for hydrogen refueling stations is reduced to 9 hours, a decrease of 87.3%, demonstrating that safety constraints effectively avoid high-risk site selection. Although the low-carbon planning scenario's total cost increases slightly by 1.56% compared to the baseline scenario, it achieves a balance between low-carbon objectives and economic efficiency by replacing high-carbon electricity with wind power and optimizing the layout of hydrogen refueling stations.

[0298] In scenario 2, after adding corresponding constraints on the basis of scenario 1, the hydrogen refueling station can optimize its own hydrogen production behavior according to the carbon emission intensity, thereby avoiding hydrogen production during the period of high carbon emission intensity and achieving further carbon emission reduction. Figure 9 As shown in the figure, in scenario 1, the hydrogen production behavior of hydrogen refueling stations and the carbon emission intensity of the power grid show a significant positive coupling relationship: when the hydrogen production volume surges during the morning peak (8:00-12:00) and evening peak (16:00-20:00), the carbon emission intensity simultaneously climbs to above 0.85kgCO2 / kWh, indicating that hydrogen production during this period is highly dependent on high-carbon electricity. After introducing the carbon emission intensity threshold constraint on the basis of scenario 1, scenario 2 significantly weakens this coupling effect by dynamically adjusting the hydrogen production load. Specifically, scenario 2 actively suppresses during the morning and evening peaks, and transfers part of the hydrogen production demand to the nighttime (0:00-4:00) and noon (12:00-16:00) periods with carbon emission intensity lower than 0.4kgCO2 / kWh, thereby achieving a precise match between the hydrogen production period and the clean energy power generation window. Furthermore, Scenario 2 further reduces hydrogen production to below 100 units between 8:00 PM and 12:00 AM, avoiding a secondary increase in carbon emissions intensity during this period due to grid peak regulation. This strategy results in a "low and flat" carbon emissions intensity curve for Scenario 2, with its full-time average decreasing by 32% compared to Scenario 1 and its peak intensity reduced by 58%. This demonstrates the potential for carbon reduction over time using a hydrogen production scheduling strategy based on dynamic carbon emissions intensity constraints.

[0299] Further analysis shows that the carbon emission intensity in the area surrounding the hydrogen refueling station has also been alleviated due to the access of wind farms, verifying the emission reduction potential of the coordinated supply of clean energy in the "electricity-hydrogen-transportation" coupling network.

[0300] Through the above analysis, on the basis of considering the basic safety constraints of hydrogen refueling stations, further considering carbon emission reduction constraints will increase the total cost, but it can achieve the established carbon emission reduction goals while further improving the safety of the system.

Claims

1. A two-layer low-carbon collaborative planning method for hydrogen refueling stations and distribution networks considering carbon emission flows, characterized by The following steps are involved: Step 1: Based on the proportional sharing principle, a carbon emission flow model of the distribution network and hydrogen refueling station is constructed; Step 2: Introduce wind farm construction decisions as a renewable energy supplement, calculate the carbon emission intensity corresponding to the hydrogen station's input electricity, and quantify the carbon emission responsibility of the hydrogen station; Step 3: Establish a traffic flow distribution model to obtain the path flow distribution in the traffic network; Step 4: Establish a hydrogen refueling station positioning model to screen candidate sites that meet hydrogen refueling needs and assess the operational risks of hydrogen refueling stations; Step 5: Establish a two-layer planning model for hydrogen refueling stations and distribution networks that takes carbon emission constraints into account. The upper-layer model transmits the unit output and power flow operation results to the lower-layer model. Step 6: Linearize the nonlinear objectives and constraints in steps 1 to 5, and use the solver to iteratively solve the hydrogen station-distribution network two-level planning model until the power load between two iterations of the hydrogen station meets the convergence requirements, and output the final coordinated planning plan for wind farm, distribution network construction and hydrogen station.

2. The method for double-layer low-carbon collaborative planning of hydrogen refueling stations and distribution networks taking carbon emission flows into consideration according to claim 1, characterized in that: In step 1, the carbon emission intensity of the grid node and the carbon flow contribution of its input power are shared proportionally to construct a carbon emission flow model for the distribution network and the hydrogen refueling station. The specific construction method of the carbon emission flow model for the distribution network and the hydrogen refueling station is as follows: In the above formula, They represent the incoming and outgoing terminal sets of node i respectively; P m ' j is the contribution of the mth incoming line to the jth outgoing line; P m,i is the active power flowing through line mi; P i G is the output power of the i-th generator; P j is the active power on the jth outgoing line; P G ' j is the contribution of the generator to the jth outgoing line; g is the incoming line node; P g is the power flow at the incoming line; R j is the carbon flow rate of the jth outlet; ρ i , ρ i G 、 are the carbon emission intensities of node i, generator at node i, and line mi respectively; R m,i 、 are the carbon flow rates of the generator on line mi and node i respectively; is the comprehensive carbon emission intensity of node j.

3. The method for a two-tier low-carbon collaborative planning of a hydrogen refueling station and a distribution network taking carbon emission flows into consideration according to claim 2, characterized in that: In step 2, while selecting the site for the hydrogen refueling station, the decision on the construction of a new energy plant is introduced. To facilitate model calculation, a wind farm is used to replace the entire new energy plant, as follows: In formula (5), is the carbon emission intensity of node i; P i W is the output power of wind power; is the carbon emission intensity of wind power; is the carbon flow rate of the branch containing the wind farm; R i is the carbon flow rate of node i; P i is the active power of node i.

4. The method for a two-tier low-carbon collaborative planning of a hydrogen refueling station and a distribution network taking carbon emission flows into consideration according to claim 3, characterized in that: In step 3, a traffic flow distribution model under traffic network equilibrium is constructed: The traffic flow distribution model is established based on two setting conditions. The specific construction method is as follows: Setting condition ①: Road traffic flow should always be kept within its maximum carrying capacity to ensure smooth and safe traffic; Setting condition ②: When the traffic flow on the road reaches its maximum capacity, the vehicle's speed will slow down, resulting in a delay in travel time; In formula (6), x a,t,s is the traffic volume on road a; t a,t,s is the actual travel time of road a; is the free travel time of road a; c a is the traffic capacity of road a; ε a,t,s is the delay time; Based on the above setting conditions, the traffic flow distribution model under social optimality is established as follows: x a,t,s ≤c a [e a,t,s ] (11); In the above formula, is the traffic flow on path k; is the road and path association matrix; is the OD demand of HPV at time t; Q OD is the total OD demand in a day; ρ HPV is the HPV penetration rate; τ t,s is the hydrogenation rate of HPV at each time period under scenario s; γ a,t,s , ε a,t,s is the dual variable of the relevant constraints; a is the road index; s is the scene index; k is the path index; t is the time index; OD represents the origin and destination; [·] is the introduced auxiliary variable.

5. The method for double-layer low-carbon collaborative planning of hydrogen refueling stations and distribution networks taking carbon emission flows into consideration according to claim 4, characterized in that: In step 4, based on the traffic capture site selection model, the step of continuously expanding the path analysis in the network is omitted, and two driving rules of HPV are proposed to establish the hydrogen station positioning model. The specific construction method is as follows: Based on the dynamic driving behavior constraints of HPV in the traffic network, the following two driving rules are proposed: Rule ①: On the same route, the distance between two adjacent hydrogen refueling stations must be less than the HPV's range; Rule ②: HPV should store a certain amount of hydrogen at the starting point and end point of any journey; Rule 1 is to prevent the HPV from stopping due to fuel exhaustion during long-distance driving, ensuring the continuity of cross-region driving; Rule 2 is to ensure that the HPV has sufficient initial hydrogen at the time of departure and retains a safe fuel reserve before reaching the destination to travel to another hydrogen refueling station or the final destination; After each hydrogen refueling, the battery state of the HPV is W=1, and the candidate location of the hydrogen refueling station is the midpoint of each road. Assuming that the cruising range of the HPV after it is fully filled with hydrogen is R, to reach the state W O Enter the traffic network and leave state W without consuming all the fuel D leaving the transport network; The planning of hydrogen refueling stations is to meet the HPV hydrogenation demand during peak hours. Therefore, the hydrogen refueling station positioning model is expressed as the following operation constraints: 1) HPV can reach the first hydrogen refueling station node: 2) The battery status of HPV when it reaches the destination is higher than w D , to satisfy rule ②: 3) HPV can reach all candidate hydrogen refueling station nodes: IN a,k,s ≥0,a∈Ω k ,k∈Ω OD (15); 4) Between two candidate hydrogen refueling stations, the HPV will not stop moving due to energy exhaustion: 5) HPV battery status at the candidate hydrogen station node: In the above formula: a, b, k are road index and path index respectively; s represents the scene; y a The construction status of the hydrogen refueling station is 1 or 0 to indicate whether the hydrogen refueling station is being built or not; is the starting road set on path k; is the set of roads close to the end point on path k; Ω OD is a set of paths; are the previous and subsequent road sets on path k respectively; Ω k is the set of roads on path k; r a 、r b is the distance between roads; W a,k,s is the HPV battery status of road a on path k under scenario s.

6. The method for a two-tier low-carbon collaborative planning of a hydrogen refueling station and a distribution network taking carbon emission flows into consideration according to claim 5, characterized in that: In step 4, after the site selection of the hydrogen refueling station is completed by comprehensively considering traffic flow, the safe operation of the hydrogen refueling station must also be taken into consideration. To this end, the TNT equivalent method is used as a safety operation indicator to measure the energy generated by the hydrogen storage tank explosion, thereby evaluating the operational risk of the hydrogen refueling station; The explosion energy of hydrogen storage tank is: In formula (19), E a,t,s is the explosion energy of the hydrogen storage tank; is the mass of hydrogen in the hydrogen storage tank; α is the explosion coefficient; E A is the TNT equivalent coefficient of the explosion; E f is the calorific value of hydrogen; E TNT It is the explosion heat generated when TNT explodes; Set the hydrogen storage tank explosion energy warning value E according to the explosion energy war and the limit value E max When the hydrogen storage tank is operating between the two, it is in a dangerous operating state. Therefore, the safety constraint of the hydrogen refueling station is set as: 0E a,t,s ≤E war +I a,t,s M(20); E war +(I a,t,s -1)M≤E a,t,s ≤E max (21); In the above formula: I a,t,s It is a binary variable, and its value is 1 or 0, indicating whether the hydrogen station is operating in the warning range. war ,E max ]; M takes the maximum value.

7. The method for double-layer low-carbon collaborative planning of hydrogen refueling stations and distribution networks taking carbon emission flows into consideration according to claim 6, characterized in that: In step 5, the lower model first calculates the carbon emission flow based on the output of the upper model, and then performs site planning for hydrogen refueling stations based on the carbon emission intensity of the nodes; The lower model transmits the power load of the hydrogen refueling station to the upper model to optimize the site selection of the hydrogen refueling station and the wind farm.

8. The method for double-layer low-carbon collaborative planning of hydrogen refueling stations and distribution networks taking carbon emission flows into consideration according to claim 7, characterized in that: In step 5, a two-level planning model of hydrogen refueling station and distribution network taking into account carbon emission constraints is established: The upper-level model is for distribution network planning: with the goal of minimizing total cost, it optimizes distribution network construction and renewable energy layout. Based on power flow calculations and carbon emission flow models, it dynamically updates the carbon emission intensity of each node in the grid, providing a low-carbon power environment for the lower-level model. The lower-level model is for hydrogen refueling station planning: based on the node carbon intensity data output by the upper-level model, with the goal of jointly minimizing the operation and maintenance costs and carbon emission costs of hydrogen refueling stations, the site selection and hydrogen production behavior of hydrogen refueling stations are optimized to ensure that they match the real-time electricity carbon intensity and meet carbon emission limits and safe operation requirements.

9. The method for double-layer low-carbon collaborative planning of hydrogen refueling stations and distribution networks taking carbon emission flows into consideration according to claim 8, characterized in that:

1. Objective function of the upper model: The upper model is used to implement the grid expansion and new energy site selection planning problem, so the objective function is set to minimize the total cost, including the investment cost F S,In and operating costs F S,Op : min f S =F S,In +F S,Op +F L,E (22); In formula (22), f S is the total cost of the upper model; F L,E represents the cost other than electricity purchase in the lower model; Investment cost F S,In It also includes the cost of line expansion and the construction cost of new energy stations: In the above formula, γ is the equipment annual equivalent coefficient; θ is the discount rate; Y is the planning period; w1 is the cost of distribution line expansion; is the number of line expansions; x i It is a binary variable, representing the construction status of new energy. When the value is 1 or 0, it means that new energy stations are being built or not. W is the unit capacity investment cost of new energy equipment; P i W,rated represents the wind power construction capacity of node i; Ω DN is the set of distribution network nodes; Ω W is the candidate location set of the wind farm station; m is the distribution network line index; i is the distribution network node index; Operating costs mainly include electricity purchase costs and equipment maintenance costs: In formula (25), F Op is the operating cost of the upper model; β M is the equipment maintenance factor; D s is the probability corresponding to the scene; The power of purchased electricity; is the time-of-use electricity price; Ω s is a typical scene set; Ω T is the time set; △t is the time interval; 2. Constraints of the upper model: The safe operation of the distribution network is the primary requirement of the upper-level model. Specifically, load changes in the lower-level model will affect the magnitude and direction of the power flow, and thus the carbon emission flow. After the hydrogen refueling station is connected to the distribution network as a load, it will be affected by the HPV traffic flow. The fluctuation of its electricity demand will affect the system power flow. Using the Distflow model to calculate the power flow and voltage distribution of the distribution network, the following constraints are imposed: 1) Power balance constraints: In the above formula, n is the distribution network node; s(i) is the set of head-end nodes corresponding to the line with end node i; e(i) is the set of end nodes corresponding to the line with end node i; is the wind power active power; P m,i,t,s , Q m,i,t,s are the active and reactive power flowing through line mi respectively; P i,n,t,s , Q i,n,t,s are the active and reactive power flowing through line ni respectively; Divided into basic active load and reactive load; is the load of the hydrogen refueling station; 2) Voltage drop constraints after expansion: When the existing line cannot meet the planned power flow, the line capacity should be expanded. The new distribution line should be built in parallel with the original line and use the same parameters as the original line. The voltage drop constraint of the expanded line is: In formula (28): U i,t,s is the node voltage of node i of line mi at time t under scenario s; U m,t,s is the node voltage of node m of line mi at time t under scenario s; New variables introduced for linearization; R m,i is the resistance of circuit mi; ​​X m,i is the reactance of line mi; U1 is the voltage at node 1; D DN Indicates the maximum number of distribution network lines that can be expanded; u indicates the segment index of the number of power grid line expansions; Linearize the voltage drop constraint: In the above formula: D DN Indicates the maximum number of distribution network lines that can be expanded; Replace the introduced binary variable Introducing new variables Linearization is performed; U1 is the voltage at node 1; U i,t,s is the voltage at node i at time t; is the upper limit of node i voltage; is the upper limit of the voltage at node m; 3) Node voltage constraints: In formula (33): are the upper and lower limits of the voltage of the distribution network node i, respectively; 4) Distribution network carbon emission cap: In formula (34): Provide a reference for carbon emission intensity for the upper-level power grid; is the total carbon emissions of the distribution network; D S is the probability of scenario s occurring; is the electricity purchase power at time t under scenario s; ρ W is the carbon emission intensity of the wind farm; is the wind power of node i at time t under scenario s; 5) Constraints on new energy equipment: Where, is the predicted per-unit output value of the wind turbine; N is the maximum capacity that can be built for a single wind farm; w is the number of wind farms to be built; M W,rated The total capacity construction limit for wind farms; P i W,rated is the wind power plant capacity at node i; X i is a 0-1 variable, indicating the construction status of the wind farm at node i; 3. Objective function of the lower model: The lower model is for hydrogen refueling station planning, which addresses the issues of hydrogen refueling station site selection and carbon emissions; The lower-level model first needs to track the carbon footprint and calculate the node carbon intensity based on the results obtained by the upper-level model; When the active power P m,i When is positive, the carbon emission intensity of line mi is equal to the carbon emission intensity of node m: In formula (39): is the carbon emission intensity of node m on line mi; is the carbon emission intensity of node i on line mi; the objective function of the lower model consists of three parts: 1)F L,In : The investment cost of planning and constructing a hydrogen refueling station and its equipment, as well as its safety investment cost; 2)F L,Op : The operating costs of hydrogen refueling stations include penalty costs for dangerous operations, carbon emission costs, maintenance costs of hydrogen refueling station equipment, and the cost of purchasing electricity from the main grid; 3) Costs other than electricity purchase in the upper model: min f L =F L,In +F L,Op +F S,E (40); Where y a It is a 0-1 variable, indicating the construction status of the hydrogen refueling station; γ L is the equal-year coefficient of the lower model; w2 is the risk penalty coefficient; w3 is the unit carbon emission cost; The planned capacity of electrolyzers, hydrogen storage tanks, and safety equipment; c HRS 、c e 、c tank 、c safe are the construction costs of hydrogen refueling stations, electrolyzers, hydrogen storage tanks, and safety equipment cost coefficients; E quo A cap on carbon emissions for hydrogen refueling stations; F S,E This is the cost other than electricity purchase in the upper model; Ω A is the road set; I a,t,s It is a 0-1 variable, indicating whether the hydrogen storage tank is operating within the warning range; E a,t,s is the explosion energy of the hydrogen storage tank; β M is the equipment maintenance factor; 4. Constraints of the lower model: 1) Carbon emission constraints of hydrogen refueling stations: In formula (43): is the power consumption of the electrolytic cell; is the carbon emission intensity of the road where the hydrogen refueling station is located; 2) Electrolyzer operation constraints: In formula (44): e The working efficiency of the electrolyzer; the quality of hydrogen produced for the electrolyser; is the electrolytic cell capacity; is the upper limit of the hydrogen production capacity of the electrolyzer; 3) Hydrogen storage tank constraints: In formula (45): is the mass of hydrogen in the hydrogen storage tank at time t+1; is the mass of hydrogen in the hydrogen storage tank at time t; The hydrogen demand of HPV; is the upper limit of the capacity of the hydrogen storage tank; 3) Hydrogenation machine operation constraints: In formula (46): Hydrogen demand of HPV; J a is the number of hydrogen refueling machines built; H ave is the average hydrogenation mass of the hydrogenator; δ HPV The mass of hydrogen refilled for each HPV refill; The maximum number of hydrogen refueling machines that can be built in a hydrogen refueling station; x a,t,s For traffic volume that requires hydrogen refueling; The power load of the lower model needs to be transferred to the upper model for update iteration, which can be calculated by the following formula: In formula (47): The power consumed by the hydrogen refueling station at node i in the distribution network; Ω A(i) is the set of roads associated with distribution network node i.

10. The method for double-layer low-carbon collaborative planning of hydrogen refueling stations and distribution networks taking carbon emission flows into consideration according to claim 9, characterized in that: In step 6, some constraints have been linearized in the above steps. In this step, the constraints of the traffic flow distribution model and the hydrogen station positioning model are linearized. The specific construction method is as follows: The traffic flow distribution model established in step 3 is converted into a linearized traffic distribution model under user equilibrium; In formula (49), is the traffic flow on path k; is the road and path association matrix; is the OD demand of HPV at time t; Q OD is the total OD demand in a day; ρ HPV is the HPV penetration rate; τ t,s is the hydrogenation rate of HPV at each time period under scenario s; γ a,t,s , ε a,t,s is the dual variable of the relevant constraint; The nonlinear operation constraints in the hydrogen station positioning model can be linearized using the large M method, and the auxiliary variable ζ is introduced. a,k,s , so that the MINLP problem is converted to a MILP problem: In formula (50), a, b, and k are the road index and path index respectively; s represents the scene; y a The construction status of the hydrogen refueling station is 1 or 0 to indicate whether the hydrogen refueling station is being built or not; is the starting road set on path k; is the set of roads close to the end point on path k; Ω OD is a set of paths; are the previous and subsequent road sets on path k respectively; Ω k is the set of roads on path k; R represents the distance before and after HPV enters and leaves the traffic network; ζ a,k,s It is an auxiliary variable introduced by the large-M linearization method.

Images