CN-121981772-A - Distributed robust optimization method for site selection of hydrogen station facilities

CN121981772ACN 121981772 ACN121981772 ACN 121981772ACN-121981772-A

Abstract

The invention discloses a distributed robust optimization method for site selection of a hydrogen station facility, which comprises the steps of 1) constructing a random planning model and corresponding constraint conditions for site selection of the hydrogen station facility with the aim of minimizing hydrogen energy infrastructure cost and transportation cost, 2) constructing an uncertainty fuzzy set on probability distribution of two random variables of hydrogen demand and pipeline interruption, 3) limiting probability distribution of a scene on the uncertainty fuzzy set by adopting a distributed robust optimization method, and establishing a distributed robust optimization model, 4) converting the model into a main problem and a sub-problem by adopting a mathematical method of linear planning and integer planning, and solving by adopting a column and constraint generation algorithm to obtain the site selection position of the hydrogen station facility. The method can effectively solve the site selection model of the hydrogen station, and reduces the comprehensive cost.

Inventors

PEI ZHI
ZHOU QIZHONG
LU HAIMIN
CHEN YONG
LI SHIYUN
YI WENCHAO

Assignees

浙江工业大学

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (6)

1. A distributed robust optimization method for site selection of a hydrogen addition station facility, comprising the steps of: Step 1, constructing a random planning model of site selection of a hydrogen station facility and corresponding constraint conditions by taking minimum hydrogen energy infrastructure cost and minimum transportation cost as targets; Step 2, extracting key parameters in historical data, and carrying out uncertainty modeling on probability distribution of two random variables of hydrogen demand and pipeline interruption so as to define a distribution fuzzy set to which the hydrogen demand and the pipeline interruption belong; step 3, limiting probability distribution of a scene by adopting a distributed robust optimization method according to the distribution fuzzy set of two types of random variable uncertainties corresponding to the uncertainty modeling in the step 2, so as to obtain a hydrogenation station infrastructure network model based on the distributed robust optimization; And 4, converting the model in the step 3 into a main problem and a sub problem by adopting a mathematical method of linear programming and integer programming, wherein the main problem is an optimal solution meeting constraint conditions under the current scene, the sub problem is a worst scene under two uncertain factors of hydrogen demand and pipeline interruption, which is searched for, and the worst scene is returned to the main problem, and the main problem is solved by adopting a column and constraint generation algorithm, so that the site selection position of the facilities of the hydrogenation station is finally obtained.
2. The distributed robust optimization method for site selection of a hydrogen plant according to claim 1, wherein in step 1, a two-stage objective function is constructed, wherein one-stage objective function is shown in formula (1), and two-stage objective function is shown in formula (10); ; the objective function (1) is to minimize the construction cost of the hydrogen adding station and the construction cost of the pipeline and the total distance of the vehicle queue to the hydrogen adding station; The objective function (10) is the sum of the cost of minimizing pipeline transportation, the cost of transporting a long tube trailer, and the penalty cost for the unmet need; Symbol definition Aggregation: a hydrogen manufacturing production factory set, wherein the index is I; J, a hydrogen adding station candidate point set, wherein the index is J; g, constructing a scale set of the hydrogen adding station, wherein the index is G; k, a vehicle queue set with hydrogen demand, wherein the index is K; e, candidate hydrogen adding station pairs with the distance smaller than the minimum safety distance; parameters: The cost of building a hydrogen addition station of scale g; Pipeline transportation cost is adopted from a hydrogen manufacturing production factory i to a hydrogenation station j; Cost of transportation by adopting a long tube trailer from a hydrogen manufacturing production factory i to a hydrogenation station j; The unit penalty cost for the demand not being met; The hydrogen leakage rate of the pipeline transportation mode; Hydrogen leakage rate of a long tube trailer transportation mode; The cost of laying pipelines from hydrogen manufacturing production factory i to hydrogenation station j; Maximum hydrogen production amount of hydrogen production factory i; Hydrogen storage capacity of the hydrogen station with the scale g; h, minimum safe distance between two hydrogenation stations; the distance between the vehicle queue k and the hydrogen adding station j; candidate hydrogen adding station j A distance therebetween; the maximum hydrogen flow rate from the hydrogen production factory i to the hydrogenation station j through the pipeline; q is hydrogen consumption per unit distance of the vehicle queue; Pipeline interruption risk budget parameters; If the pipeline interruption occurs between the hydrogen production factory i and the hydrogenation station j, the pipeline interruption is 1, otherwise, the pipeline interruption is 0; hydrogen demand of the vehicle queue k; decision variables: binary variables, 1 if a hydrogen adding station with the scale g is built at a j candidate point, or 0 if the hydrogen adding station is built at the j candidate point; The hydrogen amount transported from the hydrogen production factory i to the hydrogenation station j by adopting a long pipe trailer; the hydrogen quantity transported by adopting a pipeline from a hydrogen production factory i to a hydrogenation station j; Binary variable, if a pipeline is laid between a hydrogen production factory i and a hydrogenation station j, the binary variable is 1, otherwise, the binary variable is 0; binary variable, if the vehicle queue k selects the hydrogenation station j to hydrogenate, the binary variable is 1, otherwise, the binary variable is 0; hydrogen demand at the hydrogen addition station j is not satisfied; Vector: Respectively is And Is a vector or matrix representation of (c).
3. A distributed robust optimization method for site selection of a hydrogen plant according to claim 2, characterized in that the constraints of the objective function (1) are as shown in formulas (2) - (9); ; ; Wherein, the ; Constraint equation (2) indicates that the pipeline will only be established between the activated hydrogen plant and the hydrogen production plant; constraint formula (3) ensures that each selected hydrogen adding station can only select one scale to build stations; constraint equation (4) ensures that the safe distance between the two activated hydrogen stations is greater than the minimum safe distance; constraint equation (5) ensures that a vehicle queue can only be hydrogenated at one hydrogen station; Constraint equation (6) ensures that the vehicle queue can only be hydrogenated at the hydrogenated station that has been selected for establishment; constraint formulas (7) - (9) represent types of variables; The feasible region formed by constraint formulas (2) - (9) is defined as ; The constraint condition of the objective function (10) is shown in formulas (11) - (17); ; ; Constraint equation (11) shows that hydrogen is transported through the pipeline only if the pipeline is established and no interruption occurs; Constraint formula (12) ensures that the supply amount of hydrogen does not exceed the hydrogen storage scale of the hydrogen addition station; constraint equation (13) ensures that the supply of hydrogen does not exceed the maximum throughput of the hydrogen production plant; constraint equation (14) shows that penalty costs are incurred if the hydrogen storage capacity of the hydrogen addition station is insufficient to supply the hydrogen demand and the route hydrogen consumption of the vehicle queue, i.e., the sum of the total amount of hydrogen transported to the hydrogen addition station and the insufficient hydrogen amount needs to be greater than or equal to the hydrogen demand and the route hydrogen consumption of the supply vehicle queue; Constraint formulas (15) - (17) represent types of variables; The feasible regions of constraint formulas (11) - (17) are defined as 。
4. A distributed robust optimization method for site selection of a hydrogen plant according to claim 3, wherein in step 2, the process of constructing the uncertainty factor fuzzy set is as follows: The distribution ambiguity set definition of the hydrogen demand uncertainty is as shown in equation (18): ; Wherein, the Random vector representing hydrogen demand, its components Indicating the hydrogen demand of the kth hydrogen energy demand fleet; Random vector representing hydrogen demand Probability distribution of (2); Constraint is carried out on a lower limit vector and an upper limit vector of hydrogen demand respectively Indicating that all possible hydrogen demand facts fall within the bounded region; an average value vector representing the hydrogen demand, the components of which are Constraint(s) The first moment of the uncertainty distribution representing the probability distribution of the random variable of the hydrogen demand is known , An upper bound vector representing the mean absolute deviation of the hydrogen demand, the components of which are Constraint(s) A degree of dispersion of the probability distribution of the random variable representing the hydrogen demand around its mean value; The distribution ambiguity set definition of uncertainty of whether a pipe is broken is as shown in equation (19): ; Wherein, the A binary random variable representing whether the pipe is interrupted; Representing a set of pipeline interrupt random variables Is a joint probability distribution of (1); Indicating the total number of simultaneous interrupt pipes, i.e ; Constraints indicating whether a pipeline is interrupted or not Indicating that in all possible distributions, the total number of pipes for which an interruption is desired does not exceed 。
5. The distributed robust optimization method for site selection of a hydrogen plant of claim 4, wherein in step 3, a two-stage distributed robust optimization model is constructed based on the two-stage objective function and the two types of random variable uncertainty distribution ambiguity sets, the objective function of which is shown in formula (20): ; the corresponding constraints are shown in formulas (2) - (9) and formulas (11) - (17).
6. The distributed robust optimization method for site selection of a hydrogen plant of claim 5, wherein the two-stage distributed robust optimization model, two-stage problem, can be expressed as shown in equation (21): ; ; ; Couple equation (21) and couple the variables Is combined into The conversion yields equation (22): ; Definition of the definition Then ; Order the ; Equation (22) is for the inclusion minimization problem Is to maximize the problem (21) Solving dual; Due to Is a minimization problem and equation (23) is a maximization problem, thus for Solving the pair to convert the minimization problem into the maximization problem to obtain the formula (24) and the formulas (25) - (31) of the constraint conditions of the formula (24); ; ; ; Wherein the dual variables The two-pair variables of constraints (11) - (14), respectively, whose vector or matrix elements are respectively ; Substituting the dual problem formula (24) and constraint conditions thereof into the formula In (1) can obtain To maximize the problem, according to the linear programming pair theory and extreme point principle, in Take the value of upper critical value, lower critical value or When due to And (3) with Can be made to be Take the maximum, i.e. the optimal solution Belongs to a collection , Thus making the lead Wherein , ; And is introduced into Substituted into After that, observe 、 And Are the products of 0-1 integer variable and continuous variable, and are introduced because the integer variable and the continuous variable are multiplied to cause nonlinearity, so that linearization treatment is carried out on the integer variable and the continuous variable 、 And Linearizing it to obtain Linearization expression: ; The corresponding constraints are shown in formulas (25) - (31) and formulas (33) - (45); ; ; ; Wherein B represents an infinite number; expressions (34) - (35) are introduced Then linearizing the constraint formula; expressions (36) - (41) are linearization constraints that constrain bilinear nonlinear terms according to the mccomick envelope theory; So far as the process is concerned, The method is a mixed integer linear programming problem, and can be directly solved on a Python platform by using Gurobi solver; Sub-problems such as formula (23) Shown, and The function is a maximization problem, but includes a minimization problem The function is to be first to The function-finding pair converts the minimization problem into the maximization problem so that The function is transformed into a maximization problem as shown in (32) so that it can be solved; Definition of the definition Is that Due to the aggregation of (3) And Are all finite discrete sets of size And Definition of , Will be Any one of the sets And The pair of components is defined as a scene Then all are in common Each scene of All belong to However when When very large, the problem is virtually insoluble, thus defining a subset The objective function (20) of the two-stage distributed robust optimization model and its constraints can be expressed as: ; Equation (46) is an expression of the main problem obtained by the final transformation.

Description

Distributed robust optimization method for site selection of hydrogen station facilities Technical Field The invention relates to a distributed robust optimization method for site selection of a hydrogen station facility, which simultaneously considers various uncertainty factors. Background The large-scale commercial popularization of hydrogen fuel cell automobiles is highly dependent on a perfect, reliable, economical and efficient hydrogenation infrastructure network. Currently, the planning and construction of a hydro-station network faces a series of serious challenges and inherent contradictions. On one hand, the hydrogen station and the matched hydrogen delivery pipeline belong to heavy asset projects with high investment and long period, the hydrogen fuel requirement is influenced by multiple factors such as vehicle popularization progress, regional operation mode and the like, the method has obvious uncertainty and is easy to cause mismatching of investment benefits and market requirements, and on the other hand, the hydrogen supply needs to comprehensively plan two modes of pipeline and road transportation, and has the characteristics in cost, reliability and applicable scenes, and the cooperative optimization needs to be realized in planning. The conventional planning method is often based on deterministic assumption or a single stochastic model, and multiple uncertain factors such as requirement fluctuation, multiple transmission path coupling, interruption risk and safety constraint are difficult to comprehensively process, so that the planning scheme faces risks of insufficient economy and robustness in actual operation. Therefore, the invention focuses on the core problems of 'source-storage-transmission-addition' multi-link collaborative planning and multiple uncertainty management in a hydrogen energy supply chain, and provides a hydrogen energy infrastructure network design method based on two-stage distribution robust optimization. By constructing a distributed robust model of two kinds of uncertainties of demand randomness and pipeline transportation interruption, the system integrates a plurality of decision levels of hydrogen production supply, multi-mode transportation, hydrogenation station site selection, volume fixation and the like. And designing a column and constraint generation (C & CG) algorithm for solving. Under the condition of incomplete uncertainty probability distribution information, the method realizes comprehensive optimization and robust balance of the construction cost, the operation efficiency and the system toughness of the supply chain, and provides systematic decision theory and method support for constructing a hydrogen energy infrastructure network which is economical, efficient, reliable and stable and has risk resisting capability. Disclosure of Invention Aiming at the technical problems existing in the prior art, the invention aims to provide a distributed robust optimization method for site selection of a hydrogen station facility. The technical scheme adopted by the invention is as follows: A distributed robust optimization method for site selection of a hydrogen addition station facility, comprising the steps of: Step 1, constructing a random planning model of site selection of a hydrogen station facility and corresponding constraint conditions by taking minimum hydrogen energy infrastructure cost and minimum transportation cost as targets; Step 2, extracting key parameters in historical data, and carrying out uncertainty modeling on probability distribution of two random variables of hydrogen demand and pipeline interruption so as to define a distribution fuzzy set to which the hydrogen demand and the pipeline interruption belong; step 3, limiting probability distribution of a scene by adopting a distributed robust optimization method according to the distribution fuzzy set of two types of random variable uncertainties corresponding to the uncertainty modeling in the step 2, so as to obtain a hydrogenation station infrastructure network model based on the distributed robust optimization; And 4, converting the model in the step 3 into a main problem and a sub problem by adopting a mathematical method of linear programming and integer programming, wherein the main problem is an optimal solution meeting constraint conditions under the current scene, the sub problem is a worst scene under two uncertain factors of hydrogen demand and pipeline interruption, which is searched for, and the worst scene is returned to the main problem, and the main problem is solved by adopting a column and constraint generation algorithm, so that the site selection position of the facilities of the hydrogenation station is finally obtained. Symbol definition, collection: a hydrogen manufacturing production factory set, wherein the index is I; J, a hydrogen adding station candidate point set, wherein the index is J; g, constructing a scale set of the hydrogen adding station, wher