Hybrid electricity-hydrogen traffic station planning method and system based on multi-stage distribution robust optimization

Abstract

The invention relates to a hybrid electric hydrogen traffic station planning method and system based on multi-stage distribution robust optimization, and the method comprises the following steps: constructing a low-carbon transformation-oriented collaborative planning model based on a planning target, investment constraints and operation constraints; modeling is carried out according to the endogenous uncertainty of the influence of hybrid electric hydrogen traffic station site selection on the permeability of new energy vehicles, the endogenous uncertainty of the influence of gas station decommissioning decision on the permeability of traditional fuel vehicles and the exogenous uncertainty of wind power prediction errors, a decision-making dependency fuzzy set is obtained, and based on the decision-making dependency fuzzy set, the permeability of the new energy vehicles is determined. Converting the collaborative planning model into a distributed robust multi-stage planning problem; and solving the distributed robust multi-stage planning problem to obtain a hybrid electric hydrogen traffic station planning result of each stage. Compared with the prior art, the method has the advantages of realizing ordered and stable low-carbon transformation of the electric power traffic coupling system under the multi-source uncertainty and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a hybrid electric-hydrogen transportation station planning method and system based on multi-stage distributed robust optimization. Background Art

[0002] In the transportation sector, internal combustion vehicles (ICVs) are the main source of carbon emissions. As global concerns about climate warming and fossil fuel depletion increase, the market share of alternative fuel vehicles (AFVs), including electric vehicles (EVs) and hydrogen fuel cell vehicles (HFCVs), is expected to rise significantly in the near future. The energy demand of AFVs imposes an additional burden on the power system, which motivates the further expansion of renewable energy. The low-carbon transformation of power and transportation systems is inevitable but still in its early stages, constrained by high initial investment costs and multiple sources of uncertainty in renewable generation and energy demand. However, traditional planning models ignore the process of gradual low-carbon transformation of power and transportation systems and focus mainly on static planning decisions. Considering the development in the next few decades, it is more appropriate to plan in multiple periods. In this context, multi-stage decisions on investment timing, site selection, and capacity planning under multiple uncertainties become crucial.

[0003] Currently, the planning of hybrid charging stations (HCS) has the following shortcomings:

[0004] (1) Research on HCS planning has failed to effectively model the competition and substitution phenomena among traditional ICVs, HFCVs, and EVs.

[0005] (2) The planning issues of power and transportation systems for dynamic transformation are still less explored. The decommissioning of traditional gas stations (GS) and the construction and expansion of HCS lack coordination under a multi-stage dynamic framework.

[0006] (3) Endogenous and exogenous uncertainties have not been fully considered in the dynamic programming problem of multi-stage hybrid electric-hydrogen transportation station planning. Summary of the invention

[0007] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a hybrid electric-hydrogen transportation station planning method and system based on multi-stage distributed robust optimization, which comprehensively considers the endogenous uncertainty of the impact of the site selection of the hybrid electric-hydrogen transportation station on the planning objectives and the exogenous uncertainty of the wind power prediction error to solve the planning results of the hybrid electric-hydrogen transportation station and realize the transformation planning from traditional gas stations to hybrid charging stations.

[0008] The purpose of the present invention can be achieved by the following technical solutions:

[0009] One aspect of the present invention provides a hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization, comprising the following steps:

[0010] Build a collaborative planning model for low-carbon transformation based on planning objectives, investment constraints and operational constraints;

[0011] The endogenous uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on the penetration rate of new energy vehicles and the impact of gas station retirement decisions on the penetration rate of traditional fuel vehicles, as well as the exogenous uncertainty of wind power forecast errors, are modeled to obtain decision-dependent fuzzy sets;

[0012] Based on the decision-dependent fuzzy sets, the collaborative planning model is converted into a distributed robust multi-stage planning problem;

[0013] By solving the distributed robust multi-stage planning problem, the planning results of the hybrid electric hydrogen transportation stations in each stage are obtained.

[0014] As a preferred technical solution, the planning objectives are:

[0015]

[0016] in, and They represent the decommissioning cost, investment cost and operation cost of stage t respectively, is the recovery depreciation rate, Ω GS represents the set of existing traditional gas stations, is the refueling capacity of the retired traditional gas station, c m 、c u 、c s are the maintenance, unloading and recycling costs of traditional gas stations, r rt is the rate of decline in recovery cost, r m is the growth rate of maintenance cost, L i Plan the initial operating life of the gas station, c SCinv 、c d They are respectively, are indicator variables, which are used to indicate whether to build a new hybrid electric hydrogen transportation station at node i and whether to retire the traditional gas station at node i. is the capacity of the new hybrid electric hydrogen transportation station, Ω IP is a collection of mother hybrid electric hydrogen transportation stations used to supply hydrogen to other sub-hybrid electric hydrogen transportation stations in the region, c WT is the unit investment cost of the wind turbine, is the capacity of the new wind turbine, σ is the typical day scaling factor in the planning stage, is the time set, Ω d For device collection, is the unit operation cost of device d, is the operating power of device d, Ω SC For the sub-hybrid electric hydrogen transportation station collection, Ω Zones is the region set, Ω V is the vehicle set, c ct 、c um are the penalty costs for power curtailment and unmet demand, is the unsatisfied energy charge, Ω b is the device type set, c ls To reduce the cost of load, They are the amount of power abandonment and the amount of load reduction respectively.

[0017] As a preferred technical solution, the investment constraints include investment limit constraints, traditional gas station retirement order constraints, new hybrid electric-hydrogen transportation station order constraints, constraints on starting operation after investment, constraints on the supply of hydrogen by the main hybrid electric-hydrogen transportation station in the region for other sub-hybrid electric-hydrogen transportation stations, equipment installation capacity constraints and total capacity constraints for hydrogen refueling of sub-hybrid electric-hydrogen transportation stations based on the expansion of traditional gas stations.

[0018] As a preferred technical solution, the operating constraints include vehicle demand constraints, battery charging / discharging power constraints, battery storage level constraints, mother hybrid electric hydrogen transportation station hydrogen storage constraints, child hybrid electric hydrogen transportation station hydrogen storage constraints, distribution network power balance constraints, distribution network bus voltage constraints, and distribution network power factor constraints.

[0019] As a preferred technical solution, the endogenous uncertainty model of the impact of the site selection of the hybrid electric hydrogen transportation station on the penetration rate of new energy vehicles is:

[0020]

[0021] in, The site selection decision x for the hybrid electric hydrogen transportation station in stage t+1 t The growth rate of vehicle market share, Ω T is the stage set, Ω V is the vehicle set, ΩZones ,Ω z They are all regional collections. Indicates the location decision of the hybrid electric hydrogen transportation station x t The level of charging convenience under the influence of Indicates the location decision of the hybrid electric hydrogen transportation station x t Expected region-specific vehicle market shares under the impact, represents the basic diffusion rate, C z,j is the competition coefficient, which is used to capture the substitution effect between vehicle z and vehicle j, Δd z,v is the incentive factor, reflecting the perception level of vehicle v in region z on the location decision of the hybrid electric hydrogen transportation station, K i is the maximum theoretical market share of vehicle i, x i,t-1 For site selection decisions, is the initial region-specific vehicle market share, is the change in market share of region-specific vehicles.

[0022] As a preferred technical solution, the site selection decision x of the hybrid electric hydrogen transportation station at stage t t The decision-dependent fuzzy set under is:

[0023]

[0024] Where p represents the probability distribution, is a set of real numbers, p k is the probability, N t is the support set size, is an uncertain parameter, Ω V is the vehicle set, Ω z is a collection of regions, is the market share of vehicle v in region z, is the wind power prediction error of the i-th wind turbine, represents the first-order moment based on empirical data, is the second-order moment, and They are The mean and variance of t , ε t and is a parameter that limits the range of the fuzzy set.

[0025] As a preferred technical solution, the distributed robust multi-stage planning problem is:

[0026]

[0027] in, represents the state variables of the connection phases t and t+1, are indicator variables indicating whether to retire a traditional gas station or build a new hybrid electric hydrogen transportation station at node i, are the newly built capacity of equipment d and wind turbine i, respectively, t represents the operation variable, X t (x t-1 ,ξ t ) is a nonempty compact mixed integer polyhedron feasible set, {ξ1,…ξ T} are stage-independent random variables, where ξ1 is known with certainty, ξ t , Following the distribution pt, which lies in the decision-dependent fuzzy set In, for distributed Has a finite and phase-independent support set K| is the support set size.

[0028] As a preferred technical solution, the process of solving the distributed robust multi-stage planning problem includes the following steps:

[0029] Sampling: Sampling scenarios from the complete scenario tree constructed from the fuzzy set support set;

[0030] Forward pass: solve the forward problem and obtain an upper bound;

[0031] Backward pass: solve the backward problem and obtain the SB cut and lower bound;

[0032] The planning results of hybrid electric and hydrogen transportation stations are obtained through iterative solution.

[0033] As a preferred technical solution, the planning results of the hybrid electric-hydrogen transportation stations at each stage include the site selection of the hybrid electric-hydrogen transportation stations at each stage, the decision on the capacity of the hybrid electric-hydrogen transportation stations, and the decision on the retirement of existing traditional gas stations.

[0034] Another aspect of the present invention provides a hybrid electric hydrogen transportation station planning system based on multi-stage distributed robust optimization, comprising:

[0035] A collaborative planning model building module, which is used to build a collaborative planning model for low-carbon transformation based on planning objectives, investment constraints and operation constraints;

[0036] A module for converting distributed robust multi-stage planning problems, which is used to model the endogenous uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on the penetration rate of new energy vehicles and the impact of gas station retirement decisions on the penetration rate of traditional fuel vehicles, as well as the exogenous uncertainty of wind power prediction errors, and obtain decision-dependent fuzzy sets. Based on the decision-dependent fuzzy sets, the collaborative planning model is converted into a distributed robust multi-stage planning problem;

[0037] The hybrid electric-hydrogen transportation station planning result solving module is used to obtain the hybrid electric-hydrogen transportation station planning results of each stage by solving the distributed robust multi-stage planning problem.

[0038] Compared with the prior art, the present invention has at least one of the following beneficial effects.

[0039] (1) Realize the transformation planning from traditional gas stations to hybrid electric hydrogen transportation stations: The present invention first constructs a collaborative planning model for low-carbon transformation based on planning objectives, investment constraints and operation constraints, and then models the endogenous uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on planning objectives and the exogenous uncertainty of wind power prediction errors to obtain decision-dependent fuzzy sets, and converts the collaborative planning model into a distributed robust multi-stage planning problem. Finally, by solving the distributed robust multi-stage planning problem, the planning results of the hybrid electric hydrogen transportation stations at each stage are obtained, thereby realizing the transformation planning from traditional gas stations to hybrid electric hydrogen transportation stations.

[0040] (2) Fully consider the competition and substitution phenomena among traditional internal combustion locomotives, hydrogen fuel cell vehicles and electric vehicles: Considering that the growth of the market share of a certain type of vehicle will limit the potential growth of other types of vehicles, the present invention models the inherent uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on planning objectives, and fully considers this impact.

[0041] (3) The replacement process from traditional gas stations to hybrid electric hydrogen transportation stations is smooth: Considering that low-carbon transformation usually takes several years, sudden retirement plans may lead to serious supply shortages. The present invention converts the collaborative planning model into a distributed robust multi-stage planning problem, and coordinates the retirement of traditional gas stations with the construction and expansion of hybrid electric hydrogen transportation stations in a multi-stage dynamic framework.

[0042] (4) Fully consider endogenous uncertainty and exogenous uncertainty: In the implementation process of multi-stage dynamic planning, the present invention fully considers the endogenous uncertainty of the impact of the site selection of the hybrid electric hydrogen transportation station on the planning objectives and the exogenous uncertainty of the wind power prediction error, thereby improving the safety and economy of the system and ensuring that more reasonable decisions are made in a constantly changing environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of a hybrid electric hydrogen transportation station planning method based on multi-stage distributed blue rod optimization in an embodiment;

[0044] Figure 2 A schematic diagram of a system for verifying the application effect of the present invention in the embodiment;

[0045] Figure 3 It is a cost decomposition diagram of each stage in the embodiment;

[0046] Figure 4 Schematic diagram of a hybrid electric-hydrogen transportation station planning system based on multi-stage distributed blue-rod optimization in an embodiment. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0048] Example 1

[0049] In view of the shortcomings of the prior art, this embodiment provides a hybrid electric hydrogen transportation station planning method based on multi-stage distributional robust optimization, which is applied to the power-transportation coupling system. The method considers the competitive market share of vehicles and adopts multi-stage distributional robust programming. This method constructs a dynamic automobile market share model based on the three-dimensional competitive Lotka-Volterra equation, and then constructs a decision-dependent fuzzy set based on it, so as to clearly capture the endogenous uncertainty of the diffusion of alternative fuel vehicle market share and the contraction of internal combustion vehicle market share. Subsequently, a nested multi-stage decision-related distributional robust programming problem is established for the power grid-transportation collaborative planning problem. In order to effectively solve the computational challenges of the proposed problem, a customized solution algorithm based on enhanced Benders cut and stochastic dual dynamic integer programming algorithm (Stochastic Dual Decomposition Integer Programming, SDDIP) is provided.

[0050] See also Figure 1 , the method comprises the following steps:

[0051] Step S1, constructing a collaborative planning model for low-carbon transformation: based on planning objectives, investment constraints and operation constraints, constructing a collaborative planning model for low-carbon transformation.

[0052] (1) Objective function.

[0053] The objective function of each stage is to minimize the total net current cost:

[0054]

[0055] In the formula, and They represent the decommissioning cost, investment cost and operating cost of stage t respectively, and are given by equations (1.2) to (1.4). Includes the dismantling cost and recycling income of gasoline dispensers. Includes the fixed cost of HCS construction and the cost of purchasing equipment. Includes GS maintenance costs, HCS operation costs, and penalties for power abandonment, load shedding, and unmet fuel demand.

[0056]

[0057] (2) Investment constraints.

[0058] Constraint (1.5) limits the total investment. Constraints on the order of GS retirement and HCS investment are (1.6) and (1.7), respectively. Constraint (1.8) states that once the investment is made, the HCS will be put into operation. Constraint (1.9) limits each region to have at least one operating parent HCS (Internal Production-based HCSs, IP-HCSs) to supply hydrogen to other child HCSs (Supply Chain-based HCSs, SC-HCSs).

[0059]

[0060] The installed capacity of HCS, GS and wind turbine (WT) is calculated by (1.10)-(1.12) respectively, and the installed capacity of the equipment is constrained by (1.13)-(1.14). For SC-HCS based on GS expansion, the total capacity of hydrogen refueling is constrained by (1.15) due to land use restrictions.

[0061]

[0062]

[0063] (3) Operational constraints.

[0064] The operating power of the equipment is limited by its capacity, as shown in (1.16). For each area, the vehicle demand constraint should satisfy (1.17). The dynamic storage level of the battery is given by (1.18), and the charging and discharging powers are limited by (1.19)-(1.20), respectively. The battery storage level is limited by (1.21). The dynamics of the hydrogen stations in IP-HCS and SC-HCS are represented by (1.22)-(1.23), respectively. The hydrogen storage level is limited by (1.24). The linearized model of the distribution network operation is given by (1.25)-(1.31). Specifically, (1.25) is the power balance constraint. (1.26) represents the relationship between the bus voltage and the power flow. The constraints on the power flow and bus voltage are given by (1.27)-(1.28), respectively, and (1.29) represents the power factor constraint. Equations (1.30)-(1.31) calculate the load power of IP-HCS and SC-HCS respectively.

[0065]

[0066]

[0067] Step S2, construct a distributed robust multi-stage planning framework: model the endogenous uncertainty of the impact of hybrid electric hydrogen transportation station site selection on the penetration rate of new energy vehicles and the exogenous uncertainty of wind power forecast errors to obtain decision-dependent fuzzy sets.

[0068] In this step, a detailed model is first proposed to capture the endogenous uncertainty in the vehicle market share affected by the location decision of the hybrid electric hydrogen transportation station. Then, fuzzy sets that consider endogenous and exogenous uncertainties are proposed. On this basis, the planning model developed in step S1 is extended to a distributed robust multi-stage planning problem.

[0069] (1) Endogenous uncertainty.

[0070] In order to capture the competition and mutual inhibition phenomenon between different vehicle market shares, a three-dimensional competitive Lotka-Volterra model is adopted, and the growth rate of vehicle market share is modeled as follows.

[0071]

[0072] in, Indicates the level of charging convenience, represents the expected region-specific vehicle market share. Both are affected by the HCS location decision (denoted as x t ) and are explained in detail in (1.33)-(1.34). represents the basic diffusion rate. C z,jis a competition coefficient that captures the substitution effect between vehicle z and vehicle j. If C z,j >0, there is a competitive relationship between vehicle z and vehicle j.

[0073]

[0074] Where, Δd z,v is an incentive factor that reflects the level of influence of the HCS location decision on vehicle v in region z. Based on the definition of growth rate in (1.32), the current expected market share is modeled as t A piecewise linear function of :

[0075]

[0076] Note that in (1.32) and (1.34) there is t To solve this problem, we assume that the additional market share driven by the location decision of the hybrid electric hydrogen transportation station is negligible relative to the potential market share K. This assumption is reasonable in the early stages of the transformation process. Therefore, (1.34) It can be approximated by the incremental diffusion rate, denoted as Based on this, (1.32) and (1.34) are transformed into (1.35)-(1.36), respectively.

[0077]

[0078] For ease of representation, a parameter is introduced Defined as (1.37)

[0079]

[0080] By substituting (1.37) into (1.36), we can obtain the expected vehicle market share, i.e., the first-order moment information, as shown in (1.38). It should be noted that It is linearly related to the number of GS, while It is linearly related to the number of HCS.

[0081]

[0082] The second-order moment information of the vehicle market share is given by (1.39) and it is also decision-dependent.

[0083]

[0084] in, represents the empirical standard deviation of vehicle market share. Represents the impact of location decisions on vehicle market share fluctuations.

[0085] (2) Decisions depend on fuzzy sets.

[0086] Using the moment information obtained above, the fuzzy set of decision dependence at stage t is defined as follows.

[0087]

[0088] The uncertainty parameter is The first term is the market share of vehicle v in region z, and the second term represents the wind power forecast error. represents the first-order moment of experience, is the second-order moment, where and They are The mean and variance of ε t , ε t and is a parameter that limits the range of the fuzzy set.

[0089] The decision-dependent fuzzy set has several advantages. First, by combining the three-dimensional competitive Lotka-Volterra model, the competition effect between different types of vehicles is captured. Second, the fuzzy set considers both the endogenous uncertainty in the dynamic market share and the exogenous uncertainty in the wind power forecast error.

[0090] Step S3, constructing a distributed robust multi-stage planning problem based on the decision-dependent fuzzy set obtained in step S2.

[0091] Based on the proposed fuzzy sets, a distributionally robust multi-stage planning problem is proposed, and the goal is to minimize the total net present value cost in the entire planning period.

[0092]

[0093] in, Represents the state variable of the connection phase t and t+1. t Indicates the running variable. X t (x t-1 ,ξ t ) is a nonempty and compact mixed integer polyhedron feasible set. {ξ1,…ξ T} are stage-independent random variables, where ξ1 is known with certainty, and ξ t , Following the distribution p t , which is located in the fuzzy set of decision dependence , as defined in (40). distributed Has a finite and phase-independent support set

[0094] Due to the nested structure of the multi-stage optimization and the non-convexity of the proposed fuzzy sets, the problem is computationally infeasible and cannot be directly solved by any off-the-shelf optimization solver. Therefore, a new solution algorithm is needed to address this computational challenge.

[0095] Step S4, by solving the distributed robust multi-stage planning problem, obtain the planning results of the hybrid electric hydrogen transportation station in each stage.

[0096] (1) Equivalent feasible transformation.

[0097] Firstly, the original multi-stage distributed robustness problem is transformed into the Bellman equation form.

[0098]

[0099] for

[0100]

[0101] Furthermore, in order to utilize the SDDIP algorithm, the integer state variables must be binarized. The integer device capacity can be binarized as where j n ∈{0,1}, K represents the number of binary bits. Other state variables, such as Already binary, so no modification is needed.

[0102] Based on the duality theorem, the equivalent reconstruction form of (1.43) is as follows.

[0103]

[0104] (x t ,y t )∈X t (x t-1 ,ξ t ),α t ,Ω t ≥0 (1.47)

[0105] in, and is a dual variable. Equation (1.45) contains the bilinear term and This problem can be solved by introducing continuous auxiliary variables. Taking the first term as an example, define Due to x t,i is a 0-1 variable, this bilinear term can be easily linearized by:

[0106]

[0107] Where M is a sufficiently large integer.

[0108] (2)SDDIP decomposition algorithm.

[0109] The problem after equivalent reconstruction still faces computational infeasibility caused by the nested formulas in (1.42)-(1.44). Therefore, the MSD3RO-SDDIP algorithm is designed to solve this problem.

[0110] It should be noted that Q in (1.43) t+1 (x t ,ξ t+1 ) is a nested cost function, which is unknown at stage t. To solve this problem, the following Strengthened Benders' (SB) cut is adopted.

[0111]

[0112] Among them, ν t+1 and is the coefficient of the cutting plane. Therefore, we can obtain a lower bound approximation problem (denoted as) for (1.45)-(1.47) by replacing (1.46) with the following formula ).

[0113]

[0114] It is a traditional single-stage mixed integer linear programming (MILP). The detailed process of the algorithm is shown in Table 1, which includes four steps: 1) Sampling: sampling scenarios from the complete scenario tree constructed by the fuzzy set support set, 2) Forward pass: solving the forward problem and obtaining the upper bounds (UBs), 3) Backward pass: solving the backward problem and obtaining the SB cut and lower bounds (LBs), 4) Stopping criterion.

[0115] Table 1 MSD3RO-SDDIP algorithm

[0116]

[0117] The following numerical tests are conducted to demonstrate the effectiveness of this method. Three investment phases based on the IEEE-33 node system are used for testing. The experiments are conducted on an Intel(R) Core(TM) i7 4.70GHz personal computer equipped with 16G memory. The program is written in Pyomo 6.7.3 and the optimization problem is solved using GUROBI 10.0.2.

[0118] In order to demonstrate the superiority of the proposed method, Figure 2 Under the system shown, five cases are set for comparative analysis. Specifically, M1: the method proposed in this embodiment, M2: planning without considering vehicle market share, M3: planning without considering endogenous uncertainty, M4: planning without considering GS retirement, and M5: planning without considering the joint construction of GS and HCS.

[0119] Table 2 shows the planning results under different cases. The experimental results show that in the second stage, all methods built four IP-HCSs to meet the hydrogen demand in the three regions. The results of M1 show the most new SC-HCSs and retired GSs. M2 invested in a SC-HCS co-built with GS, which is a conservative result because it ignores the competitive dynamics between different types of vehicles. M3 only considers exogenous uncertainty and is more conservative in new construction and retirement decisions. In M4, the hydrogen refueling capacity of a single co-built SC-HCS is limited because the retirement of GS is not considered, so more GS are converted into co-built SC-HCSs. Since M5 prohibits the co-building of SC-HCSs, it results in the construction of only a single SC-HCS. In the third stage, as the demand for new energy in transportation increases further, all methods expand or build new HCSs. M2 has a slow deployment in the second stage, so it compensates by quickly building independent SC-HCSs to meet energy demand. Since M3 ignores the uncertainty of vehicle market share, it takes a more conservative approach to GS retirement and new HCS construction than M1. M4 relies heavily on converting GS into co-built HCSs, and ultimately the planning results under this method have the most co-built HCSs.

[0120] Table 2 Planning results of each stage

[0121]

[0122] exist Figure 3 In the table, we show the breakdown of planning costs under different methods. It can be seen that the main part of the total cost comes from investment and operating costs. The method proposed in this embodiment achieves the lowest total cost. M2 and M3 incur higher operating costs because their conservative planning results lead to high penalties for unmet energy demand. This highlights the importance of endogenous uncertainty in HCS planning models. M4 and M5 incur higher investment costs, which is consistent with the planning results listed in Table 1.

[0123] Table 3 is the definition of relevant variables in the examples.

[0124] Table 3 Variable definitions

[0125]

[0126] Example 2

[0127] Based on Example 1, see Figure 4 This embodiment provides a hybrid electric hydrogen transportation station planning system based on multi-stage distributed blue stick optimization, including:

[0128] (1) Collaborative planning model building module: used to build a collaborative planning model for low-carbon transformation based on planning objectives, investment constraints, and operation constraints;

[0129] (2) Distributed Robust Multi-stage Planning Problem Conversion Module: used to model the endogenous uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on the penetration rate of new energy vehicles and the impact of gas station retirement decisions on the penetration rate of traditional fuel vehicles, as well as the exogenous uncertainty of wind power prediction errors, and obtain decision-dependent fuzzy sets. Based on the decision-dependent fuzzy sets, the collaborative planning model is converted into a distributed robust multi-stage planning problem;

[0130] (3) Hybrid electric-hydrogen transportation station planning result solving module: used to obtain the hybrid electric-hydrogen transportation station planning results of each stage by solving the distributed robust multi-stage planning problem.

[0131] The present invention has the following characteristics:

[0132] (1) Aiming at low-carbon transformation, the problem of coordinated planning of power system and HCS is modeled as a multi-stage distributed robust model for the first time. This model dynamically optimizes the configuration of different sites to achieve an orderly and smooth transformation path under multi-source uncertainty.

[0133] (2) In order to accurately describe the evolution of competitive market share of different vehicles, a decision-dependent three-dimensional Lotka-Volterra model is proposed. On this basis, a decision-dependent fuzzy set is constructed to capture the endogenous uncertainty of the changing charging demand.

[0134] (3) By leveraging strong duality theory and equivalent linearization techniques, the stage-by-stage decision-dependent distributionally robust optimization problem is reformulated as a MILP. Subsequently, a customized solution is developed based on the SDDIP algorithm to efficiently handle the nested nature of the multi-stage distributionally robust optimization problem with good computational performance.

[0135] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A hybrid electric hydrogen transportation station planning method based on multi-stage distributed blue stick optimization, characterized in that: The steps include: Build a collaborative planning model for low-carbon transformation based on planning objectives, investment constraints and operational constraints; The endogenous uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on the penetration rate of new energy vehicles and the impact of gas station retirement decisions on the penetration rate of traditional fuel vehicles, as well as the exogenous uncertainty of wind power forecast errors, are modeled to obtain decision-dependent fuzzy sets; Based on the decision-dependent fuzzy sets, the collaborative planning model is converted into a distributed robust multi-stage planning problem; By solving the distributed robust multi-stage planning problem, the planning results of the hybrid electric hydrogen transportation stations in each stage are obtained.

2. According to claim 1, a hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization is characterized in that: The planning objectives are: in, and They represent the retirement cost, investment cost and operation cost of stage t respectively, ω is the recovery depreciation rate, Ω GS represents the set of existing traditional gas stations, is the refueling capacity of the retired traditional gas station, c m 、c u 、c s are the maintenance, unloading and recycling costs of traditional gas stations, r rt is the decline rate of recovery cost, r m is the growth rate of maintenance cost, L i Plan the initial operating life of the gas station, c SCinv 、c d They are respectively, are indicator variables, which are used to indicate whether to build a new hybrid electric hydrogen transportation station at node i and whether to retire the traditional gas station at node i. is the capacity of the new hybrid electric hydrogen transportation station, Ω IP is a collection of mother hybrid electric hydrogen transportation stations used to supply hydrogen to other sub-hybrid electric hydrogen transportation stations in the region, c WT is the unit investment cost of the wind turbine, is the capacity of the new wind turbine, σ is the typical day scaling factor in the planning stage, is the time set, Ω d For device collection, is the unit operation cost of device d, is the operating power of device d, Ω SC For the sub-hybrid electric hydrogen transportation station collection, Ω Zones is the region set, Ω V is the vehicle set, c ct 、c um are the penalty costs for power curtailment and unmet demand, is the unsatisfied energy charge, Ω b is the device type set, c ls To reduce the cost of load, They are the amount of power abandonment and load reduction respectively.

3. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The investment constraints include investment limit constraints, traditional gas station retirement order constraints, new hybrid electric-hydrogen transportation station order constraints, constraints on starting operation after investment, constraints on the supply of hydrogen by the main hybrid electric-hydrogen transportation station in the region for other sub-hybrid electric-hydrogen transportation stations, equipment installation capacity constraints and total capacity constraints for hydrogen refueling of sub-hybrid electric-hydrogen transportation stations based on the expansion of traditional gas stations.

4. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The operating constraints include vehicle demand constraints, battery charging / discharging power constraints, battery storage level constraints, main hybrid electric-hydrogen transportation station hydrogen storage constraints, sub-hybrid electric-hydrogen transportation station hydrogen storage constraints, distribution network power balance constraints, distribution network bus voltage constraints, and distribution network power factor constraints.

5. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The endogenous uncertainty model of the impact of the site selection of the hybrid electric hydrogen transportation station on the penetration rate of new energy vehicles is: in, The site selection decision x for the hybrid electric hydrogen transportation station in stage t+1 t The growth rate of vehicle market share, Ω T is the stage set, Ω V is the vehicle set, Ω Zones ,Ω z They are all regional collections. Indicates the location decision of the hybrid electric hydrogen transportation station x t The level of charging convenience under the influence of Indicates the location decision of the hybrid electric hydrogen transportation station x t Expected region-specific vehicle market shares under the impact, represents the basic diffusion rate, C z,j is the competition coefficient, which is used to capture the substitution effect between vehicle z and vehicle j, Δd z,v is the incentive factor, reflecting the perception level of vehicle v in region z on the location decision of the hybrid electric hydrogen transportation station, K i is the maximum theoretical market share of vehicle i, x i,t-1 For site selection decisions, is the initial region-specific vehicle market share, is the change in market share of region-specific vehicles.

6. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The site selection decision x for the hybrid electric-hydrogen transportation station at stage t t The decision-dependent fuzzy set under is: Where p represents the probability distribution, is a set of real numbers, p k is the probability, |K| is the support set size, is an uncertain parameter, Ω V is the vehicle set, Ω z is a collection of regions, is the market share of vehicle v in region z, is the wind power prediction error of the i-th wind turbine, represents the first-order moment based on empirical data, is the second-order moment, and They are The mean and variance of t , ε t and is a parameter that limits the range of the fuzzy set.

7. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The distributed robust multi-stage planning problem is: in, represents the state variables of the connection phases t and t+1, are indicator variables indicating whether to retire a traditional gas station or build a new hybrid electric hydrogen transportation station at node i, are the newly built capacity of equipment d and wind turbine i, respectively, t represents the operation variable, X t (x t*1 ,ξ t ) is a nonempty compact mixed integer polyhedron feasible set, {ξ1,…ξ T } are stage-independent random variables, where ξ1 is known with certainty, Following the distribution pt, which lies in the decision-dependent fuzzy set In, for distributed Has a finite and phase-independent support set K| is the support set size.

8. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The process of solving the distributed robust multi-stage planning problem includes the following steps: Sampling: Sampling scenarios from the complete scenario tree constructed from the fuzzy set support set; Forward pass: solve the forward problem and obtain an upper bound; Backward pass: solve the backward problem and obtain the SB cut and lower bound; The planning results of hybrid electric and hydrogen transportation stations are obtained through iterative solution.

9. The hybrid electric hydrogen transportation station planning method based on multi-stage distributed robust optimization according to claim 1 is characterized in that: The planning results of the hybrid electric-hydrogen transportation stations at each stage include the site selection of the hybrid electric-hydrogen transportation stations at each stage, the decision on the capacity of the hybrid electric-hydrogen transportation stations, and the decision on the retirement of existing traditional gas stations.

10. A hybrid electric hydrogen transportation station planning system based on multi-stage distributed blue stick optimization, characterized in that: include: A collaborative planning model building module, which is used to build a collaborative planning model for low-carbon transformation based on planning objectives, investment constraints and operation constraints; A module for converting distributed robust multi-stage planning problems, which is used to model the endogenous uncertainty of the impact of the site selection of hybrid electric hydrogen transportation stations on the penetration rate of new energy vehicles and the impact of gas station retirement decisions on the penetration rate of traditional fuel vehicles, as well as the exogenous uncertainty of wind power prediction errors, and obtain decision-dependent fuzzy sets. Based on the decision-dependent fuzzy sets, the collaborative planning model is converted into a distributed robust multi-stage planning problem; The hybrid electric-hydrogen transportation station planning result solving module is used to obtain the hybrid electric-hydrogen transportation station planning results of each stage by solving the distributed robust multi-stage planning problem.

Images