Multi-energy station cluster scheduling method, system, electronic terminal and computer readable storage medium

Abstract

The application discloses a kind of multi-energy hybrid charging station cluster scheduling method, the multi-energy hybrid charging station is hydrogen-electric hybrid charging station, the scheduling method includes the following steps: pre-analysis before scheduling, including photovoltaic uncertainty modeling and electric vehicle schedulable potential determination;Multi-energy hybrid charging station cluster is modeled, in the modeling process, hydrogen-electric hybrid charging station is upper layer of problem, and regional clearing management business is lower layer of problem, constructs Stackelberg game model, and power quotation and winning bid electric quantity are used as the interactive variables of upper and lower layers;Dual-level model is converted into single-level model using KKT condition;Single-level model is solved by programming using commercial solver, and scheduling result is obtained.The application provides an economic and efficient scheduling method, which can reduce the influence of new energy uncertainty, significantly improve the system energy utilization efficiency, and avoid a large amount of time consumption caused by multiple iterations of algorithm.

Description

Technical Field

[0001] This invention relates to the fields of integrated energy systems and electric vehicle technology, and in particular to a method, system, electronic terminal, and computer-readable storage medium for scheduling multi-energy hybrid charging station clusters. Background Technology

[0002] In recent years, the development of new energy vehicles such as electric vehicles and hydrogen-powered vehicles has played an irreplaceable role in achieving global decarbonization goals. The widespread adoption of new energy vehicles largely depends on the support of charging / energy-charging stations. There are two main approaches to modeling electric vehicle charging load: one is to propose a queuing theory model based on long-term operating data of electric vehicle charging stations, assuming that EV arrival follows a Poisson process; the other is to model the EV load based on actual road traffic conditions, comprehensively considering the interaction between EV users and road traffic. The first approach only requires determining relevant parameters to complete the modeling, but its accuracy is weak for EV users who are significantly affected by traffic congestion. The second approach has a more accurate model, but its ability to quickly predict the charging load of large-scale EV clusters is weak.

[0003] Solving the aforementioned technical problems is a research direction that those skilled in the art are dedicated to. Summary of the Invention

[0004] The first objective of this invention is to provide a multi-energy hybrid charging station cluster scheduling method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a multi-energy hybrid charging station cluster scheduling method, wherein the multi-energy hybrid charging station is a hydrogen-electric hybrid charging station, and the scheduling method includes the following steps: S1, conduct pre-scheduling analysis, including photovoltaic uncertainty modeling and determination of the schedulable potential of electric vehicles; S2. Problem modeling is performed on the multi-energy hybrid charging station cluster. In the modeling process, the hydrogen-electric hybrid charging station is the upper layer of the problem, and the regional clearing manager is the lower layer of the problem. A Stackelberg game model is constructed. The Stackelberg game model is a two-layer model, with the electricity price and the winning bid electricity volume as the interaction variables between the upper and lower layers. S3, using KKT conditions, the Stackelberg game model is transformed into a single-layer model; S4 uses a commercial solver to solve a single-layer model and obtains the scheduling results.

[0006] As a specific implementation method, the specific process of step S1 is as follows: S101. A stochastic optimization model for photovoltaic output in a hydrogen-electric hybrid charging station is established using a multi-scenario approach. The specific implementation process is as follows: First, define photovoltaic power generation as Where pv represents photovoltaic, t represents time, and ω represents a specific scene in the scene set, and To obtain a scene set with a time span T, based on the multivariate Gaussian distribution X... ω ~N(μ0, ∑) generates a random vector X = (X1, X2, ..., X... T ) T The vector with mean μ0 is a 1×T zero vector, and the covariance matrix ∑ is a symmetric T×T matrix whose diagonal elements are equal to 1, and whose off-diagonal elements are calculated by the following formula: (1), Wherein, the range parameter v controls the correlation strength of the random variable at different times, and t and t' represent two different times; In random vector X ω At time t of the ω-th scene, the inverse function φ(•) is applied to predict the cumulative distribution function. The normalized photovoltaic power generation curve for each scenario ω within the scheduling interval is obtained through the following transformation. for: (2), probability density function By Beta distribution Modeling, transforming the photovoltaic output curve μ t Normalize and treat the photovoltaic output curve as a conditional expectation describing the stochastic process of photovoltaic power generation; S102, Modeling the schedulable potential of electric vehicles, the implementation process is as follows: All slow-charging electric vehicles within the charging station are modeled as generalized energy storage devices using the Minkowski summation method, and used to participate in electricity market bidding. The established generalized energy storage device model is as follows: (4), in, and Let HCS represent the total charging power and discharging power of the generalized energy storage devices in the j-th HCS at time t. HCS stands for hydrogen-electric hybrid charging station. and These represent the maximum values ​​of charging and discharging of the generalized energy storage device in the j-th HCS at time t, respectively; Let represent the change in electricity consumption caused by the change in the grid-connected status of electric vehicles in the j-th HCS generalized energy storage device at time t. Let be the charge of the generalized energy storage device in the j-th HCS at time t. and Let Δt be the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS at time t, and Δt be the time interval of the optimization problem. and These are the charging and discharging efficiencies, respectively.

[0007] As a specific implementation method, in step S101, the number of scenes in the scene set is reduced through the following steps: First, use the improved version. k The -means clustering algorithm divides the original scene into different clusters, and then uses a synchronous back-substitution elimination algorithm based on Kantorovich probabilistic distance to reduce the scene set in each cluster to a unique typical scene. As shown below: (3) in and The original scene set and the retained scene set in cluster c; For the scene and The Euclidean distance; For the scene and The product of probabilities; For scene set Mid-scene The probability of; For scene set Mid-scene The probability of.

[0008] As one specific implementation method, k The improvement process of the -means clustering algorithm is as follows: 1) Initialize the number of population particles and the number of hypersphere centers, set the maximum number of iterations, the upper and lower limits of the hypersphere search radius, the probability of hypersphere motion angle change, and the initial number of cluster centers; 2) Construct and evaluate the fitness function F, calculate the fitness value of each individual, obtain the current fitness of the individual, and sort the entire population according to the fitness. 3) Individual position and fitness: Based on the distance relationship between the hypersphere center and the particle, guide the particle to move towards the current optimal center, redetermine the position and search radius of the hypersphere center, evaluate its fitness, and update the individual ranking until the set maximum number of iterations is reached or the early termination judgment condition is met. 4) Obtain the optimal initial cluster centers and perform clustering, using the optimal individual output by the hypersphere search algorithm as the cluster center. k- Clustering is performed using the initial cluster centers of -means, and the clustering results are continuously updated through iterative loops until the termination criteria are met or the maximum number of iterations is reached, and the final clustering results are output. 5) Calculate the average displacement. Calculate the displacement between the cluster center and all points in the cluster to obtain the average displacement.

[0009] As a specific implementation method, the specific process of step S2 is as follows: S201, Constructing the upper-level problem of the Stackelberg game model, the objective function of a single hydrogen-electric hybrid charging station is as follows: (5), Where t is the time index, T is the time set used for scheduling, Δt is the time interval of the optimization problem; HCS represents a hydrogen-electric hybrid charging station; Let RLMP represent the day-ahead clearing price of the j-th HCS at time t, and let RLMP represent the regional node marginal price. and Let be the total charging power and discharging power of the generalized energy storage device in the j-th HCS at time t. Let be the input power of the electrolytic cell unit in the j-th HCS at time t. Let be the photovoltaic output value of the j-th HCS at time t. The market value of the j-th HCS at time t. Let be the charge of the generalized energy storage device in the j-th HCS at time t. Let β be the amount of hydrogen stored in the j-th HCS hydrogen storage device at time t, and β be the degradation cost coefficient. The constraints for a single hydrogen-electric hybrid charging station are: (6); (7); (8); (9); (10); (11); (12); in, Let represent the change in electricity consumption caused by the change in the grid-connected status of electric vehicles in the j-th HCS generalized energy storage device at time t. and These are the charging and discharging efficiencies, respectively. To meet the energy needs of hydrogen fuel cell vehicles, The power-to-hydrogen conversion coefficient, This represents the efficiency value of the electrolytic cell unit. Energy dissipation rate of hydrogen storage equipment; and Let represent the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS at time t. and Let be the upper and lower limits of the hydrogen storage capacity of the hydrogen storage device in the j-th HCS at time t; and These are the energy values ​​that the generalized energy storage device and the hydrogen storage device should retain at the end of the scheduling in the j-th HCS, respectively. and Let these be the upper and lower limits of the j-th HCS quote; S202, the lower-level problem for constructing the Stackelberg game model, the objective function of the regional clearing manager is: (13) In the formula, N HCS This refers to the collection of hybrid charging stations contained in the HCS cluster. Let i be the power transmitted from the i-th HCS to the j-th HCS in the HCS cluster, and let i be the parent node of j and j be the child node of i. The constraints corresponding to the objective function are: (14); (15); (16); (17); (18); (19); (20); (twenty one); (twenty two); In equations (14) to (21), the right side of the colon represents the dual variables corresponding to the constraints of the original problem, K. j Let j be the set of child nodes of the j-th HCS. This represents the standard basic electric vehicle load and lacks dispatchable potential; K0 is the set of child nodes of the balancing node. The power of the child node of the balancing node, i.e., the j-th HCS, at time t; L represents the maximum power transferred from the i-th HCS to the j-th HCS. HCS This is the set of routes between the i-th HCS and the j-th HCS; This indicates the maximum capacity offered by the l-th pricing segment; and These represent the maximum values ​​of charging and discharging of the j-th HCS at time t, respectively; and These represent the maximum allowable values ​​for the electrolytic cell unit and photovoltaic power generation of the j-th HCS at time t, respectively.

[0010] As a specific implementation method, the specific process of step S3 is as follows: S301. Based on the upper and lower level problems of the Stackelberg game model constructed in step 2, the Lagrangian function of the lower level regional clearing manager problem is obtained, as shown in the following formula: (twenty three); S302, using the partial derivatives of the variables to obtain the KKT equilibrium conditions: (twenty four); S303, Write down the KKT directions and complementary constraint conditions corresponding to the inequality constraints: (25); S304 introduces Boolean variables and uses the Big-M method to transform complementary constraints into a form that is easy for the solver to solve: (26); In the formula, M is a constant, and all All are Boolean variables. for A set of vectors; S305, In summary, the KKT system representation of the lower-level problem is as follows: (27); S306, transformed into a linear term using strong duality theory, yields: (28); S307, further derived from the KKT equilibrium condition, yields: (29); S308, which leads to: (30); S309, Based on the formula obtained in step S308, substitute the corresponding terms of the KKT complementary constraint condition obtained in step 305, and combine them with... get: (31); S3010, At this point, the mathematical programming problem with equilibrium constraints corresponding to the original two-layer master-slave game has been transformed into a single-layer mixed-integer nonlinear programming problem. The final joint bidding optimization problem is in the form of: (32).

[0011] A second objective of this invention is to provide a multi-energy hybrid charging station cluster scheduling system, comprising: a regional clearing manager and a multi-energy hybrid charging station cluster, wherein, The multi-energy hybrid charging station cluster includes several hydrogen-electric hybrid charging stations. Each hydrogen-electric hybrid charging station includes photovoltaic panels, a charging station and a hydrogen refueling station. Each hydrogen refueling station includes an electrolyzer and hydrogen storage equipment for refueling hydrogen-powered vehicles. The regional clearing manager is responsible for calculations and energy management.

[0012] A third objective of this invention is to provide an electronic terminal, comprising: A memory on which computer programs or instructions are stored; A processor is used to load and execute the computer program or instructions to implement the above-described multi-energy hybrid charging station cluster scheduling method.

[0013] A fourth objective of this invention is to provide a computer-readable storage medium having a computer program or instructions stored thereon, wherein when the computer program or instructions are executed by a processor, the above-described multi-energy hybrid charging station cluster scheduling method is implemented.

[0014] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention proposes a multi-energy hybrid charging station cluster scheduling method and system, primarily consisting of a regional clearing manager and a hydrogen-electric hybrid charging station (HCS) cluster. It also considers the uncertainty of renewable energy output, providing an economical and efficient operating method for multi-energy hybrid charging station cluster scheduling. This method can mitigate the impact of new energy uncertainty, significantly improve system energy utilization efficiency, and avoid the significant time consumption caused by multiple algorithm iterations. The system achieves the minimization of electricity costs and the maximization of social welfare through two-layer game theory. The model considers the uncertainty of new energy output and transforms the two-layer model into a single-layer optimization problem solvable by a solver using KKT conditions. This improves the operating efficiency and stability of the multi-energy hybrid charging station cluster system, ensuring the system's economic and social benefits, and achieving a win-win situation for both individuals and society. Attached Figure Description

[0015] Figure 1 The flowchart illustrates the multi-energy hybrid charging station cluster scheduling method provided by this invention. Figure 2The Stackelberg game relationship diagram of the upper and lower layer optimization models in the multi-energy hybrid charging station cluster scheduling method provided by this invention; Figure 3 This is a schematic diagram of the structure of a multi-energy hybrid charging station cluster scheduling system provided by the present invention; Figure 4 This is a load curve diagram of the basic load of the fast-charging vehicle in the embodiment; Figure 5 This is a graph showing all the parameters of the slow-charging vehicle in the embodiment. Figure 6 This is a diagram showing the load and maximum photovoltaic power output of the hydrogen-powered vehicle in the embodiment; Figure 7 The following are the slow charging load scheduling results for the four power stations in the example; Figure 8 The hydrogen production rate of the electrolyzer unit in this embodiment is shown. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] like Figure 1 As shown, this invention discloses a multi-energy hybrid charging station cluster scheduling method, wherein the multi-energy hybrid charging station is a hydrogen-electric hybrid charging station, and the scheduling method includes the following steps: S1, conduct pre-scheduling analysis, including photovoltaic uncertainty modeling and determination of the schedulable potential of electric vehicles; S2. Problem modeling is performed on the multi-energy hybrid charging station cluster. In the modeling process, the hydrogen-electric hybrid charging station is the upper layer of the problem, and the regional clearing manager is the lower layer of the problem. A Stackelberg game model is constructed. The Stackelberg game model is a two-layer model, with the electricity price and the winning bid electricity volume as the interaction variables between the upper and lower layers. S3. To make the Stackelberg game model in step S2 easier to solve, the two-layer model is transformed into a single-layer model using KKT conditions. S4 uses a commercial solver to solve a single-layer model and obtains the scheduling results.

[0018] The multi-energy hybrid charging station cluster scheduling system here, such as Figure 3 As shown, it includes: regional clearing managers and multi-energy hybrid charging station clusters, among which, The multi-energy hybrid charging station cluster includes several hydrogen-electric hybrid charging stations. Each hydrogen-electric hybrid charging station includes photovoltaic panels, a charging station (for fast and slow charging) and a hydrogen refueling station. Each hydrogen refueling station includes an electrolyzer and hydrogen storage equipment for refueling hydrogen-powered vehicles. The regional clearing manager is responsible for computation and energy management. It uses commercial solvers to solve single-layer models and obtain scheduling results.

[0019] Here, the specific process of step S1 is as follows: S101. A stochastic optimization model for photovoltaic output in a hydrogen-electric hybrid charging station is established using a multi-scenario approach. The specific implementation process is as follows: First, define photovoltaic power generation as Where pv represents photovoltaic, t represents time, and ω represents a specific scene in the scene set, and To obtain a scene set with a time span T, based on the multivariate Gaussian distribution X... ω ~N(μ0, ∑) generates a random vector X = (X1, X2, ..., X... T ) T The vector with mean μ0 is a 1×T zero vector, and the covariance matrix ∑ is a symmetric T×T matrix whose diagonal elements are equal to 1, and whose off-diagonal elements are calculated by the following formula: (1) Wherein, the range parameter v controls the correlation strength of the random variable at different times, and t and t' represent two different times; In random vector X ω At time t of the ω-th scene, the inverse function φ(•) is applied to predict the cumulative distribution function. The normalized photovoltaic power generation curve for each scenario ω within the scheduling interval is obtained through the following transformation. for: (2), probability density function By Beta distribution Modeling, transforming the photovoltaic output curve μ t Normalize and treat the photovoltaic output curve as a conditional expectation describing the stochastic process of photovoltaic power generation; By applying this method, any number of photovoltaic power generation scenarios can be sampled around the expected curve, with all scenarios having equal probability. However, to correctly select the number of scenarios representing the uncertainty of photovoltaic power generation, this can be determined in the case analysis by the Wasserstein distance between each set of scenarios and its predecessor. Furthermore, computation using a large number of scenarios is time-consuming, therefore further scenario reduction is needed while ensuring computational accuracy.

[0020] Therefore, this example proposes an improved method. k A scenario reduction method combining the -means clustering algorithm and the synchronous back-substitution elimination algorithm. k The -means clustering algorithm divides the original scene into different clusters, and then uses a synchronous back-substitution elimination algorithm based on Kantorovich probabilistic distance to reduce the scene set in each cluster to a unique representative scene. Therefore, the reduced scene set can replace the original scene set.

[0021] Improvements k The -means clustering algorithm divides the original scene into different clusters. This differs from traditional... k -The means algorithm, an improved algorithm that integrates the hypersphere search algorithm to optimize the selection of initial cluster centers, has a fitness function based on the overall squared cluster error, specifically expressed as: , Where k represents the number of cluster categories, x i μ represents the i-th data object. k This represents the cluster center of the k-th cluster.

[0022] The specific steps to improve the clustering algorithm are as follows: Step 1: Initialize the number of population particles and the number of hypersphere centers, set the maximum number of iterations, the upper and lower limits of the hypersphere search radius, the probability of hypersphere motion angle change, and initialize the number of cluster centers; Step 2: Construct and evaluate the fitness function F, calculate the fitness value of each individual, obtain the current fitness of the individual, and sort the entire population according to the fitness. Step 3: Update individual positions and fitness. Based on the distance relationship between the hypersphere center and the particle, guide the particle to move towards the current optimal center, redetermine the position and search radius of the hypersphere center, evaluate its fitness, and update the individual ranking until the set maximum number of iterations is reached or the early termination judgment condition is met. Step 4: Obtain the optimal initial cluster centers and perform clustering, using the optimal individuals output by the hypersphere search algorithm as... k - Clustering is performed using the initial cluster centers of -means, and the clustering results are continuously updated through iterative loops until the termination criteria are met or the maximum number of iterations is reached, and the final clustering results are output. Step 5: Calculate the average displacement. Calculate the displacement between the cluster center and all points in the cluster to obtain the average displacement.

[0023] Then, a synchronous back-substitution elimination algorithm based on Kantorovich probabilistic distance is used to reduce the scene set in each cluster to a unique typical scene. Therefore, the reduced scene set can replace the original scene set. (Kantorovich probabilistic distance) As shown below: (3) in and The original scene set and the retained scene set in cluster c; For the scene and The Euclidean distance; For the scene and The product of probabilities; For scene set Mid-scene The probability of; For scene set Mid-scene The probability of.

[0024] It is important to note that improvements k The -means clustering algorithm does not consider scene probabilities, therefore a synchronous back-substitution elimination method is needed to reduce the number of scenes based on their probabilities. Furthermore, for scene reduction, the sum of the probabilities of all scenes in the scene set is 1. Considering that all scenes in the original scene set have the same probability, and that only one typical scene is retained in each cluster's scene set, the reduced scene probabilities can be modified as follows: pr c =N c / N Ω , Among them, pr c Let N be the probability of a typical scenario in cluster c; c To reduce the number of scenarios in a certain cluster c; N Ω The number of scenes in the original scene set Ω; By giving the number of scenarios and their corresponding probabilities, the mathematical expectation of the objective function containing photovoltaic power generation is calculated, which is the result of stochastic optimization.

[0025] S102, Modeling the schedulable potential of electric vehicles, the implementation process is as follows: This application does not model the individual travel characteristics of electric vehicles, but focuses on the schedulable charging and discharging behavior of slow-charging electric vehicles at the charging station level. Therefore, the charging station is regarded as a virtual aggregator. Combining the characteristics of electric vehicles in slow charging mode, which have load shifting and reverse power supply capabilities, all slow-charging electric vehicles in the hydrogen-electric hybrid charging station are modeled as generalized energy storage devices to participate in electricity market bidding using the Minkowski summation method.

[0026] Parameters of generalized energy storage devices This determines the potential of electric vehicle clusters as a flexible storage resource, namely, the dispatchable potential of electric vehicles. and These represent the maximum values ​​of charging and discharging of the generalized energy storage device in the j-th HCS at time t, respectively; Let represent the change in electricity consumption caused by the change in the grid-connected status of electric vehicles in the j-th HCS generalized energy storage device at time t. and These represent the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS during time period t.

[0027] The dispatchable potential is derived from historical data of electric vehicles at each charging station, including the arrival time, departure time, battery level of the electric vehicle upon arrival, battery level of the electric vehicle upon departure, and electric vehicle battery type (including battery capacity limits and maximum charging and discharging power).

[0028] If two variable spaces in the parameters of a generalized energy storage device have the same domain, then performing a Minkowski summation on the two variables yields the variables and their envelope, expressed as follows: , Here, A and B are two variable spaces; a and b are elements of A and B, respectively; A ⊕ B Let A and B be Minkowski sums.

[0029] Due to the differences in the grid connection time of individual electric vehicles, its domain Due to heterogeneity, it needs to be extended to the same scheduling time set T, resulting in the following generalized energy storage device model: (4).

[0030] in, and Let HCS represent the total charging power and discharging power of the generalized energy storage devices in the j-th HCS at time t. HCS stands for hydrogen-electric hybrid charging station. and These represent the maximum values ​​of charging and discharging of the generalized energy storage device in the j-th HCS at time t, respectively; Let represent the change in electricity consumption caused by the change in the grid-connected status of electric vehicles in the j-th HCS generalized energy storage device at time t. Let be the charge of the generalized energy storage device in the j-th HCS at time t. and Let Δt be the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS at time t, and Δt be the time interval of the optimization problem. and These are the charging and discharging efficiencies, respectively.

[0031] Step S2 involves modeling the multi-energy hybrid charging station cluster system, including constructing an upper-level problem and a lower-level problem in a game theory model. The game relationship between the upper and lower-level problems is as follows: Figure 2 As shown, the specific process is as follows: S201, a cluster consists of multiple hydrogen-electric hybrid charging stations (HCS), where the objective function for each HCS is to minimize the electricity cost. The objective function for a single hydrogen-electric hybrid charging station is shown below: (5) Where t is the time index, T is the time set used for scheduling, Δt is the time interval of the optimization problem; HCS represents a hydrogen-electric hybrid charging station; Let RLMP represent the day-ahead clearing price of the j-th HCS at time t, and let RLMP represent the regional node marginal price. and Let J represent the total charging power and discharging power of the generalized energy storage devices in the j-th HCS. Let be the input power of the electrolytic cell unit in the j-th HCS at time t. Let be the photovoltaic output value of the j-th HCS at time t. The market value of the j-th HCS at time t. and Let represent the electrical energy (kWh) of the generalized energy storage device in the j-th HCS at time t and the hydrogen storage capacity (m³) of the hydrogen storage device at time t, respectively. 3 β is the degradation cost coefficient. In this example, the value of β ranges from [0.005, 0.01]. The constraints for a single hydrogen-electric hybrid charging station are: (6); (7); (8); (9); (10); (11); (12); In equation (6), Let be the charge of the generalized energy storage device in the j-th HCS at time t. The change in electricity consumption caused by the change in the grid-connected status of electric vehicles at time t is represented by a generalized energy storage device. and These are the charging and discharging efficiencies, respectively; in equation (7), Let be the amount of hydrogen stored in the j-th HCS generalized hydrogen storage device at time t. To meet the energy needs of hydrogen fuel cell vehicles, The power-to-hydrogen conversion coefficient, This represents the efficiency value of the electrolytic cell unit. The energy dissipation rate of the hydrogen storage device; in equations (8) and (9), and Let represent the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS at time t. and Let be the upper and lower limits of the hydrogen storage capacity of the hydrogen storage device in the j-th HCS at time t; in equations (10) and (11), and Let be the energy values ​​that the generalized energy storage device and the hydrogen storage device should retain in the j-th HCS at the end of the scheduling process, respectively; in equation (12), and Let these be the upper and lower limits of the j-th HCS quote; Equations (5)-(12) are the bidding decision model for the j-th HCS. The decision variables have been indicated in equation (5). By adding the objective functions of all HCSs together, we obtain the joint bidding optimization problem of HCSs, which is also the upper-level optimization problem of the Stackelberg game.

[0032] S202, for the regional clearing manager, defines a lower-level optimization problem with the objective of maximizing social welfare. The objective function is: (13) In the formula, N HCS This refers to the collection of hybrid charging stations contained in the HCS cluster. Let i be the power transmitted from the i-th HCS to the j-th HCS in the HCS cluster, and let i be the parent node of j and j be the child node of i. The constraints corresponding to the objective function are: (14); (15); (16); (17); (18); (19); (20); (twenty one); (twenty two); In equations (14) to (21), the right side of the colon represents the dual variables corresponding to each constraint in the original problem. Equation (14) represents the balance constraint that the j-th HCS must satisfy. j Let j be the set of child nodes of the j-th HCS. The load is a conventional basic electric vehicle load (i.e., fast charging load), which has no scheduling potential; Equation (15) is the power balance constraint of the entire cluster. Since all HCSs in the region are aggregated together, their aggregation point is called the balance node, and K0 is the set of child nodes of the balance node. Let be the power of the child node of the balancing node, i.e., the j-th HCS, at time t; Equation (16) is the capacity constraint for allowed transmission between the i-th HCS and the j-th HCS. L represents the maximum power transferred from the i-th HCS to the j-th HCS. HCS Let be the set of routes between the i-th HCS and the j-th HCS; Equation (17) represents the segmented capacity constraint of the regional clearing manager. This represents the maximum capacity offered by the l-th bidding segment; Equations (18) and (19) are the charge and discharge constraints of the generalized energy storage device, respectively. and represent the maximum charging and discharging values ​​of the j-th HCS at time t, respectively; Equations (20) and (21) are the power constraints of the electrolytic cell unit and photovoltaic power generation, respectively. and These represent the maximum allowable values ​​of the electrolytic cell unit and photovoltaic power generation for the j-th HCS at time t, respectively; HCS settles electricity costs based on regional clearing tariffs, and regional clearing tariffs are... The relationship represented by equation (22) is satisfied.

[0033] Step S3 specifically includes the following steps: S301, based on the bi-level optimization problem composed of equations (5) to (12) and equations (13) to (22), the Lagrangian function of the lower-level regional clearing management problem is obtained, as shown in equation (23): (twenty three) S302, using the partial derivatives of the variables to obtain the KKT equilibrium conditions: (twenty four) S303, Next, write down the KKT directions and complementary constraint conditions corresponding to the inequality constraints: (25) S304 introduces Boolean variables and uses the Big-M method to transform complementary constraints into a form that is easy for the solver to solve: (26) In the formula, M is a constant, and all All are Boolean variables. for A set of vectors; S305, In summary, the KKT system representation of the lower-level problem is as follows: (27) S306, since the upper-level optimization problem contains bilinear terms and is difficult to solve, it is transformed into linear terms using strong duality theory. From strong duality theory, we get: (28) S307, further derived from the KKT equilibrium condition, yields: (29) S308, which leads to: (30) S309, based on equation (30), then substitute the corresponding terms in the KKT complementary constraint equation (25) and combine them with... get: (31) S3010, At this point, the mathematical programming problem with equilibrium constraints corresponding to the original two-layer master-slave game has been transformed into a single-layer mixed-integer nonlinear programming problem. The final joint bidding optimization problem is in the form of: (32) The aforementioned multi-energy hybrid charging station cluster scheduling system achieves the minimization of electricity costs and the maximization of social welfare through two-level game theory. The model considers the uncertainty of new energy output and transforms the two-level model into a single-level optimization problem that can be solved by a solver through KKT conditions. This improves the operating efficiency and stability of the multi-energy hybrid charging station cluster system, ensures the system's economic and social benefits, and achieves a win-win goal for both individuals and society.

[0034] Step S4 involves solving the single-layer model using a commercial solver through programming. Various commercial solvers are available; any solver capable of solving mixed-integer nonlinear programming problems will suffice. Alternatively, a custom-designed solution algorithm can be created. The solution outputs the scheduling results. In this example, either GUROBI or COPT is used as the solver. Example

[0035] This example uses the YALMIP toolbox and CPLEX 12.8 solver in MATLAB 2022a to verify the scheduling method of this invention. The scheduling cycle is one day, and the time interval is 15 minutes.

[0036] 1) Set parameters: This example uses four HCS, namely {HCS1, HCS2, HCS3, HCS4}. (The lower limit of the j-th HCS quote) and The upper limits of the j-th HCS price are $0.01 / kWh and $0.1 / kWh, respectively. (The energy value that the generalized energy storage device should retain at the end of the scheduling in the j-th HCS) is taken as {760kWh, 670kWh, 90kWh, 1150kWh}. The power-to-hydrogen conversion coefficient is 360 m. 3 / MWh, The efficiency of the electrolytic cell unit is 60%. (The maximum input power of the electrolytic cell unit in the j-th HCS at time t) is 75m. 3 , The maximum power transferred from the i-th HCS to the j-th HCS is 1000kW. The energy dissipation rate of the hydrogen storage device is 0.006%. (The energy value that the hydrogen storage device should retain at the end of the scheduling in the j-th HCS) is 0 m 3 Four HCS (The lower limit of the hydrogen storage capacity of the hydrogen storage device in the j-th HCS at time t) and (The upper limit of hydrogen storage capacity of the hydrogen storage device in the j-th HCS at time t) is 0 m 3 and 250 m 3 The parameters required for an HCS cluster are as follows: Figure 4 and Figure 5 As shown. Figure 4 The load curve for fast-charging vehicles (base load) is shown. Figure 5 In the figure, (a)-(d) show all the parameter curves for the slow-charging vehicles from HCS1 to HCS4, respectively. Because and They are numerically equal, therefore only the values ​​shown in the graph are equal. The load of hydrogen-powered vehicles and the maximum output of photovoltaic power are as follows: Figure 6 As shown.

[0037] 2) The Stackelberg game model constructed in this invention is used, and the two-layer model is transformed into a single-layer model using KKT conditions. Then, the single-layer model is solved using a commercial solver through programming to obtain the scheduling results. See [link to relevant documentation]. Figure 7 , 8 As shown.

[0038] Figure 7 Figures (a)-(d) show the slow-charging load scheduling results for the four power stations. We can see from the figures that the charging time of the four HCS vehicles is mainly concentrated during periods of lower regional electricity prices, and they discharge when necessary to meet the overall system optimization objective. For example, from 22:00 to 3:00, the regional electricity price is high, so none of the four HCSs engage in significant charging activity. Furthermore, since the regular load in this example is not very large, the discharge time of the four HCSs is not frequent. In addition, HCS4 exhibits simultaneous charging and discharging behavior, which is due to some vehicles charging while others discharge within the group. From this, we can see that the charging and discharging curves demonstrate the flexibility of the dispatchable load, proving that the method of this invention has good application prospects.

[0039] For an electrolyzer unit, the hydrogen gas produced by consuming electrical energy is as follows: Figure 8 As shown. Figure 8 (a)-(d) show the results for the four HCSs, respectively. The variations are mainly related to hydrogen load, regional electricity prices at different times, and constraints on hydrogen storage tanks. For example, between 12:00 and 15:00, due to lower regional electricity prices, the hydrogen load demand of the four HCSs was high, so all four HCSs purchased large amounts of electricity to convert into hydrogen for storage and consumption. Between 23:00 and 06:00, although the hydrogen load demand was high at certain times, the relative hydrogen storage was sufficient, so there was no large-scale electricity purchase for hydrogen production. After 04:00, since the hydrogen load demand was basically zero, there was virtually no electricity purchase.

[0040] The present invention also provides an electronic terminal, comprising: A memory on which computer programs or instructions are stored; A processor is used to load and execute the computer program or instructions to implement the above-described multi-energy hybrid charging station cluster scheduling method.

[0041] The present invention also provides a computer-readable storage medium having a computer program or instructions stored thereon, wherein when the computer program or instructions are executed by a processor, the above-described multi-energy hybrid charging station cluster scheduling method is implemented.

[0042] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0043] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0044] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0045] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0046] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for scheduling multi-energy hybrid charging station clusters, characterized in that, The multi-energy hybrid charging station is a hydrogen-electric hybrid charging station, and the scheduling method includes the following steps: S1, conduct pre-scheduling analysis, including photovoltaic uncertainty modeling and determination of the schedulable potential of electric vehicles; S2. Problem modeling is performed on the multi-energy hybrid charging station cluster. In the modeling process, the hydrogen-electric hybrid charging station is the upper layer of the problem, and the regional clearing manager is the lower layer of the problem. A Stackelberg game model is constructed. The Stackelberg game model is a two-layer model, with the electricity price and the winning bid electricity volume as the interaction variables between the upper and lower layers. S3, using KKT conditions, the Stackelberg game model is transformed into a single-layer model; S4 uses a commercial solver to solve a single-layer model and obtains the scheduling results.

2. The multi-energy hybrid charging station cluster scheduling method according to claim 1, characterized in that, The specific process of step S1 is as follows: S101. A stochastic optimization model for photovoltaic output in a hydrogen-electric hybrid charging station is established using a multi-scenario approach. The specific implementation process is as follows: First, define photovoltaic power generation as Where pv represents photovoltaic, t represents time, and ω represents a specific scene in the scene set, and To obtain a scene set with a time span T, based on the multivariate Gaussian distribution X... ω ~N(μ0, ∑) generates a random vector X = (X1, X2, ..., X... T ) T The vector with mean μ0 is a 1×T zero vector, and the covariance matrix ∑ is a symmetric T×T matrix whose diagonal elements are equal to 1, and whose off-diagonal elements are calculated by the following formula: （1）， Wherein, the range parameter v controls the correlation strength of the random variable at different times, and t and t' represent two different times; In random vector X ω At time t of the ω-th scene, the inverse function φ(•) is applied to predict the cumulative distribution function. The normalized photovoltaic power generation curve for each scenario ω within the scheduling interval is obtained through the following transformation. for: （2）， probability density function By Beta distribution Modeling, transforming the photovoltaic output curve μ t Normalize and treat the photovoltaic output curve as a conditional expectation describing the stochastic process of photovoltaic power generation; S102, Modeling the schedulable potential of electric vehicles, the implementation process is as follows: All slow-charging electric vehicles within the charging station are modeled as generalized energy storage devices using the Minkowski summation method, and used to participate in electricity market bidding. The established generalized energy storage device model is as follows: （4）， in, and Let HCS represent the total charging power and discharging power of the generalized energy storage devices in the j-th HCS at time t. HCS stands for hydrogen-electric hybrid charging station. and These represent the maximum values ​​of charging and discharging of the generalized energy storage device in the j-th HCS at time t, respectively; Let represent the change in electricity consumption caused by the change in the grid-connected status of electric vehicles in the j-th HCS generalized energy storage device at time t. Let be the charge of the generalized energy storage device in the j-th HCS at time t. and Let Δt be the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS at time t, and Δt be the time interval of the optimization problem. and These are the charging and discharging efficiencies, respectively.

3. The method for scheduling a multi-energy hybrid charging station cluster according to claim 2, characterized in that, In step S101, the number of scenes in the scene set is reduced through the following steps: First, use the improved version. k The -means clustering algorithm divides the original scene into different clusters, and then uses a synchronous back-substitution elimination algorithm based on Kantorovich probabilistic distance to reduce the scene set in each cluster to a unique typical scene. As shown below: （3） in and The original scene set and the retained scene set in cluster c; For the scene and The Euclidean distance; For the scene and The product of probabilities; For scene set Mid-scene The probability of; For scene set Mid-scene The probability of.

4. The multi-energy hybrid charging station cluster scheduling method according to claim 3, characterized in that, k The improvement process of the -means clustering algorithm is as follows: 1) Initialize the number of population particles and the number of hypersphere centers, set the maximum number of iterations, the upper and lower limits of the hypersphere search radius, the probability of hypersphere motion angle change, and the initial number of cluster centers; 2) Construct and evaluate the fitness function F, calculate the fitness value of each individual, obtain the current fitness of the individual, and sort the entire population according to the fitness. 3) Individual position and fitness: Based on the distance relationship between the hypersphere center and the particle, guide the particle to move towards the current optimal center, redetermine the position and search radius of the hypersphere center, evaluate its fitness, and update the individual ranking until the set maximum number of iterations is reached or the early termination judgment condition is met. 4) Obtain the optimal initial cluster centers and perform clustering, using the optimal individual output by the hypersphere search algorithm as the cluster center. k - Clustering is performed using the initial cluster centers of -means, and the clustering results are continuously updated through iterative loops until the termination criteria are met or the maximum number of iterations is reached, and the final clustering results are output. 5) Calculate the average displacement. Calculate the displacement between the cluster center and all points in the cluster to obtain the average displacement.

5. The multi-energy hybrid charging station cluster scheduling method according to claim 1, characterized in that, The specific process of step S2 is as follows: S201, Constructing the upper-level problem of the Stackelberg game model, the objective function of a single hydrogen-electric hybrid charging station is as follows: （5）， Where t is the time index, T is the time set used for scheduling, Δt is the time interval of the optimization problem; HCS represents a hydrogen-electric hybrid charging station; Let RLMP represent the day-ahead clearing price of the j-th HCS at time t, and let RLMP represent the regional node marginal price. and Let be the total charging power and discharging power of the generalized energy storage device in the j-th HCS at time t. Let be the input power of the electrolytic cell unit in the j-th HCS at time t. Let be the photovoltaic output value of the j-th HCS at time t. The market value of the j-th HCS at time t. Let be the charge of the generalized energy storage device in the j-th HCS at time t. Let β be the amount of hydrogen stored in the j-th HCS hydrogen storage device at time t, and β be the degradation cost coefficient. The constraints for a single hydrogen-electric hybrid charging station are: （6）； （7）； （8）； （9）； （10）； （11）； （12）； in, Let represent the change in electricity consumption caused by the change in the grid-connected status of electric vehicles in the j-th HCS generalized energy storage device at time t. and These are the charging and discharging efficiencies, respectively. To meet the energy needs of hydrogen fuel cell vehicles, The power-to-hydrogen conversion coefficient, This represents the efficiency value of the electrolytic cell unit. Energy dissipation rate of hydrogen storage equipment; and Let represent the upper and lower limits of the energy of the generalized energy storage device in the j-th HCS at time t. and Let be the upper and lower limits of the hydrogen storage capacity of the hydrogen storage device in the j-th HCS at time t; and These are the energy values ​​that the generalized energy storage device and the hydrogen storage device should retain at the end of the scheduling in the j-th HCS, respectively. and Let these be the upper and lower limits of the j-th HCS quote; S202, the lower-level problem for constructing the Stackelberg game model, the objective function of the regional clearing manager is: （13）， In the formula, N HCS This refers to the collection of hybrid charging stations contained in the HCS cluster. Let i be the power transmitted from the i-th HCS to the j-th HCS in the HCS cluster, and let i be the parent node of j and j be the child node of i. The constraints corresponding to the objective function are: （14）； （15）； （16）； （17）； （18）； （19）； （20）； （21）； （22）； In equations (14) to (21), the right side of the colon represents the dual variables corresponding to the constraints of the original problem, K. j Let j be the set of child nodes of the j-th HCS. This represents the standard basic electric vehicle load and lacks dispatchable potential; K0 is the set of child nodes of the balancing node. The power of the child node of the balancing node, i.e., the j-th HCS, at time t; L represents the maximum power transferred from the i-th HCS to the j-th HCS. HCS This is the set of routes between the i-th HCS and the j-th HCS; This indicates the maximum capacity offered by the l-th pricing segment; and These represent the maximum values ​​of charging and discharging of the j-th HCS at time t, respectively; and These represent the maximum allowable values ​​for the electrolytic cell unit and photovoltaic power generation of the j-th HCS at time t, respectively.

6. The multi-energy hybrid charging station cluster scheduling method according to claim 5, characterized in that, The specific process of step S3 is as follows: S301, Based on the upper and lower level problems of the Stackelberg game model constructed in step 2, the Lagrangian function of the lower level regional clearing manager problem is obtained, as shown in the following formula: （23）； S302, using the partial derivatives of the variables to obtain the KKT equilibrium conditions: （24）； S303, Write down the KKT directions and complementary constraint conditions corresponding to the inequality constraints: （25）； S304 introduces Boolean variables and uses the Big-M method to transform complementary constraints into a form that is easy for the solver to solve: （26）； In the formula, M is a constant, and all All are Boolean variables. for A set of vectors; S305, In summary, the KKT system representation of the lower-level problem is as follows: （27）； S306, transformed into a linear term using strong duality theory, yields: （28）； S307, further derived from the KKT equilibrium condition, yields: （29）； S308, which leads to: （30）； S309, Based on the formula obtained in step S308, substitute the corresponding terms of the KKT complementary constraint condition obtained in step 305, and combine them with... get: （31）； S3010, At this point, the mathematical programming problem with equilibrium constraints corresponding to the original two-layer master-slave game has been transformed into a single-layer mixed-integer nonlinear programming problem. The final joint bidding optimization problem is in the form of: （32）。 7. A multi-energy hybrid charging station cluster scheduling system, used to execute the scheduling method according to any one of claims 1 to 6, characterized in that, include: Regional clearing management companies and multi-energy hybrid charging station clusters, among which, The multi-energy hybrid charging station cluster includes several hydrogen-electric hybrid charging stations. Each hydrogen-electric hybrid charging station includes photovoltaic panels, a charging station and a hydrogen refueling station. Each hydrogen refueling station includes an electrolyzer and hydrogen storage equipment for refueling hydrogen-powered vehicles. The regional clearing manager is responsible for calculations and energy management.

8. An electronic terminal, characterized in that, include: A memory on which computer programs or instructions are stored; A processor is configured to load and execute the computer program or instructions to implement the multi-energy hybrid charging station cluster scheduling method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by the processor, they implement the multi-energy hybrid charging station cluster scheduling method as described in any one of claims 1 to 6.

Images