Dynamic layout planning method for electric vehicle charging and battery swapping station based on double-layer model

Abstract

The invention discloses a double-layer model-based dynamic layout planning method for an electric vehicle charging and swapping station, and relates to the field of site layout of the electric vehicle charging and swapping station. The invention aims to solve the problems that the electric vehicle charging station is low in coverage rate and long in charging time, and the existing charging mode is single and the site selection of the power station is not planned. The invention relates to a dynamic layout planning method for an electric vehicle charging and swapping station based on a double-layer model. The method comprises the following steps: firstly, predicting a future electricity demand of a target area according to electric vehicle demand data of the target area; and then, establishing an upper-layer user selection model and a lower-layer power station constant volume model, respectively solving a minimum value of the upper-layer user selection model and a maximum value of the lower-layer power station constant volume model by utilizing a genetic algorithm, obtaining an optimal power station coordinate, capacity and power utilization price, and further carrying out dynamic layout planning on the electric vehicle charging and swapping power station.

Description

Technical Field

[0001] The invention belongs to the field of site layout of charging and swapping stations for electric vehicles. Background Art

[0002] As global attention to environmental issues increases, the drawbacks of fossil fuels have become increasingly apparent, and their use has led to serious pollution, resource depletion and climate change. As an alternative, electric vehicles do not rely on fossil fuels and reduce harmful emissions and greenhouse gases. Especially when electricity comes from renewable energy, electric vehicles can significantly reduce carbon footprints, improve energy efficiency, and promote environmental protection and sustainable development. However, with the development of the electric vehicle industry and the widespread popularity of electric vehicles, the problem of charging electric vehicles has become more and more serious. The massive increase in charging infrastructure is far behind the growth rate of electric vehicles. Secondly, because electric vehicles have a relatively short range and require frequent charging, coupled with insufficient charging infrastructure, low coverage of charging stations and long charging time, the problem of charging electric vehicles needs to be solved urgently.

[0003] How to reasonably plan and layout charging stations and battery swap stations to meet the electricity needs of different groups of people has become a top priority. In previous studies, such as the application number 202111066707.9 disclosed in the Chinese patent literature, the application publication date 2021.09.10, the invention name "Electric vehicle charging station layout method and electronic equipment based on a two-layer planning model", this method calculates the 24-hour electricity demand distribution of each charging station based on the preference model. The lower model takes the minimum construction cost and the maximum charging station profit as the objective function, solves the model based on the particle swarm optimization algorithm, and outputs the layout plan of the electric vehicle charging station. Another example is the application number 201811408046.1 disclosed in the Chinese patent literature, the application publication date is 2018.11.23, and the invention name is "Electric Vehicle Charging Station Planning and Layout Method and System Based on Hierarchical Analysis". This method first uses the hierarchical analysis method to comprehensively consider various factors affecting the location of charging stations, such as road traffic, land planning, and distribution network acceptance capacity, to screen out candidate addresses, and then construct an electric vehicle planning model. Based on the electric vehicle planning model, the capacity constraints and grid constraints of the charging station address are calculated, and the layout plan of the vehicle charging station is obtained based on the capacity constraints and grid constraints. However, the above method only considers a single charging method, ignores the needs of users of other charging methods, and does not dynamically plan the location of charging stations according to changes in electricity demand. Summary of the invention

[0004] The present invention aims to solve the problems of low coverage and long charging time of electric vehicle charging stations, single existing charging method and no planning of power station site selection. A dynamic layout planning method of electric vehicle charging and swapping stations based on a double-layer model is now provided.

[0005] A dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model, comprising:

[0006] Predict the future electricity demand in the target area based on the electric vehicle demand data in the target area;

[0007] Establish an upper user selection model, the upper user selection model includes: charging selection model and battery swap selection model And the expressions are:

[0008]

[0009]

[0010] The upper-level user selection model constraints are as follows:

[0011] θ1, θ2, θ3∈{0,1}, θ1+θ2+θ3=1, ∑hij=1, d il ≤D max ,

[0012] Among them, T ij ', AF ij ' m , d i ' l , T il ′ respectively represent the ij , AF ijm , d il , T il The normalized result, T ij and T il are the time taken for users at electricity demand point i to travel to candidate charging station j and candidate swap station l, respectively. ijm is the attractiveness of candidate charging station j to m types of charging users in electricity demand point i, is the expected waiting time for charging users of type m in candidate charging station j, d il is the distance from the electricity demand point i to the candidate battery swap station l, is the battery replacement cost of candidate battery replacement station l, h ij is the user decision variable, θ1, θ2 and θ3 are constraint weights, D max is the maximum driving distance of the electric vehicle, i∈N, j∈M, l∈L, N is the total number of electricity demand points, M is the total number of candidate charging stations, and L is the total number of candidate battery swap stations;

[0013] A lower-layer power station capacity model is established, and the expression of the lower-layer power station capacity model f(·) is:

[0014] f(S jmc ,S lm ,p t )=P to -FC EV -C T ,

[0015] The constraints of the lower power station capacity model are as follows:

[0016]

[0017] Among them, S jmc is the number of type m charging piles in charging station j, S lm is the capacity of the battery swap station l, p t is the average charging electricity price in the tth year, P to is the power station revenue, F is the power station construction and operation cost, C EV is the carbon emission cost of the power station, C T Travel costs for users;

[0018] Genetic algorithms are used to solve the minimum value of the upper user selection model and the maximum value of the lower power station capacity model, respectively, to obtain the optimal power station coordinates, capacity and electricity price, and then dynamically plan the layout of electric vehicle charging and swapping stations.

[0019] Furthermore, the above-mentioned prediction of the future electricity demand of the target area based on the electric vehicle demand data of the target area includes:

[0020] Collecting data on the number of electric vehicles in the target area and its ancillary information, permanent population, per capita income, charging prices and oil prices. The ancillary information includes: quarterly sales volume of electric vehicles, market share and number of charging piles;

[0021] The collected data is input into the support vector machine prediction model for training, and the trained support vector machine prediction model is used to predict the future electricity demand of the target area.

[0022] Furthermore, the above-mentioned establishment of the upper user selection model includes:

[0023] Collect the coordinates of public parking lots, residential areas and commercial areas in the target area respectively;

[0024] Perform cluster analysis on the coordinates of public parking lots, and select the coordinate points of public parking lots closest to each cluster center as candidate power stations;

[0025] Perform cluster analysis on the coordinates of residential and commercial areas, and use the coordinate points of residential or commercial areas closest to each cluster center as electricity demand points;

[0026] Divide the future electricity demand of the target area into three modes: fast charging, slow charging and battery replacement, and allocate different electricity demands to each electricity demand point, so that the ratio of fast charging demand, slow charging demand and battery replacement demand in each electricity demand point is 16:4:5;

[0027] Calculate the driving time, driving distance, electricity price and electricity queuing time from each electricity demand point to each candidate power station respectively, the electricity price includes charging price and battery swap price, the electricity queuing time includes charging queuing time and battery swap queuing time, and the candidate power stations include candidate charging stations and candidate battery swap stations;

[0028] An upper-level user selection model is constructed using the travel time, travel distance, electricity price, and electricity queuing time.

[0029] Furthermore, the time T for the user at the power demand point i to go to the candidate charging station j is ij The expression is:

[0030] T ij =t0[1+α(qc) β ],

[0031] Among them, t0 is the free flow travel time from the electricity demand point i to the candidate charging station j, α and β are regression coefficients, c is the actual traffic capacity from the electricity demand point i to the candidate charging station j, and q is the traffic volume;

[0032] The time T for the user at the electricity demand point i to go to the candidate battery swap station l il The expression is:

[0033] T il =t0′[1+α(qc′) β ],

[0034] Among them, t0′ is the free flow travel time within the section from the power demand point i to the candidate battery swap station l, and c′ is the actual traffic capacity within the section from the power demand point i to the candidate battery swap station l;

[0035] The attractiveness AF of the candidate charging station j to the m types of charging users in the electricity demand point i ijm The expression is:

[0036]

[0037] Among them, AOS jm is the service capacity of candidate charging station j for electric vehicles with charging type m, Pr j is the average charging price of candidate charging station j, d ij is the distance from the electricity demand point i to the candidate charging station j, d ij ≤Dmax ;

[0038] The expected waiting time W of the m-type charging users in the candidate charging station j is m j The expression is:

[0039]

[0040] Among them, ρ m is the service intensity of type m charging pile, P m0 is the probability that all m types of charging piles are idle, λ jm The probability of an electric vehicle of type m charging arriving at candidate charging station j;

[0041] The battery replacement cost of the candidate battery replacement station l The expression is:

[0042]

[0043] Among them, C fc is the fixed cost, P bc For charging price.

[0044] Furthermore, the service capability AOS of the candidate charging station j for electric vehicles with charging type m is jm The expression is:

[0045] AOS jm =S jmc P m ,

[0046] Among them, P m is the rated power of the Class M charging pile;

[0047] The service intensity ρ of the m-type charging pile m The expression is:

[0048]

[0049] Among them, μ m Charging time for electric vehicles of type M charging;

[0050] The probability P that all the m types of charging piles are idle m0 The expression is:

[0051]

[0052] Where τ=0,1,...,S jmc -1;

[0053] The fixed cost C fc The expression is:

[0054] C fc =P BR +P ED +P LC ,

[0055] P BR is the battery replacement cost, P ED is the equipment depreciation cost, P LC For labor costs.

[0056] Furthermore, the above-mentioned establishment of the lower power station capacity determination model includes:

[0057] Calculate the power station construction and operation cost F and user travel cost C respectively T , power plant carbon emission cost C EV and power station revenue P to , and establish a fixed capacity model for the lower power station.

[0058] Furthermore, the above power station construction and operation cost F includes the construction cost F1 of the charging station and the construction cost F2 of the battery swap station, and the expressions are:

[0059]

[0060]

[0061] Among them, r is the depreciation rate, t is the operating life of the power station, G mc is the unit price of type m charging pile, F ck is the construction and operation cost of a charging station with k charging devices, P p_m F is the cost of purchasing and storing batteries. ek is the construction cost of the battery swap station, F rk is the operating cost of the battery swap station, K represents the total number of charging equipment, and h jk and h lk are the decision variables for charging and battery swapping equipment respectively;

[0062] The power station income P to Including charging station revenue P to1 and the revenue of the battery swap station P to2 , and the expressions are:

[0063]

[0064]

[0065] Among them, Pr j is the average charging price of candidate charging station j, p w is the grid electricity price, e ciis the average power consumption of electric vehicles per charge at the power demand point i, γ is the power consumption of electric vehicles per kilometer, and ε c and ε s are the average number of times each electric vehicle is charged and replaced per day, and are the proportions of charging users and battery swapping users in electricity demand points, a i is the electricity demand at electricity demand point i, P bc For charging price;

[0066] The user travel cost C T The expression is:

[0067] C T =C wT +C vT +C eT ,

[0068] Among them, C wT is the waiting time cost at the station, C vT is the time cost in the process, C eT The time cost of battery replacement;

[0069] The power station carbon emission cost C EV The expression is:

[0070]

[0071] Where μ is the emission factor of the electricity generated by the grid, d ij is the distance from the electricity demand point i to the candidate charging station j and d ij ≤D max , p D is the carbon dioxide treatment fee, X ij and X il are the decision variables for charging and battery replacement time, respectively;

[0072] X ij ,X il ,h jk ,h lk ∈{0,1},Xij≤hjk,X il ≤h lk ,

[0073] n E and n L are the number of candidate charging stations and the number of candidate battery swap stations, respectively, Jmax and S Jmin are the maximum and minimum capacity of the candidate charging stations, S Lmax and S Lminare the maximum and minimum capacities of the candidate battery swap stations respectively.

[0074] Furthermore, the waiting time cost C wT 、Time cost in the process C vT and the battery replacement time cost C eT The expressions are:

[0075]

[0076]

[0077]

[0078] Among them, δ is the urban travel time cost coefficient, and are the expected waiting time for fast-charging and slow-charging users in candidate charging station j, p1 is the probability of electric vehicles using fast charging, P G The unit time cost of the user.

[0079] Furthermore, the transportation conditions of the above-mentioned lower power station capacity model include:

[0080] d jj′ ≥d min , j≠j′,

[0081]

[0082]

[0083]

[0084] d jj′ is the distance from candidate charging station j to candidate charging station j′, d min is the minimum proximity distance between candidate charging stations, Pr j is the average charging electricity price of candidate charging station j, and Pr They are respectively the upper and lower limits of charging prices set by market regulators.

[0085] Furthermore, the charging equipment decision variable h jk The number of rows is the number of candidate charging station addresses in the candidate power station set, and the number of columns is the level of the candidate charging station. The battery swapping equipment decision variable h lk The number of rows is the number of candidate battery swap station addresses in the candidate station set, and the number of columns is the level of the candidate battery swap station.

[0086] The beneficial effects of the present invention are:

[0087] The present invention comprehensively considers the interests of electric vehicle power station investors and operators and users, and considers the charging experience of users with different electricity demands.

[0088] The present invention adopts an annual dynamic programming method to optimize and adjust the layout of electric vehicle charging and swapping stations according to market demand. In each year, the distribution and scale of power stations are dynamically adjusted taking into account the behavior patterns of charging users and changes in electricity demand, thereby maximizing the profits of power station operators while ensuring the convenience of charging for users.

[0089] The present invention takes into account the needs of different charging groups, comprehensively considers the three needs of fast charging, slow charging, and battery replacement, and takes into account the interests of both charging groups and power station operators. It analyzes and dynamically plans the site and capacity of power stations year by year according to the growth of electricity demand. It maximizes the interests of power station operators while maximizing the charging experience of users. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 It is a flow chart of a dynamic layout planning method of electric vehicle charging and swapping stations based on a double-layer model;

[0091] Figure 2 A specific flow chart planned for a particular year;

[0092] Figure 3 It is the growth graph of electricity demand;

[0093] Figure 4 Candidate addresses and electricity demand map for the target area;

[0094] Figure 5 Find the target path addressing diagram for a certain demand point;

[0095] Figure 6 The fitness function curve diagram of the best choice of battery swap station for users of the upper model;

[0096] Figure 7 The fitness function curve diagram of the best fast charging station for users of the upper model;

[0097] Figure 8 The fitness function curve diagram of the best slow charging station for users of the upper model;

[0098] Fig. 9 This is the fitness function curve of the lower model in the first year;

[0099] Fig.10 This is the fitness function curve of the lower model in the second year;

[0100] Fig.11 This is the fitness function curve of the lower model in the third year. DETAILED DESCRIPTION

[0101] The following will be combined with the accompanying 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 only part of the embodiments of the present invention, rather than 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 belong to the scope of protection of the present invention. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0102] Reference Figures 1 to 11 Specifically describing this embodiment, a dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model described in this embodiment includes:

[0103] S1: Predict the future electricity demand in the target area based on the electric vehicle demand data in the target area.

[0104] S11: Obtain the data on the number of electric vehicles in the target area and its ancillary information, including: quarterly sales of electric vehicles, market share, number of charging piles, etc.; obtain the permanent population and per capita income in the target area; obtain local charging prices, oil prices, etc.

[0105] S12: The data obtained in step S11 is preprocessed (abnormal points are removed and missing values ​​are supplemented) and then input into the support vector machine (SVM) prediction model for training. The trained SVM model is used to predict the electricity demand (the number of electric vehicles in the future) for the next three years. The prediction results are as follows: Figure 3 shown.

[0106] S2: Establish upper-level user selection model.

[0107] S21: Perform cluster analysis on the data.

[0108] The coordinate information of public parking lots, residential areas and commercial areas is obtained through data collection.

[0109] like Figure 4As shown in the figure, the coordinates of the public parking lot are clustered according to different electricity demands to obtain several cluster centers. The cluster center can be found through an iterative algorithm, and the condition that the sum of the distances from all data points to the cluster center is the smallest is required. The coordinates of the cluster center and the public parking lot (Point of Information, POI) are imported into the electronic map of QGIS (Quantum Geographic Information System), and the coordinate points of the public parking lot closest to each cluster center are taken as candidate points of the public parking lot, that is, candidate power stations.

[0110] Similarly, for residential and commercial areas, cluster analysis is performed on the coordinates of residential and commercial areas, and the coordinate points of the residential or commercial areas closest to each cluster center are used to obtain the electricity demand points.

[0111] The future electricity demand of the target area obtained by S1 is divided into three types: fast charging, slow charging and battery replacement. Different electricity demands are allocated to each electricity demand point, so that the ratio of users with fast charging demand, slow charging demand and battery replacement demand in each electricity demand point is 16:4:5.

[0112] S22: Calculate the driving time, driving distance, electricity price and electricity queuing time from each electricity demand point to each candidate power station respectively, the electricity price includes the charging price and the battery replacement price, the electricity queuing time includes the charging queuing time and the battery replacement queuing time, and the candidate power stations include candidate charging stations and candidate battery replacement stations.

[0113] S221: The time T of the charging user from the power demand point i to the candidate charging station j ij Affected by factors such as traffic volume and urban road resistance, it can be calculated by the following formula:

[0114] T ij =t0[1+α(qc) β ] (1),

[0115] Where t0 is the free flow travel time from the electricity demand point i to the candidate charging station j, α and β are regression coefficients (the typical value of α is 0.15, and the typical value of β is 4), c is the actual traffic capacity from the electricity demand point i to the candidate charging station j, and q is the traffic volume.

[0116] The time T for the battery swap user to travel from the power demand point i to the candidate battery swap station l il The expression is:

[0117] T il =t0′[1+α(qc′) β ],

[0118] Among them, t0′ is the free flow travel time within the section from the electricity demand point i to the candidate battery swap station l, and c′ is the actual traffic capacity within the section from the electricity demand point i to the candidate battery swap station l.

[0119] S222: Calculate driving distance using QGIS.

[0120] QGIS is an open source geographic information system software that provides a wealth of spatial analysis functions, one of which is network analysis. Network analysis is one of the important functions in QGIS, used for path planning, shortest path analysis, etc. First, obtain the national road network Shapefile data and import it into the layer, import the longitude and latitude coordinates of the demand point and the coordinates of the candidate address into QGIS through point features, and then use the path planning function in QGIS network analysis to calculate the shortest path and travel distance from the demand point to the candidate address. The results are as follows: Figure 5 shown.

[0121] S223: Combine the charging price with the charging distance to construct a charging station attraction function, and use this function to reflect the attractiveness of charging station j to the user at the electricity demand point i. The calculation formula is as follows:

[0122]

[0123] AOS jm =S jmc P m (3)

[0124] Where AF ijm is the attractiveness of candidate charging station j to type m charging users at electricity demand point i, AOS jm is the service capacity of candidate charging station j for electric vehicles with charging type m, S jmc is the number of type m charging piles in charging station j, P m is the rated power of the m-type charging pile, Pr j is the average charging electricity price (charging price) of candidate charging station j, d ij is the distance from electricity demand point i to candidate charging station j.

[0125] For battery replacement users, the charging standards for electric vehicle battery replacement services vary according to the different needs of users and the type of battery, including battery replacement, equipment depreciation, labor and other costs. The user battery replacement fee calculation formula is as follows:

[0126]

[0127] C fc =P BR +P ED +P LC (5)

[0128] Where: is the battery replacement cost of candidate battery replacement station l; C fc is a fixed cost, which is determined by the battery replacement cost P BR , Equipment depreciation cost P ED and labor cost P LC Composition; P bc It is the charging price, which is related to the battery type and electricity price.

[0129] S224: Charging / swapping queue time. The fast or slow charging of electric vehicles at charging stations is attributed to the M / M / c model of queuing theory, taking into account the limitation of vehicle queue waiting time. When the number of vehicles is constant, the queue waiting time decreases as the number of charging piles increases, and the charging piles should be configured under the specified queue waiting time expectation.

[0130]

[0131] Where: is the expected waiting time for m types of charging users in candidate charging station j; ρ m is the service intensity of type m charging pile; P m0 is the probability that all m types of charging piles are idle; jm The probability that an electric vehicle of type m charging type arrives at candidate charging station j; μ m is the charging time of an electric vehicle of type m; τ=0,1,...,S jmc -1.

[0132] The calculation method of travel time and driving distance of battery swap users is the same as that of users who use fast and slow batteries. Considering the charging time of users who use battery swap, since the battery swap time of electric vehicles is short, generally 5 to 10 minutes, the waiting time for battery swap can be ignored. Therefore, the total time for battery swapping to the battery swap station is the time T from the user at the power demand point i to the candidate battery swap station l. il .

[0133] S23: Based on the user selection model, the user selection function is calculated, and the user's best selected power station site is determined according to the selection function.

[0134] Construct an upper-level user selection model. Since we need to consider users with different charging methods, we set two upper-level user selection models, namely user charging selection model Battery swap selection model with users

[0135]

[0136] Where: T ij ', AF ij 'm , d i ' l , T il ′ respectively represent the ij , AF ijm , d il , T il The normalized result, T il is the time it takes for the user at the electricity demand point i to go to the candidate battery swap station l, h ij is the user decision variable, which determines which charging / swapping station the user wants to go to, il is the distance from the electricity demand point i to the candidate battery swap station l.

[0137] The objective function is:

[0138]

[0139] Objective function value The corresponding power station serial number is the desired candidate power station set.

[0140] The constraints of the upper model are as follows:

[0141] User weight constraints:

[0142] θ1, θ2, θ3∈{0,1}, θ1+θ2+θ3=1 (13).

[0143] Decision variable constraints:

[0144] ∑h ij =1 (14).

[0145] Maximum driving distance constraints for electric vehicles:

[0146] d ij ≤D max ,i∈N,j∈M(15),

[0147] d il ≤D max ,i∈N,l∈L(16),

[0148] N is the total number of electricity demand points, M is the total number of charging stations, L is the total number of battery swap stations, and D max The maximum driving distance of an electric vehicle.

[0149] The decision variable constraint indicates that users at each demand point can only go to the same charging station to receive charging services during the same time period.

[0150] The results of solving the upper model are as follows Figure 6,7,8, we get the best selection of fast charging station set, slow charging station set and battery swap station set for users. The results of the lower model are as follows Fig. 9 ,10,11,The site addresses and objective function values ​​of the three years are shown in Table 1.

[0151] Table 13 Site addresses and objective function values ​​in 2013

[0152]

[0153] S3: Based on the user's best choice of power station site set calculated by the user selection model, the lower-level charging and swapping station capacity model is established. The results of the upper-level model are used as the input of the lower-level model, and the interests of electric vehicle power station investors and users are comprehensively considered to maximize the power station revenue, minimize the power station construction and operation costs, minimize the power station carbon emission costs, and minimize the user travel costs. Build a power station layout optimization model.

[0154] S31: Comprehensively consider the interests of electric vehicle power station investors, operators and users, that is, maximize the power station revenue, minimize the power station construction and operation costs, minimize the power station carbon emission costs, and minimize user travel costs.

[0155] (a) The power station construction and operation cost F consists of two parts: charging station and battery swap station. The formula is as follows:

[0156] F=F1+F2.

[0157] F1 is the construction cost of the charging station, which mainly includes infrastructure cost and land acquisition cost. Infrastructure cost mainly consists of charger, distribution transformer, other auxiliary equipment, auxiliary road construction cost, etc.; land acquisition cost depends on the geographical location and land type of the charging station. F1 can be calculated by formula (18):

[0158]

[0159] Where: r is the depreciation rate, t is the operating life of the power station, S jmc is the number of type m charging piles, G mc is the unit price of type m charging pile, F ck is the construction and operation cost of a charging station with charging equipment k. jk is the charging equipment decision variable, h jk The number of rows is the number of candidate charging station addresses in the candidate power station set, and the number of columns is the level of the candidate charging station. jk When , first randomly select a candidate address in the candidate power station set to determine the position of the row, then check the number of charging stations in the position to determine the level of the candidate charging station, so as to find the position of the column, and finally set 1 in the corresponding position and set the rest of the positions in this row to 0.

[0160] F2 is the construction cost of the battery swap station, including infrastructure costs, battery swap equipment costs, and battery costs. Infrastructure costs mainly include costs of business buildings and road paving, battery swap equipment costs mainly include costs of distribution transformers, chargers, battery swap lines (battery swap stations), battery swap robots, etc. Battery costs mainly include battery purchase costs and battery storage costs. F2 can be calculated by formula (19):

[0161]

[0162] Where: S lm is the capacity of the battery swap station l, which is expressed as the reserve battery storage capacity of the battery swap station, P p_m F is the battery purchase cost and battery storage cost, ek is the construction cost of the battery swap station, which mainly includes infrastructure costs, battery swap equipment costs, and battery costs. rk The operating cost of the battery swap station mainly includes wage costs, equipment maintenance costs and land rental costs. lk is the decision variable of the battery swapping equipment, h lk The number of rows is the number of candidate power station addresses in the candidate power station set, and the number of columns is the level of the candidate power station. lk When , first randomly select a candidate address in the candidate power station set to determine the position of the row, then check the number of charging stations in the position to determine the level of the candidate battery swap station, so as to find the position of the column, and finally set 1 in the corresponding position and set the rest of the positions in this row to 0.

[0163] (b) User travel costs include time cost in the road, queuing time cost, and battery replacement waiting time cost, which can be calculated by formula (20):

[0164] C T =C wT +C vT +C eT (20)

[0165] Where: C T is the user's travel cost, C wT is the waiting time cost at the station, C vT is the time cost in the process, C eT The time cost of battery replacement.

[0166] The annual waiting time cost of queuing at the station is determined by the expected waiting time of queuing at the station, which can be calculated by formula (21):

[0167]

[0168] Where: δ is the urban travel time cost coefficient, and are the expected waiting time of fast charging and slow charging users in charging station j, p1 is the probability of electric vehicles using fast charging, The proportion of users charging at demand points, a i is the electricity demand at demand point i, P G is the user's unit time cost, X ij is the charging time decision variable.

[0169] The time cost in the road can be calculated by formula (22):

[0170]

[0171] The battery replacement time cost can be calculated by formula (23):

[0172]

[0173] Where: T ij and T il are the time taken for users at electricity demand point i to travel to candidate charging station j and candidate battery swap station l, respectively, c and ε s are the average number of times each electric vehicle is charged and replaced per day, and are the proportion of charging users and battery swapping users in electricity demand points, X ij and X il are the decision variables for charging and battery replacement time respectively.

[0174] (c) Carbon emission costs

[0175] Electric vehicles consume electricity on the way to the battery swap station, and a certain amount of carbon emissions will be generated during the production of electricity. By introducing carbon emission costs into the battery swap station site selection and planning model, the environmental impact of users on the way to the battery swap station can be quantified. The carbon emission cost can be calculated by formula (24):

[0176]

[0177] Where: μ is the emission factor of the electricity generated by the grid, d ij d il are the distances from the user at the electricity demand point i to the candidate charging station j and the candidate battery swap station l, γ is the power consumption of the electric vehicle per kilometer, and p D For the carbon dioxide treatment fee.

[0178] (d) Power station revenue

[0179] The charging station first purchases electricity from the grid, and then provides charging services to electric vehicle users. Its revenue comes from the difference in electricity prices and the service fee for each charging. The charging station revenue P to1It can be calculated by formula (25):

[0180]

[0181] Where: Pr j is the average charging electricity price at charging station j, including charging unit price and service fee, p w is the grid electricity price, e ci is the average power consumption of electric vehicles per charge at demand point i, γ is the power consumption of electric vehicles per kilometer, e full The amount of electricity required to fully charge an electric vehicle, e initial is the initial charge of the electric vehicle.

[0182] The battery swap station uses the spare batteries stored in the warehouse to provide battery swap services to users. Its revenue comes from the battery swap service fee minus the cost of charging the empty battery. The revenue of the battery swap station is P to2 It can be calculated by formula (27):

[0183]

[0184] S32: Construct a capacity determination model for the lower-level charging and battery swapping stations, that is, construct a power station layout optimization model with the goals of maximizing power station revenue, minimizing power station construction and operation costs, minimizing power station carbon emission costs, and minimizing user travel costs.

[0185] The model is expressed as:

[0186] f(S jmc ,S lm ,p t )=P to -FC EV -C T ,

[0187] The objective function is expressed as:

[0188] maxf(S jmc ,S lm ,p t )=P to -FC EV -C T (28)

[0189] Where: S jmc is the number of type m charging piles in charging station j, S lm is the capacity of the battery swap station, p t is the average charging electricity price in the tth year, P to is the power station revenue, F is the power station construction and operation cost, C EV is the carbon emission cost of the power station, C T Travel costs for users.

[0190] Constraints:

[0191] Variable constraints: users at each demand point can only receive power supply services from the same power supply station during the same time period:

[0192]

[0193] Only when a power supply station is established at the candidate point can services be provided to users:

[0194]

[0195] n E and n L They are the number of charging stations and the number of battery swap stations respectively.

[0196] Formulas (29) and (30) indicate that users at each demand point can only go to the same charging station to receive charging services during the same time period. Formulas (31) and (32) are the constraints on the number of charging and swapping stations.

[0197] Decision variables:

[0198] X ij ,X il ,h jk ,h lk ∈{0,1},

[0199] X ij ≤h jk (33),

[0200] X il ≤h lk (34),

[0201] i∈N, j∈M, k∈K, where K represents the total number of charging devices.

[0202] Formulas (33) and (34) are constraints between decision variables, indicating that charging stations can only provide services to users if charging stations are established at candidate points.

[0203] Constraints on the number of charging stations:

[0204]

[0205] Constraints on the number of battery swap stations:

[0206]

[0207] Formula (35) is the constraint on the number of charging stations, and formula (36) is the constraint on the number of battery swap stations.

[0208] Where: S Jmax and S Jminare the maximum and minimum capacity of the charging station, S Lmax and S Lmin They are the maximum and minimum capacities of the battery swap station respectively.

[0209] Service capacity restrictions of candidate power plants:

[0210]

[0211] Battery swap station capacity limit:

[0212] 30≤S lm ≤90(38),

[0213] Formula (37) is the service capacity limit of the candidate power station, and formula (38) is the capacity limit of the battery swap station.

[0214] Distance constraint: the minimum distance constraint should be met between charging stations:

[0215] d jj′ ≥d min , j′∈J, j≠j′ (39),

[0216] Where: d jj′ is the distance from charging station j to charging station j′, d min is the minimum proximity distance between charging stations.

[0217] Charging price constraints:

[0218]

[0219] The minimum charging price is usually set to cover the cost of providing the service, including electricity cost, infrastructure cost, maintenance cost, tax and profit. On the other hand, the maximum charging price is mainly determined by electricity market competition and consumer demand. In some cases, market regulators may set a cap on charging prices to protect consumers from price gouging. Formula (40) is the charging price constraint.

[0220] Where: and Pr They are respectively the upper and lower limits of charging prices set by the market regulatory authorities.

[0221] Charging / battery replacement demand constraints:

[0222]

[0223] Formulas (41) and (42) are constraints on the service capacity of the power station. The service capacity of the power station should meet the electricity demand of electric vehicle users within the region.

[0224] S4: Genetic algorithm is an optimization algorithm that simulates natural selection and genetic mechanisms and is often used to solve complex optimization problems. The two-layer model is a nested structure containing two optimization problems. The solution of the upper model will affect the optimization of the lower model. Therefore, this step uses an intelligent optimization algorithm to solve the two-layer model. According to the solution results, the optimal charging and battery swap station locations, station capacity and charging price for the first year are output.

[0225] S41: First initialize the population:

[0226] In the upper model, the weights θ1, θ2, and θ3 in the objective function are encoded as floating point numbers, then x1 = [θ1, θ2, θ3].

[0227] In the lower model, for S jmc , S lm 、p t Mixed coding, S jmc Contains 2 parameters: S j1c and S j2c , respectively the number of fast charging and slow charging piles. lm Only for itself. t Contains two parameters: Pr j and P bct , then x2=[S j1c ,S j2c ,S lm ,Pr j ,P bct ].

[0228] S42: Set the fitness function. For the upper model, its fitness function is formula (9) or formula (10), depending on the user's electricity usage. For the lower model, its fitness function is: f(S jmc ,S lm ,p t )=P to -FC EV -C T .

[0229] S43: Execute the genetic operator to iterate continuously to obtain the optimal solution.

[0230] S5: Update data and make future plans. Based on the output of the lower-level model in the first year, update the regional data and make plans for the next two years.

[0231] After executing step S4, a set of 2×n decision variable matrices is first output, the first row is the candidate address of the battery swap station to be established, and the second row is the candidate address of the charging station to be established. Then the output matrix x2=[S j1c ,S j2c ,S lm ,Prj ,P bct ], give the charging cost and specific service capacity of the charging and swapping stations in the decision variable matrix, that is, the number of charging piles or battery stocks. Update the candidate address set for the second year, the service capacity and price of the built power stations, and update the electricity demand in the area to the predicted data for the second year. The same is true for the third year, and it is updated after the second execution of step S4. Until the final solution is obtained.

[0232] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that the features described in conjunction with a single embodiment may be used in other described embodiments.

Claims

1. A dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model, characterized in that: include: Predict the future electricity demand in the target area based on the electric vehicle demand data in the target area; Establish an upper user selection model, the upper user selection model includes: charging selection model and battery swap selection model And the expressions are: The upper-level user selection model constraints are as follows: θ1,θ2,θ3∈{0,1},θ1+θ2+θ3=1,∑hij=1,d il ≤D max , Among them, T ij ', AF ij ' m , d i ' l , T il ′ respectively represent the ij , AF ijm , d il , T il The normalized result, T ij and T il are the time taken for users at electricity demand point i to travel to candidate charging station j and candidate swap station l, respectively. ijm is the attractiveness of candidate charging station j to m types of charging users in electricity demand point i, is the expected waiting time for charging users of type m in candidate charging station j, d il is the distance from the electricity demand point i to the candidate battery swap station l, is the battery replacement cost of candidate battery replacement station l, h ij is the user decision variable, θ1, θ2 and θ3 are constraint weights, D max is the maximum driving distance of the electric vehicle, i∈N, j∈M, l∈L, N is the total number of electricity demand points, M is the total number of candidate charging stations, and L is the total number of candidate battery swap stations; A lower-layer power station capacity model is established, and the expression of the lower-layer power station capacity model f(·) is: f(S jmc ,S lm ,p t )=P to -F-C EV -C T , The constraints of the lower power station capacity model are as follows: 30≤S lm ≤90， Among them, S jmc is the number of type m charging piles in charging station j, S lm is the capacity of the battery swap station l, p t is the average charging electricity price in the tth year, P to is the power station revenue, F is the power station construction and operation cost, C EV is the carbon emission cost of the power station, C T Travel costs for users; Genetic algorithms are used to solve the minimum value of the upper user selection model and the maximum value of the lower power station capacity model, respectively, to obtain the optimal power station coordinates, capacity and electricity price, and then dynamically plan the layout of electric vehicle charging and swapping stations.

2. According to claim 1, a dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model is characterized in that: The predicting of the future electricity demand of the target area according to the electric vehicle demand data of the target area includes: Collecting data on the number of electric vehicles in the target area and its ancillary information, permanent population, per capita income, charging prices and oil prices. The ancillary information includes: quarterly sales volume of electric vehicles, market share and number of charging piles; The collected data is input into the support vector machine prediction model for training, and the trained support vector machine prediction model is used to predict the future electricity demand of the target area.

3. A dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model according to claim 1 or 2, characterized in that: The establishing of the upper user selection model comprises: Collect the coordinates of public parking lots, residential areas and commercial areas in the target area respectively; Perform cluster analysis on the coordinates of public parking lots, and select the coordinate points of public parking lots closest to each cluster center as candidate power stations; Perform cluster analysis on the coordinates of residential and commercial areas, and use the coordinate points of residential or commercial areas closest to each cluster center as electricity demand points; Divide the future electricity demand of the target area into three modes: fast charging, slow charging and battery replacement, and allocate different electricity demands to each electricity demand point, so that the ratio of fast charging demand, slow charging demand and battery replacement demand in each electricity demand point is 16:4:5; Calculate the driving time, driving distance, electricity price and electricity queuing time from each electricity demand point to each candidate power station respectively, the electricity price includes charging price and battery swap price, the electricity queuing time includes charging queuing time and battery swap queuing time, and the candidate power stations include candidate charging stations and candidate battery swap stations; An upper-level user selection model is constructed using the travel time, travel distance, electricity price, and electricity queuing time.

4. The method for dynamic layout planning of electric vehicle charging and swapping stations based on a double-layer model according to claim 3 is characterized in that: The time T for the user at the electricity demand point i to go to the candidate charging station j ij The expression is: T ij =t0[1+α(q / c) β ], Among them, t0 is the free flow travel time from the electricity demand point i to the candidate charging station j, α and β are regression coefficients, c is the actual traffic capacity from the electricity demand point i to the candidate charging station j, and q is the traffic volume; The time T for the user at the electricity demand point i to go to the candidate battery swap station l il The expression is: T il =t0′[1+α(q / c′) β ], Among them, t0′ is the free flow travel time within the section from the power demand point i to the candidate battery swap station l, and c′ is the actual traffic capacity within the section from the power demand point i to the candidate battery swap station l; The attractiveness AF of the candidate charging station j to the m types of charging users in the electricity demand point i ijm The expression is: Among them, AOS jm is the service capacity of candidate charging station j for electric vehicles with charging type m, Pr j is the average charging price of candidate charging station j, d ij is the distance from the electricity demand point i to the candidate charging station j, d ij ≤D max ; The expected waiting time of m types of charging users in candidate charging station j The expression is: Among them, ρ m is the service intensity of type m charging pile, P m0 is the probability that all m types of charging piles are idle, λ jm The probability of an electric vehicle of type m charging arriving at candidate charging station j; The battery replacement cost of the candidate battery replacement station l The expression is: Among them, C fc is the fixed cost, P bc For charging price.

5. The method for dynamic layout planning of electric vehicle charging and swapping stations based on a double-layer model according to claim 4 is characterized in that: The service capability AOS of the candidate charging station j for electric vehicles with charging type m is jm The expression is: TO THE jm =S jmc P m , Among them, P m is the rated power of the Class M charging pile; The service intensity ρ of the m-type charging pile m The expression is: Among them, μ m Charging time for electric vehicles of type M charging; The probability P that all the m types of charging piles are idle m0 The expression is: Where τ=0,1,...,S jmc -1; The fixed cost C fc The expression is: C fc =P BR +P ED +P LC , P BR is the battery replacement cost, P ED is the equipment depreciation cost, P LC For labor costs.

6. A dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model according to claim 1 or 2, characterized in that: The establishing of the lower power station capacity model comprises: Calculate the power station construction and operation cost F and user travel cost C respectively T , power plant carbon emission cost C EV and power station revenue P to , and establish a capacity model for the lower power station.

7. A dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model according to claim 6, characterized in that: The power station construction and operation cost F includes the construction cost F1 of the charging station and the construction cost F2 of the battery swap station, and the expressions are: Among them, r is the depreciation rate, t is the operating life of the power station, G mc is the unit price of type m charging pile, F ck is the construction and operation cost of a charging station with k charging devices, P p_m F is the cost of purchasing and storing batteries. ek is the construction cost of the battery swap station, F rk is the operating cost of the battery swap station, K represents the total number of charging equipment, and h jk and h lk are the decision variables for charging and battery swapping equipment respectively; The power station income P to Including charging station revenue P to1 and the revenue of the battery swap station P to2 , and the expressions are: Among them, Pr j is the average charging price of candidate charging station j, p w is the grid electricity price, e ci is the average power consumption of electric vehicles per charge at the power demand point i, γ is the power consumption of electric vehicles per kilometer, and ε c and ε s are the average number of times each electric vehicle is charged and replaced per day, and are the proportions of charging users and battery swapping users in electricity demand points, a i is the electricity demand at electricity demand point i, P bc For charging price; The user travel cost C T The expression is: C T =C wT +C vT +C eT , Among them, C wT is the waiting time cost at the station, C vT is the time cost in the process, C eT The time cost of battery replacement; The power station carbon emission cost C EV The expression is: Where μ is the emission factor of the electricity generated by the grid, d ij is the distance from the electricity demand point i to the candidate charging station j and d ij ≤D max , p D is the carbon dioxide treatment fee, X ij and X il are the decision variables for charging and battery replacement time, respectively; X ij ,X il ,h jk ,h lk ∈{0,1},Xij≤hjk,X il ≤h lk , n E and n L are the number of candidate charging stations and the number of candidate battery swap stations, respectively, Jmax and S Jmin are the maximum and minimum capacity of the candidate charging stations, S Lmax and S Lmin are the maximum and minimum capacities of the candidate battery swap stations respectively.

8. The method for dynamic layout planning of electric vehicle charging and swapping stations based on a double-layer model according to claim 7 is characterized in that: The waiting time cost C wT 、Time cost in the process C vT and the battery replacement time cost C eT The expressions are: Among them, δ is the urban travel time cost coefficient, and are the expected waiting time for fast-charging and slow-charging users in candidate charging station j, p1 is the probability of electric vehicles using fast charging, P G The unit time cost of the user.

9. A dynamic layout planning method for electric vehicle charging and swapping stations based on a double-layer model according to claim 8, characterized in that: The transportation conditions of the lower power station capacity model include: d jj′ ≥d min ,j≠j′, d jj′ is the distance from candidate charging station j to candidate charging station j′, d min is the minimum proximity distance between candidate charging stations, Pr j is the average charging electricity price of candidate charging station j, and Pr They are respectively the upper and lower limits of charging prices set by market regulators.

10. The method for dynamic layout planning of electric vehicle charging and swapping stations based on a double-layer model according to claim 7, characterized in that: The charging equipment decision variable h jk The number of rows is the number of candidate charging station addresses in the candidate power station set, and the number of columns is the level of the candidate charging station. The battery swapping equipment decision variable h lk The number of rows is the number of candidate battery swap station addresses in the candidate station set, and the number of columns is the level of the candidate battery swap station.

Images