A charging station planning method considering the “road-network-station” network and load

Abstract

A charging station planning method that considers the "road-network-station" network and load, comprising the following steps: Step 1: Constructing a road-network-station coupled network topology, taking into account the impact of traffic conditions, overall grid operation, and charging station queues on the planning results; Step 2: Constructing a charging decision model for electric vehicle users based on a speed-flow model and charging load distribution predictions; Step 3: Constructing a multi-objective planning model for charging stations that considers load balancing effects; and Step 4: Solving the constructed multi-objective planning model for charging stations using an improved immune genetic algorithm. The present invention aims to address the technical issues in existing charging station planning, such as the difficulty in significantly reducing the time it takes electric vehicles to reach charging stations by optimizing their layout, as well as the technical issues that can easily lead to uneven load distribution within a region and increase regional grid voltage fluctuation.

Description

Technical Field

[0001] The present invention belongs to the technical field of power grids, and in particular relates to a charging station planning method that takes into account a "road-grid-station" network and loads. Background Art

[0002] While the booming electric vehicle market plays a key role in clean transportation and the transition to low-carbon energy, it also presents a series of challenges in charging station operations, their rational layout and planning, and the optimized operation of power systems. Unguided and blind construction of electric vehicle charging stations will not only prevent them from fully realizing their charging service capabilities and unleashing their economic and social benefits, but may also seriously impact the safe and stable operation of regional power grids.

[0003] Some literature has introduced the concept of load balancing into charging station planning research. Most of these approaches begin with a preliminary site selection for the charging station, then use methods such as Voronoi diagrams and clustering algorithms to divide the service area of ​​the charging station and determine the charging demand for each charging station. However, this approach presents two problems:

[0004] 1) For example, the literature "A Correction Planning Method for Expressway Fast Charging Stations Considering Distribution Network Flow Constraints" and the literature "A Planning Method for Urban Centralized Charging Stations" achieve the division of the charging station service area by introducing methods such as Voronoi diagrams or clustering algorithms. However, methods such as Voronoi diagrams use geometric calculations to cluster nodes close to the center point into point sets to generate polygonal networks, where the distance from the node to the center point is mostly Euclidean distance. In fact, due to the increasing complexity and diversity of urban traffic and road network structures, specific point-to-point distances are difficult to measure using traditional Euclidean distance methods, and the impact of traffic flow on user time cannot be measured by simply calculating the shortest path. Therefore, it is difficult for the above methods to significantly reduce the time it takes for electric vehicles to reach charging stations by optimizing the reasonable layout of charging stations.

[0005] 2) Due to the spatiotemporal distribution of EV charging demand and the uneven distribution of EVs at each transportation node, the aforementioned literature inevitably results in uneven load distribution within each area after dividing the service area using methods such as Voronoi diagrams. This not only leads to persistent queues in charging hotspots but also increases voltage volatility and network loss costs in the regional power grid, significantly impacting the economic and stable operation of the regional power grid.

[0006] In light of the above situation, rational planning and layout of charging facilities can provide an important guarantee for meeting the growing charging demand of electric vehicles. Influenced by user travel behavior, rational planning of charging infrastructure based on the load distribution of electric vehicles is an urgent problem that needs to be solved. Summary of the Invention

[0007] The purpose of the present invention is to solve the technical problems in the existing charging station planning, such as the difficulty in significantly reducing the time it takes for electric vehicles to reach the charging station by optimizing the reasonable layout of the charging station, and the technical problems of easily causing uneven load distribution in the region and easily increasing the voltage fluctuation of the regional power grid.

[0008] In order to solve the above technical problems, the following technical solutions are adopted:

[0009] A charging station planning method considering the "road-network-station" network and load includes the following steps:

[0010] Step 1: Consider the impact of traffic conditions, overall grid operation, and charging station queues on planning results and construct a road-grid-station coupled network topology;

[0011] Step 2: Build a charging decision model for electric vehicle users based on the speed-flow model and charging load distribution prediction results;

[0012] Step 3: Construct a multi-objective planning model for charging stations considering the load balancing effect;

[0013] Step 4: Use the improved immune genetic algorithm to solve the constructed multi-objective planning model of the charging station.

[0014] In step 1, when constructing the road-network-station coupling network topology, the details are as follows:

[0015] The topology of the traffic network is shown in formula (1):

[0016]

[0017] Where G L is the road network topology; V L It is the traffic node in the road network structure, that is, the intersection node of the road; E L Indicates the road network edge; is the adjacency matrix of the road network, which is used to describe the connection relationship between traffic nodes and the road length; n L Indicates the total number of traffic nodes; and represent the i-th and j-th traffic nodes respectively; Representation node and connected roads; eij is the length of the road section, and the specific expression is shown in formula (2);

[0018]

[0019] Where, d ij Indicates road section The length of ; inf means infinity;

[0020] It can be expressed by formula (3):

[0021]

[0022] Considering only the connection edges and nodes of the power grid, the topological structure of the power grid is obtained, and its topological structure can be expressed by formula (4):

[0023]

[0024] Where G D is the grid topology; V D is the node set of the power grid; E D is the set of grid edges; is the parameter matrix of the relevant information of the power grid branch; v G Indicates the total number of grid nodes; Representation node and The connected power grid line; xy , z xy , c xy and They represent the resistance, reactance, susceptance of the power grid branch xy and the transmission power limit of the branch;

[0025] The topology of the power station layer is shown in formula (5):

[0026]

[0027] Where G C is the charging station topology; V C A set of nodes representing charging stations; Represents the relevant information parameter matrix of the charging station; represents the number of the kth charging station; v C Indicates the total number of charging stations; is the number of charging piles in the kth charging station; Indicates the capacity of the newly built supporting transformer; represents the land price and labor costs at the location of the k-th charging station;

[0028] The road-network-station coupling network topology also includes the interconnection relationship between the road-network-station levels, and its specific expression is shown in formula (6):

[0029]

[0030] Where G represents the “road-network-station” road-network-station coupling network topology; E L-D-C is the coupling network edge, which represents the coupling between the road network node, the power grid node and the charging station access node; is the coupling network edge that is not connected to the charging station, that is, the coupling between the road network node and the power grid node; It is the coupling network edge connected to the charging station, that is, the coupling of the three nodes of "road-network-station".

[0031] In step 2, when building the charging decision model for electric vehicle users, the following steps are adopted:

[0032] Step 2-1: Simplify the distribution of electric vehicles to dispersed at the intersection nodes of the road network. Calculate the total time it takes for each node to reach each charging station in a day. With the goal of minimizing the total time, decide which charging station service area the charging load of the node belongs to.

[0033] Step 2-2: Nodes with the same decision-making plan are assigned to the same charging station service area.

[0034] In step 2-1, the objective function is shown in formula (7):

[0035]

[0036] Where, is the total time spent on the road from node i to the kth charging station; η is the decision variable; is the time it takes for the electric vehicle at node i to reach the kth charging station at time t; N i,t is the number of electric vehicles at node i at time t; Ω t is the time set; Ω i is a set of road nodes;

[0037]

[0038] Where, L i,k,q is the road section q that the electric vehicle at node i passes through to the kth charging station; v q (t) is the speed of the electric vehicle at the road section q at time t; Ω q is the set of paths passed by the electric vehicle;

[0039] v q (t) can be calculated by formula (9):

[0040]

[0041] Where, represents the average driving speed of road section q; ρ q (t) is the traffic flow density of the qth road section at time t; C q,max represents the maximum traffic flow of the qth road segment; a, b and μ are the adaptive coefficients of the road; N q (t) represents the number of electric vehicles passing the qth road segment at time t; θ is the market penetration rate of electric vehicles;

[0042] In step 2-2, nodes with the same decision-making scheme are assigned to the same charging station service area. The specific charging demand distribution of charging stations is shown in formulas (10)-(11):

[0043]

[0044]

[0045] Among them, D t,k represents the charging demand of the kth charging station at time t;

[0046] The constraints of the electric vehicle user charging decision model can be expressed as:

[0047] 1) Driving distance constraint:

[0048] L i,k ≤L max (12)

[0049] Where, L max is the maximum distance between the electric vehicle and the charging station;

[0050] 2) Queuing waiting time constraints:

[0051]

[0052] Where, The maximum waiting time for electric vehicles at charging stations.

[0053] In step 3, the multi-objective planning model of the charging station considering the load balancing effect is constructed as follows:

[0054] The objective function of the multi-objective planning model of the charging station is:

[0055] maxF1=C S -C I,CS -C O,CS -C I,TR -C O,TR -C L,G -C L,TR(14)

[0056] Where C S is the planned total revenue of the charging station; C I,CS is the total planned investment cost of the charging station; C O,CS is the total operation and maintenance cost of the charging station; C I,TR The investment cost of the transformer for the charging station; C O,TR The operation and maintenance cost of the supporting transformer; C L,G is the network loss cost of the distribution network; C L,TR is the loss cost of the transformer;

[0057] The total revenue from the charging station plan is:

[0058]

[0059] Where, The time-of-use electricity price for charging stations; The time-of-use electricity purchase price for the charging station from the power grid company; B t,k is the number of charging piles that are busy at the kth charging station at the tth moment; P S is the rated power of the charging pile;

[0060] The total planned investment cost of the charging station is:

[0061]

[0062] Where, is the area occupied by the charging station, which is directly proportional to the number of charging piles in the charging station; The land price; Planning capacity for charging stations; is the unit price of the charging pile; r0 is the discount rate; T is the planned service life;

[0063] The total operation and maintenance cost of the charging station is:

[0064]

[0065] Where C M,CS The unit charging equipment operation and maintenance cost;

[0066] The investment cost of the transformer for the charging station is:

[0067]

[0068]

[0069] Where, The planned capacity of the transformer supporting the charging station; The unit price of its corresponding capacity; the transformer capacity planning should be adapted to the number of charging equipment in the charging station, and Plan the capacity of the transformer for the number of charging piles;

[0070] The operation and maintenance cost of the supporting transformer is:

[0071]

[0072] Where C M,TR is the unit operation and maintenance cost of the transformer; T max is the maximum annual utilization hours;

[0073] The loss cost of the transformer is:

[0074]

[0075] Where c TR is the unit loss cost of the transformer; L fe and L cu are the rated iron loss and copper loss of the transformer respectively;

[0076] The network loss cost of the distribution network is:

[0077]

[0078] Where c loss is the unit network loss cost of the distribution network; is the line loss of the lth line of the power grid; Ω l is a collection of lines;

[0079] The number of charging piles at a charging station is configured as shown in formula (23):

[0080]

[0081] Where, The maximum number of charging piles planned for the charging station is The minimum value of the maximum number of electric vehicles arriving at the service area in one day;

[0082] The number of electric vehicles queuing for charging at the kth charging station at time t can be obtained by formula (24):

[0083] Q t,k =min{Q t-1,k +D t,k -I t,k ,0} (24)

[0084] Where Q t-1,kis the number of electric vehicles queuing for charging at the kth charging station at time t-1; I t,k represents the number of charging piles that are not busy at the kth charging station at time t, which can be calculated by formula (25);

[0085]

[0086] Where B t,k and M t,k They represent the number of charging piles in the busy state and the state of charging completion at the kth charging station at time t respectively;

[0087] B t,k and M t,k It can be calculated by formula (26) and (27) respectively:

[0088] B t,k =B t-1,k -M t-1,k +K t-1,k (26)

[0089] M t,k =K t-2,k (27)

[0090] Where B t-1,k and M t-1,k K represents the number of charging piles that are busy and about to be charged at the kth charging station at time t-1; t-1,k and K t-2,k They represent the number of newly charged electric vehicles at the k-th charging station at time t-1 and time t-2 respectively;

[0091] The number of newly charged electric vehicles in the kth charging station at time t can be expressed as:

[0092] K t,k =min{Q t-1,k +D t,k ,I t,k} (28)

[0093] The queuing time of electric vehicles at the kth charging station at time t can be obtained from formulas (24)-(28):

[0094]

[0095] Where K q is the queue impact coefficient;

[0096] After taking into account the queuing time, the total time the user spends traveling to the charging station is:

[0097]

[0098] Where, is the waiting time of electric vehicles at the kth charging station and time t.

[0099] In step 3, the constraints of the multi-objective programming model of the charging station can be expressed as:

[0100] 1) Upper and lower limit constraints of node amplitude:

[0101] U min ≤U x ≤U max (31)

[0102] Where U x is the node voltage at node x, U max and U min are its upper and lower limits respectively;

[0103] 2) Branch flow constraints:

[0104]

[0105] Where: P x,t and Q x,t are the active power and reactive power of node x at time t; U x,t and U y,t are the voltage amplitudes of nodes x and y at time t; G xy and B xy are the conductance and susceptance of branch xy respectively; θ xy is the phase angle difference between the voltages at nodes x and y; Ω D is the node set of the power grid;

[0106] 3) Planning the number of charging piles and transformer capacity constraints within the charging station:

[0107]

[0108] Where, and To plan the upper and lower limits of the number of charging piles in the charging station;

[0109] 4) Planning the number of charging stations:

[0110]

[0111]

[0112] Where, and The planned number of charging stations is M CS The upper and lower limits of N totalis the total number of electric vehicles in the planning area; Q is the battery capacity; T av is the average charging time; Indicates rounding down;

[0113] 5) Voltage deviation constraint:

[0114]

[0115] Where U0 is the reference voltage value; f vd is the voltage fluctuation rate; is the maximum volatility constraint;

[0116] 6) Distance constraints between charging stations:

[0117]

[0118] Where, and Represents the distance between any two charging stations Maximum and minimum constraints;

[0119] 7) Substation main transformer load rate constraints:

[0120]

[0121] Where, is the basic power load at the grid node x to which the kth charging station is connected; represents the electric vehicle charging load at the grid node x to which the kth charging station is connected; Configure the rated capacity of the transformer for the kth charging station; α max To plan the upper limit of transformer load rate;

[0122] 8) Transformer capacity constraints:

[0123]

[0124] Where, and To plan the upper and lower limits of transformer capacity.

[0125] In step 4, when solving the constructed multi-objective programming model of the charging station, the following steps are adopted:

[0126] Step 4-1: Initialize parameters;

[0127] Step 4-2: Generate initial antibody population and memory library;

[0128] Step 4-3: Calculate individual fitness; substitute the corresponding parameters into the objective function to obtain the corresponding F1 and F2; considering that the profit calculation of the charging station is to maximize the profit, while the goal of electric vehicle users is to minimize the time spent on charging; in order to facilitate the comparison of the advantages and disadvantages of different planning schemes, define the global variable operator F 1,max and F 2,min , used to normalize the multi-objective values; after normalizing F1 and F2, the global fitness can be calculated using formula (40):

[0129]

[0130] Where ω1 and ω2 are weights;

[0131] Step 4-4: Calculate the affinity and antibody concentration between individual antibodies; the affinity S between individual antibodies u,v It represents the similarity between antibodies u and v, and the antibody concentration Con represents the overall similarity of the antibody population;

[0132]

[0133]

[0134] Where, L u,v Indicates the length of the antibody with the same value as u and v; ls represents the total length of the antibody individual; m is the total number of antibody populations;

[0135] Step 4-5: Determine whether the antibody population meets the conditions; determine whether the antibody concentration of the antibody population is equal to 1. If so, obtain the planning result and terminate the cycle; if not, continue with steps 6-8;

[0136] Steps 4-6: Selection operation: select excellent individuals from the exchanged individual antibodies, and use the "roulette wheel" method to select individuals from the antibody population to generate a new population;

[0137] Steps 4-7: Crossover operation: select two antibody individuals from the antibody population according to a certain exchange probability and randomly exchange them to generate new individuals;

[0138] Step 4-8: Select antibody individuals according to a certain probability, and randomly change the value of a gene in the string structure data of the selected individuals to generate new individuals; and return to step 3 for a new round of iteration.

[0139] A method for constructing an electric vehicle user charging decision model comprises the following steps:

[0140] Step S1: Simplify the distribution of electric vehicles to dispersed at the intersection nodes of the road network, calculate the total time it takes for each node to reach each charging station in a day, and decide which charging station service area the charging load of the node belongs to with the goal of minimizing the total time;

[0141] Step S2: Nodes with the same decision-making plan are assigned to the same charging station service area.

[0142] In step S1, the objective function is as shown in formula (7):

[0143]

[0144] Where, is the total time spent on the road from node i to the kth charging station; η is the decision variable; is the time it takes for the electric vehicle at node i to reach the kth charging station at time t; N i,t is the number of electric vehicles at node i at time t; Ω t is the time set; Ω i is a set of road nodes;

[0145]

[0146] Where, L i,k,q is the road section q that the electric vehicle at node i passes through to the kth charging station; v q (t) is the speed of the electric vehicle at the road section q at time t; Ω q is the set of paths passed by the electric vehicle;

[0147] v q (t) can be calculated by formula (9):

[0148]

[0149] Where, represents the average driving speed of road section q; ρ q (t) is the traffic flow density of the qth road section at time t; C q,max represents the maximum traffic flow of the qth road segment; a, b and μ are the adaptive coefficients of the road; N q (t) represents the number of electric vehicles passing the qth road segment at time t; θ is the market penetration rate of electric vehicles;

[0150] In step S2, nodes with the same decision-making scheme are assigned to the same charging station service area. The specific charging demand distribution of the charging station is shown in formulas (10)-(11):

[0151]

[0152]

[0153] Among them, D t,k represents the charging demand of the kth charging station at time t;

[0154] The constraints of the electric vehicle user charging decision model can be expressed as:

[0155] 1) Driving distance constraint:

[0156] L i,k ≤L max (12)

[0157] Where, L max is the maximum distance between the electric vehicle and the charging station;

[0158] 2) Queuing waiting time constraints:

[0159]

[0160] Where, The maximum waiting time for electric vehicles at charging stations.

[0161] A multi-objective planning model for charging stations considering load balancing effects is constructed.

[0162] The objective function of the multi-objective planning model of the charging station is:

[0163] maxF1=C S -C I,CS -C O,CS -C I,TR -C O,TR -C L,G -C L,TR (14)

[0164] Where C S is the planned total revenue of the charging station; C I,CS is the total planned investment cost of the charging station; C O,CS is the total operation and maintenance cost of the charging station; C I,TR The investment cost of the transformer for the charging station; C O,TR The operation and maintenance cost of the supporting transformer; C L,G is the network loss cost of the distribution network; C L,TR is the loss cost of the transformer;

[0165] The total revenue from the charging station plan is:

[0166]

[0167] Where, The time-of-use electricity price for charging stations; The time-of-use electricity purchase price for the charging station from the power grid company; B t,k is the number of charging piles that are busy at the kth charging station at the tth moment; P S is the rated power of the charging pile;

[0168] The total planned investment cost of the charging station is:

[0169]

[0170] Where, is the area occupied by the charging station, which is directly proportional to the number of charging piles in the charging station; The land price; Planning capacity for charging stations; is the unit price of the charging pile; r0 is the discount rate; T is the planned service life;

[0171] The total operation and maintenance cost of the charging station is:

[0172]

[0173] Where C M,CS The unit charging equipment operation and maintenance cost;

[0174] The investment cost of the transformer for the charging station is:

[0175]

[0176]

[0177] Where, The planned capacity of the transformer supporting the charging station; The unit price of its corresponding capacity; the transformer capacity planning should be adapted to the number of charging equipment in the charging station, and Plan the capacity of the transformer for the number of charging piles;

[0178] The operation and maintenance cost of the supporting transformer is:

[0179]

[0180] Where C M,TR is the unit operation and maintenance cost of the transformer; T max is the maximum annual utilization hours;

[0181] The loss cost of the transformer is:

[0182]

[0183] Where c TRis the unit loss cost of the transformer; L fe and L cu are the rated iron loss and copper loss of the transformer respectively;

[0184] The network loss cost of the distribution network is:

[0185]

[0186] Where c loss is the unit network loss cost of the distribution network; is the line loss of the lth line of the power grid; Ω l is a collection of lines;

[0187] The number of charging piles at a charging station is configured as shown in formula (23):

[0188]

[0189] Where, The maximum number of charging piles planned for the charging station is The minimum value of the maximum number of electric vehicles arriving at the service area in one day;

[0190] The number of electric vehicles queuing for charging at the kth charging station at time t can be obtained by formula (24):

[0191] Q t,k =min{Q t-1,k +D t,k -I t,k ,0} (24)

[0192] Where Q t-1,k is the number of electric vehicles queuing for charging at the kth charging station at time t-1; I t,k represents the number of charging piles that are not busy at the kth charging station at time t, which can be calculated by formula (25);

[0193]

[0194] Where B t,k and M t,k They represent the number of charging piles in the busy state and the state of charging completion at the kth charging station at time t respectively;

[0195] B t,k and M t,k It can be calculated by formula (26) and (27) respectively:

[0196] B t,k =B t-1,k -M t-1,k +K t-1,k (26)

[0197] M t,k =K t-2,k (27)

[0198] Where B t-1,k and M t-1,k K represents the number of charging piles that are busy and about to be charged at the kth charging station at time t-1; t-1,k and K t-2,k They represent the number of newly charged electric vehicles at the k-th charging station at time t-1 and time t-2 respectively;

[0199] The number of newly charged electric vehicles in the kth charging station at time t can be expressed as:

[0200] K t,k =min{Q t-1,k +D t,k ,I t,k} (28)

[0201] The queuing time of electric vehicles at the kth charging station at time t can be obtained from formulas (24)-(28):

[0202]

[0203] Where K q is the queue impact coefficient;

[0204] After taking into account the queuing time, the total time the user spends traveling to the charging station is:

[0205]

[0206] Where, is the waiting time of electric vehicles at the kth charging station and time t;

[0207] The constraints of the multi-objective planning model of the charging station can be expressed as:

[0208] 1) Upper and lower limit constraints of node amplitude:

[0209] U min ≤U x ≤U max (31)

[0210] Where U x is the node voltage at node x, U max and U min are its upper and lower limits respectively;

[0211] 2) Branch flow constraints:

[0212]

[0213] Where: P x,t and Q x,t are the active power and reactive power of node x at time t; U x,t and U y,t are the voltage amplitudes of nodes x and y at time t; G xy and B xy are the conductance and susceptance of branch xy respectively; θ xy is the phase angle difference between the voltages at nodes x and y; Ω D is the node set of the power grid;

[0214] 3) Planning the number of charging piles and transformer capacity constraints within the charging station:

[0215]

[0216] Where, and To plan the upper and lower limits of the number of charging piles in the charging station;

[0217] 4) Planning the number of charging stations:

[0218]

[0219]

[0220] Where, and The planned number of charging stations is M CS The upper and lower limits of N total is the total number of electric vehicles in the planning area; Q is the battery capacity; T av is the average charging time; Indicates rounding down;

[0221] 5) Voltage deviation constraint:

[0222]

[0223] Where U0 is the reference voltage value; f vd is the voltage fluctuation rate; is the maximum volatility constraint;

[0224] 6) Distance constraints between charging stations:

[0225]

[0226] Where, and Represents the distance between any two charging stations Maximum and minimum constraints;

[0227] 7) Substation main transformer load rate constraints:

[0228]

[0229] Where, is the basic power load at the grid node x to which the kth charging station is connected; represents the electric vehicle charging load at the grid node x to which the kth charging station is connected; Configure the rated capacity of the transformer for the kth charging station; α max To plan the upper limit of transformer load rate;

[0230] 8) Transformer capacity constraints:

[0231]

[0232] Where, and To plan the upper and lower limits of transformer capacity.

[0233] Compared with the prior art, the present invention has the following technical effects:

[0234] 1) Compared with traditional approaches that only consider the impact of real-time traffic conditions, power grid flow distribution, and station queues on charging station planning and layout, this method builds a more comprehensive model and is more suitable for solving the current charging station planning problem dominated by power grid companies;

[0235] 2) Compared with the traditional method of using Voronoi diagrams to divide charging station service areas, the load clustering method proposed in this invention can minimize the distance between the actual edge nodes and the central node, and fully consider the impact of the real-time traffic network on user travel. This can not only reduce the time spent by electric vehicle users charging, but also improve the utilization rate of planned charging equipment, thereby improving the economic efficiency of charging station planning;

[0236] 3) To address the impact of balanced charging load distribution within a charging station's service area on the power network, this invention incorporates load balancing into the charging station planning process. By calculating revenue and factoring in network losses and transformer costs, this effectively reduces grid losses, lowers voltage fluctuations, and improves overall grid reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0237] The present invention will be further described below with reference to the accompanying drawings and examples:

[0238] Figure 1 This is a schematic diagram of the network topology based on hierarchical road-network-station road-network-station coupling in the present invention;

[0239] Figure 2 This is a flow chart of the immune genetic algorithm in the present invention;

[0240] Figure 3 Schematic diagram of time-sharing electricity purchase and sales prices for a charging station in an embodiment of the present invention;

[0241] Figure 4 1. This is an architecture diagram of a road-power-station coupling network according to an embodiment of the present invention;

[0242] Figure 5 A schematic diagram of the service area division of a charging network in an embodiment of the present invention;

[0243] Figure 6 Schematic diagram showing the comparison of the variance of the number of electric vehicles arriving at the charging station per hour in methods 1 and 2 in an embodiment of the present invention;

[0244] Figure 7 Schematic diagram of the division of the charging station service area according to method 1 in an embodiment of the present invention;

[0245] Figure 8 Schematic diagram of the division of the charging station service area according to method 2 in an embodiment of the present invention;

[0246] Figure 9 A comparison chart of queue lengths for methods 1 and 2 in an embodiment of the present invention;

[0247] Figure 10 Schematic diagram of node voltage fluctuation rate in methods 2 and 3 in an embodiment of the present invention. DETAILED DESCRIPTION

[0248] A charging station planning method considering a multi-coupling network of "road-network-station" includes the following steps:

[0249] Step 1: Build a charging station planning model considering the “road-network-station” coupling network

[0250] Step 1.1: About the "Road-Network-Station" coupling network topology

[0251] In order to further analyze the interaction mechanism between the complex and ever-changing transportation network, charging stations and power network, this patent introduces the graph theory method to construct the "road-network-station" coupling network topology to further express the multi-layer coupling relationship between the three. Figure 1 shown.

[0252] The topology of the traffic network is shown in formula (1):

[0253]

[0254] Where GL is the road network topology; V L It is the traffic node in the road network structure, that is, the intersection node of the road; E L Indicates the road network edge; is the adjacency matrix of the road network, which is used to describe the connection relationship between traffic nodes and the road length; n L Indicates the total number of traffic nodes; and represent the i-th and j-th traffic nodes respectively; Representation node and connected roads; e ij is the length of the road section, and the specific expression is shown in formula (2).

[0255]

[0256] Where, d ij Indicates road section The length of ; inf means infinity.

[0257] It can be expressed by formula (3):

[0258]

[0259] This patent simplifies the structure of the power grid and only considers the connection edges and nodes of the power grid. Its topological structure can be expressed by formula (4):

[0260]

[0261] Where G D is the grid topology; V D is the node set of the power grid; E D is the set of grid edges; is the parameter matrix of the relevant information of the power grid branch; v G Indicates the total number of grid nodes; Representation node and The connected power grid line; xy , z xy , c xy and They respectively represent the resistance, reactance, susceptance of the power grid branch xy and the transmission power limit of the branch.

[0262] The topology of the power station layer is shown in formula (5):

[0263]

[0264] Where G C is the charging station topology; VC A set of nodes representing charging stations; Represents the relevant information parameter matrix of the charging station; represents the number of the kth charging station; v C Indicates the total number of charging stations; is the number of charging piles in the kth charging station; Indicates the capacity of the newly built supporting transformer; represents the land price and labor costs at the location of the k-th charging station.

[0265] The coupling network also includes the interconnection relationship between the “road-network-station” level, and its specific expression is shown in formula (6):

[0266]

[0267] Where G represents the “road-network-station” coupling topology; E L-D-C is the coupling network edge, which represents the coupling between the road network node, the power grid node and the charging station access node; is the coupling network edge that is not connected to the charging station, that is, the coupling between the road network node and the power grid node; It is the coupling network edge connected to the charging station, that is, the coupling of the three nodes of "road-network-station".

[0268] Step 2: Build an electric vehicle user charging decision model

[0269] This patent simplifies the distribution of electric vehicles to dispersed at the intersection nodes of the road network. It calculates the total time it takes for each node to reach each charging station in a day, and determines which charging station service area the node's charging load belongs to with the goal of minimizing the total time. Its objective function is shown in formula (7):

[0270]

[0271] Where, is the total time spent on the road from node i to the kth charging station; η is the decision variable; is the time it takes for the electric vehicle at node i to reach the kth charging station at time t; N i,t is the number of electric vehicles at node i at time t; Ω t is the time set; Ω i is a set of road nodes.

[0272]

[0273] Where, L i,k,q is the road section q that the electric vehicle at node i passes through to the kth charging station; v q (t) is the speed of the electric vehicle at the road section q at time t; Ωq is the set of paths taken by the electric vehicle.

[0274] v q (t) can be calculated by formula (9):

[0275]

[0276] Where, represents the average driving speed of road section q; ρ q (t) is the traffic flow density of the qth road section at time t; C q,max represents the maximum traffic flow of the qth road section; a, b and μ are the adaptive coefficients of the road, which can be obtained by fitting the traffic road data

[54] ; N q (t) represents the number of electric vehicles passing the qth road section at time t; θ is the market penetration rate of electric vehicles.

[0277] Finally, nodes with the same decision-making plan are assigned to the same charging station service area. The specific charging demand distribution of charging stations is shown in formulas (10)-(11):

[0278]

[0279]

[0280] Among them, D t,k represents the charging demand of the kth charging station at time t.

[0281] The constraints of the electric vehicle user charging decision model can be expressed as:

[0282] 1) Driving distance constraint:

[0283] L i,k ≤L max (12)

[0284] Where, L max is the maximum distance from an electric vehicle to a charging station.

[0285] 2) Queuing waiting time constraints:

[0286]

[0287] Where, The maximum waiting time for electric vehicles at charging stations.

[0288] Step 3: Charging station planning model considering load balancing effect

[0289] The solution to this problem proposed in this patent is to not only consider the location and capacity of the charging station during planning, but also consider adding corresponding supporting facilities such as the configuration of the transformer, fully taking into account the investment cost, operating cost, loss cost and network loss cost of the transformer, and planning the charging station with the goal of maximizing profits. In addition, considering that if the queues at the charging station are too long, it will cause users to flow to other charging stations or go home to charge, this patent also takes into account the profits caused by user loss due to long queues, and optimizes the planning decision-making plan with a market-oriented approach.

[0290] The objective function of the charging station planning model is:

[0291] maxF1=C S -C I,CS -C O,CS -C I,TR -C O,TR -C L,G -C L,TR (14)

[0292] Where C S is the planned total revenue of the charging station; C I,CS is the total planned investment cost of the charging station; C O,CS is the total operation and maintenance cost of the charging station; C I,TR The investment cost of the transformer for the charging station; C O,TR The operation and maintenance cost of the supporting transformer; C L,G is the network loss cost of the distribution network; C L,TR is the loss cost of the transformer.

[0293] The total revenue from the charging station plan is:

[0294]

[0295] Where, The time-of-use electricity price for charging stations; The time-of-use electricity purchase price for the charging station from the power grid company; B t,k is the number of charging piles that are busy at the kth charging station at the tth moment; P S is the rated power of the charging pile.

[0296] The total planned investment cost of the charging station is:

[0297]

[0298] Where, is the area occupied by the charging station, which is directly proportional to the number of charging piles in the charging station; The land price; Planning capacity for charging stations; is the unit price of the charging pile; r0 is the discount rate; T is the planned service life.

[0299] The total operation and maintenance cost of the charging station is:

[0300]

[0301] Where C M,CS It is the operation and maintenance cost of unit charging equipment.

[0302] The investment cost of the transformer for the charging station is:

[0303]

[0304]

[0305] Where, The planned capacity of the transformer supporting the charging station; The unit price of its corresponding capacity; the transformer capacity planning should be adapted to the number of charging equipment in the charging station, and Plan the capacity of the transformer corresponding to the number of charging piles.

[0306] The operation and maintenance cost of the supporting transformer is:

[0307]

[0308] Where C M,TR is the unit operation and maintenance cost of the transformer; T max The maximum annual utilization hours.

[0309] The loss cost of the transformer is:

[0310]

[0311] Where c TR is the unit loss cost of the transformer; L fe and L cu are the rated iron loss and copper loss of the transformer respectively.

[0312] The network loss cost of the distribution network is:

[0313]

[0314] Where c loss is the unit network loss cost of the distribution network; is the line loss of the lth line of the power grid; Ω l A collection of lines.

[0315] The number of charging piles at a charging station is configured as shown in formula (23):

[0316]

[0317] Where, The maximum number of charging piles planned for the charging station is The minimum value of the maximum number of electric vehicles arriving at the service area within a day.

[0318] The number of electric vehicles queuing for charging at the kth charging station at time t can be obtained by formula (24):

[0319] Q t,k =min{Q t-1,k +D t,k -I t,k ,0} (24)

[0320] Where Q t-1,k is the number of electric vehicles queuing for charging at the kth charging station at time t-1; I t,k represents the number of charging piles that are not busy at the kth charging station at time t, which can be calculated by formula (25).

[0321]

[0322] Where B t,k and M t,k They represent the number of charging piles in the busy state and the state of charging completion at the kth charging station at time t.

[0323] B tk and M tk It can be calculated by formula (26) and (27) respectively:

[0324] B t,k =B t-1,k -M t-1,k +K t-1,k (26)

[0325] M t,k =K t-2,k (27)

[0326] Where B t-1,k and M t-1,k K represents the number of charging piles that are busy and about to be charged at the kth charging station at time t-1; t-1,k and K t-2,k They represent the number of electric vehicles newly charged at the k-th charging station at time t-1 and time t-2 respectively.

[0327] The number of newly charged electric vehicles in the kth charging station at time t can be expressed as:

[0328] K t,k =min{Q t-1,k +D t,k ,I t,k} (28)

[0329] The queuing time of electric vehicles at the kth charging station at time t can be obtained from formulas (24)-(28):

[0330]

[0331] Where K q is the queue impact coefficient.

[0332] After taking into account the queuing time, the total time the user spends traveling to the charging station is:

[0333]

[0334] Where, is the waiting time of electric vehicles at the kth charging station and time t.

[0335] The constraints of the charging station planning model can be expressed as:

[0336] 1) Upper and lower limit constraints of node amplitude:

[0337] U min ≤U x ≤U max (31)

[0338] Where U x is the node voltage at node x, U max and U min are its upper and lower limits respectively.

[0339] 2) Branch flow constraints:

[0340]

[0341] Where: P x,t and Q x,t are the active power and reactive power of node x at time t; U x,t and U y,t are the voltage amplitudes of nodes x and y at time t; G xy and B xy are the conductance and susceptance of branch xy respectively; θ xy is the phase angle difference between the voltages at nodes x and y; Ω D is the node set of the power grid.

[0342] 3) Planning the number of charging piles and transformer capacity constraints within the charging station:

[0343]

[0344] Where, and To plan the upper and lower limits of the number of charging piles in a charging station.

[0345] 4) Planning the number of charging stations:

[0346]

[0347]

[0348] Where, and The planned number of charging stations is M CS The upper and lower limits of N total is the total number of electric vehicles in the planning area; Q is the battery capacity; T av is the average charging time; Indicates rounding down.

[0349] 5) Voltage deviation constraint:

[0350]

[0351] Where U0 is the reference voltage value; f vd is the voltage fluctuation rate; is the maximum volatility constraint.

[0352] 6) Distance constraints between charging stations:

[0353]

[0354] Where, and Represents the distance between any two charging stations The maximum and minimum constraints.

[0355] 7) Substation main transformer load rate constraints:

[0356]

[0357] Where, is the basic power load at the grid node x to which the kth charging station is connected; represents the electric vehicle charging load at the grid node x to which the kth charging station is connected; Configure the rated capacity of the transformer for the kth charging station; α max It is the upper limit of the planned transformer load rate.

[0358] 8) Transformer capacity constraints:

[0359]

[0360] Where, and To plan the upper and lower limits of transformer capacity

[0361] Step 4: Use the improved immune genetic algorithm to solve the constructed charging station multi-objective programming model as follows:

[0362] The immune genetic algorithm is an intelligent algorithm that combines the immune theory of biology with the genetic algorithm. While retaining the advantages of fast convergence and strong robustness of the traditional genetic algorithm, the algorithm extracts a "vaccine" to effectively suppress the degradation phenomenon that occurs during the iteration process, thereby constructing a more adaptable and selective information defense system. While ensuring the continuous optimization and iteration of the population, it reduces the occurrence of invalid search events in the genetic algorithm and ensures the validity of the iterative information. The specific algorithm process is as follows Figure 2 shown.

[0363] Step 1: Initialize parameters, including antibody population size, number of iterations, crossover probability, mutation probability, and other related parameters;

[0364] Step 2: Generate initial antibody population and memory library;

[0365] Step 3: Calculate individual fitness. Substitute the corresponding parameters into the objective function to obtain the corresponding F1 and F2. Considering that the revenue calculation of the charging station is to maximize the revenue, the goal of the electric vehicle user is to minimize the time spent on charging. In order to facilitate the comparison of the advantages and disadvantages of different planning schemes, this patent defines a global variable operator F 1,max and F 2,min , used to normalize the multi-objective values. After normalizing F1 and F2, the global fitness can be calculated using formula (40):

[0366]

[0367] Where ω1 and ω2 are weights.

[0368] Step 4: Calculate the affinity between individual antibodies and antibody concentration. u,v Indicates the similarity between antibodies u and v. Antibody concentration Con indicates the overall similarity of the antibody population.

[0369]

[0370]

[0371] Where, L u,vrepresents the length of the antibody with the same value as u and v; ls represents the total length of the antibody individual; and m is the total number of the antibody population.

[0372] Step 5: Determine whether the antibody population meets the conditions. Determine whether the antibody concentration of the antibody population is equal to 1. If so, obtain the planning result and terminate the loop; if not, continue with steps 6-8;

[0373] Step 6: Select an operation. Select the best individuals from the exchanged individual antibodies and use the "roulette wheel" method to select individuals from the antibody population to generate a new population.

[0374] Step 7: Crossover operation: Two antibody individuals are randomly selected from the antibody population according to a certain exchange probability to generate new individuals.

[0375] Step 8: Select antibody individuals according to a certain probability and randomly change the value of a gene in the string structure data for the selected individuals to generate new individuals. Return to step 3 for a new round of iteration.

[0376] The present invention also includes a method for constructing an electric vehicle user charging decision model, which includes the following steps:

[0377] Step S1: Simplify the distribution of electric vehicles to dispersed at the intersection nodes of the road network, calculate the total time it takes for each node to reach each charging station in a day, and decide which charging station service area the charging load of the node belongs to with the goal of minimizing the total time;

[0378] Step S2: Nodes with the same decision-making plan are assigned to the same charging station service area.

[0379] In step S1, the objective function is as shown in formula (7):

[0380]

[0381] Where, is the total time spent on the road from node i to the kth charging station; η is the decision variable; is the time it takes for the electric vehicle at node i to reach the kth charging station at time t; N i,t is the number of electric vehicles at node i at time t; Ω t is the time set; Ω i is a set of road nodes;

[0382]

[0383] Where, L i,k,q is the road section q that the electric vehicle at node i passes through to the kth charging station; v q(t) is the speed of the electric vehicle at the road section q at time t; Ω q is the set of paths passed by the electric vehicle;

[0384] v q (t) can be calculated by formula (9):

[0385]

[0386] Where, represents the average driving speed of road section q; ρ q (t) is the traffic flow density of the qth road section at time t; C q,max represents the maximum traffic flow of the qth road segment; a, b and μ are the adaptive coefficients of the road; N q (t) represents the number of electric vehicles passing the qth road segment at time t; θ is the market penetration rate of electric vehicles;

[0387] In step S2, nodes with the same decision-making scheme are assigned to the same charging station service area. The specific charging demand distribution of the charging station is shown in formulas (10)-(11):

[0388]

[0389]

[0390] Among them, D t,k represents the charging demand of the kth charging station at time t;

[0391] The constraints of the electric vehicle user charging decision model can be expressed as:

[0392] 1) Driving distance constraint:

[0393] L i,k ≤L max (12)

[0394] Where, L max is the maximum distance between the electric vehicle and the charging station;

[0395] 2) Queuing waiting time constraints:

[0396]

[0397] Where, The maximum waiting time for electric vehicles at charging stations.

[0398] The present invention also includes a method for constructing a multi-objective planning model for a charging station taking into account the load balancing effect.

[0399] The objective function of the multi-objective planning model of the charging station is:

[0400] maxF1=C S -C I,CS -C O,CS -C I,TR -C O,TR -C L,G -C L,TR (14)

[0401] Where C S is the planned total revenue of the charging station; C I,CS is the total planned investment cost of the charging station; C O,CS is the total operation and maintenance cost of the charging station; C I,TR The investment cost of the transformer for the charging station; C O,TR The operation and maintenance cost of the supporting transformer; C L,G is the network loss cost of the distribution network; C L,TR is the loss cost of the transformer;

[0402] The total revenue from the charging station plan is:

[0403]

[0404] Where, The time-of-use electricity price for charging stations; The time-of-use electricity purchase price for the charging station from the power grid company; B t,k is the number of charging piles that are busy at the kth charging station at the tth moment; P S is the rated power of the charging pile;

[0405] The total planned investment cost of the charging station is:

[0406]

[0407] Where, is the area occupied by the charging station, which is directly proportional to the number of charging piles in the charging station; The land price; Planning capacity for charging stations; is the unit price of the charging pile; r0 is the discount rate; T is the planned service life;

[0408] The total operation and maintenance cost of the charging station is:

[0409]

[0410] Where C M,CS The unit charging equipment operation and maintenance cost;

[0411] The investment cost of the transformer for the charging station is:

[0412]

[0413]

[0414] Where, The planned capacity of the transformer supporting the charging station; The unit price of its corresponding capacity; the transformer capacity planning should be adapted to the number of charging equipment in the charging station, and Plan the capacity of the transformer for the number of charging piles;

[0415] The operation and maintenance cost of the supporting transformer is:

[0416]

[0417] Where C M,TR is the unit operation and maintenance cost of the transformer; T max is the maximum annual utilization hours;

[0418] The loss cost of the transformer is:

[0419]

[0420] Where c TR is the unit loss cost of the transformer; L fe and L cu are the rated iron loss and copper loss of the transformer respectively;

[0421] The network loss cost of the distribution network is:

[0422]

[0423] Where c loss is the unit network loss cost of the distribution network; is the line loss of the lth line of the power grid; Ω l is a collection of lines;

[0424] The number of charging piles at a charging station is configured as shown in formula (23):

[0425]

[0426] Where, The maximum number of charging piles planned for the charging station is The minimum value of the maximum number of electric vehicles arriving at the service area in one day;

[0427] The number of electric vehicles queuing for charging at the kth charging station at time t can be obtained by formula (24):

[0428] Q t,k =min{Qt-1,k +D t,k -I t,k ,0} (24)

[0429] Where Q t-1,k is the number of electric vehicles queuing for charging at the kth charging station at time t-1; I t,k represents the number of charging piles that are not busy at the kth charging station at time t, which can be calculated by formula (25);

[0430]

[0431] Where B t,k and M t,k They represent the number of charging piles in the busy state and the state of charging completion at the kth charging station at time t respectively;

[0432] B t,k and M t,k It can be calculated by formula (26) and (27) respectively:

[0433] B t,k =B t-1,k -M t-1,k +K t-1,k (26)

[0434] M t,k =K t-2,k (27)

[0435] Where B t-1,k and M t-1,k K represents the number of charging piles that are busy and about to be charged at the kth charging station at time t-1; t-1,k and K t-2,k They represent the number of newly charged electric vehicles at the k-th charging station at time t-1 and time t-2 respectively;

[0436] The number of newly charged electric vehicles in the kth charging station at time t can be expressed as:

[0437] K t,k =min{Q t-1,k +D t,k ,I t,k} (28)

[0438] The queuing time of electric vehicles at the kth charging station at time t can be obtained from formulas (24)-(28):

[0439]

[0440] Where K q is the queue impact coefficient;

[0441] After taking into account the queuing time, the total time the user spends traveling to the charging station is:

[0442]

[0443] Where, is the waiting time of electric vehicles at the kth charging station and time t;

[0444] The constraints of the multi-objective planning model of the charging station can be expressed as:

[0445] 1) Upper and lower limit constraints of node amplitude:

[0446] U min ≤U x ≤U max (31)

[0447] Where U x is the node voltage at node x, U max and U min are its upper and lower limits respectively;

[0448] 2) Branch flow constraints:

[0449]

[0450] Where: P x,t and Q x,t are the active power and reactive power of node x at time t; U x,t and U y,t are the voltage amplitudes of nodes x and y at time t; G xy and B xy are the conductance and susceptance of branch xy respectively; θ xy is the phase angle difference between the voltages at nodes x and y; Ω D is the node set of the power grid;

[0451] 3) Planning the number of charging piles and transformer capacity constraints within the charging station:

[0452]

[0453] Where, and To plan the upper and lower limits of the number of charging piles in the charging station;

[0454] 4) Planning the number of charging stations:

[0455]

[0456]

[0457] Where, and The planned number of charging stations is M CS The upper and lower limits of N total is the total number of electric vehicles in the planning area; Q is the battery capacity; T av is the average charging time; Indicates rounding down;

[0458] 5) Voltage deviation constraint:

[0459]

[0460] Where U0 is the reference voltage value; f vd is the voltage fluctuation rate; is the maximum volatility constraint;

[0461] 6) Distance constraints between charging stations:

[0462]

[0463] Where, and Represents the distance between any two charging stations Maximum and minimum constraints;

[0464] 7) Substation main transformer load rate constraints:

[0465]

[0466] Where, is the basic power load at the grid node x to which the kth charging station is connected; represents the electric vehicle charging load at the grid node x to which the kth charging station is connected; Configure the rated capacity of the transformer for the kth charging station; α max To plan the upper limit of transformer load rate;

[0467] 8) Transformer capacity constraints:

[0468]

[0469] Where, and To plan the upper and lower limits of transformer capacity.

[0470] Example

[0471] This patent adopts a traffic network consisting of 51 roads and 33 traffic nodes. The specific road topology length data is shown in Table 1. It is divided into commercial areas, work areas and living areas according to the specific urban functional structure. Referring to the land prices of Shanghai functional areas, the land prices of the corresponding planning areas are 120,000, 80,000 and 60,000 yuan / m2 respectively. 2 The grid structure is composed of IEEE-33 power-saving system. There are 2,000 electric vehicles in the planning area, with an average speed of 30km / h. The market penetration rate of electric vehicles is 10%. The rated power of the planned charging pile is 30kw / h, the planned service life is 10 years, the discount rate is 0.1, the unit price of the charging pile is 10,000 yuan / unit, and the time-of-use electricity purchase and sales price of the charging station is as follows: Figure 3 The corresponding table of transformer capacity and charging pile planning is shown in Table 1. The node and line information parameters of the distribution network are shown in Tables 2 and 3.

[0472] Table 1 Correspondence between transformer planning capacity and planned number of charging piles

[0473]

[0474] In order to verify the effectiveness of the method proposed in this chapter, three methods were used for verification and their calculation results were compared and analyzed.

[0475] Method 1: Using the Voronoi diagram method for planning, but without considering the balancing effect of electric vehicle charging load;

[0476] Method 2: Use load clustering method, but do not consider the balancing effect of electric vehicle charging load;

[0477] Method 3: Use load clustering method and consider the balancing effect of electric vehicle charging load.

[0478] The software platforms used to solve the simulations by the above methods are all implemented in the Matlab 2019 environment. The operating environment is: main frequency 3.20 GHz, memory 8.0 GB, and Intel Core i5 processor. The mpt toolkit of Voronoi diagram is used to verify method 1.

[0479] 2) Example simulation results and analysis

[0480] Combining the number of electric vehicles and urban traffic network information, and according to formulas (33)-(34), it is planned to build 5-8 charging stations in the area, and each charging station contains a maximum of 50 charging piles. Taking methods 2 and 3 as examples, by solving the model constructed by this patent, the corresponding relationship table between the number of charging stations, their benefits and overall fitness can be obtained as shown in Table 2.

[0481] Table 2 Trends in the number of charging stations, their benefits, and overall fitness

[0482]

[0483] As shown in Table 2, when planning six charging stations, the total revenue of the charging stations in Methods 2 and 3 reaches its maximum, and the overall fitness also reaches its minimum at this time. As the number of charging stations increases, the investment and operating costs of the charging stations also increase, which in turn increases the total revenue and overall fitness of the charging stations. This patent uses the example of a power grid company guiding the planning of six charging stations to conduct a simulation comparison analysis. The simulation comparison results of Methods 1, 2, and 3 are shown in Table 4.

[0484] Table 3 Simulation comparison results of different methods

[0485]

[0486] As shown in Table 3, the patent method selects the traffic nodes 10, 24, 19, 6, 3 and 21 to plan and build charging stations, and the corresponding number of charging piles is 43, 50, 50, 29, 50 and 50 respectively. Figure 4 The road network-grid coupling node connection comparison table is shown in Table 4.

[0487] Depend on Figure 4 It can be seen that the "road-electricity-station" coupling network obtained by the method proposed in this patent plans new charging stations at nodes 10, 24, 19, 6, 3 and 21, which are coupled with nodes 10, 29, 19, 24, 2 and 33 of the power grid respectively. In the road network structure, this patent classifies road network nodes 31, 32, 1, 2, 3, 4, 6, 11, 16, 21, 25, 33, 26 and 29 as residential areas, divides road network nodes 5, 10, 15, 20, 24, 28, 30 and 27 into working areas, and divides road network nodes 9, 7, 8, 12, 13, 14, 17, 18, 19, 22 and 23 into commercial areas. Based on the proposed load clustering method, this patent divides all nodes with the same charging decision into the same service area. The charging service zoning diagram of the charging station based on the method of this patent is as follows Figure 5 shown.

[0488] Depend on Figure 5It can be seen that the method of this patent divides road network traffic nodes 5, 9, 10 and 15 into charging station No. 1, divides road network traffic nodes 23, 24, 27, 28 and 30 into charging station No. 2, divides road network traffic nodes 13, 14, 17, 18, 19 and 20 into charging station No. 3, divides road network traffic nodes 6, 7, 11 and 12 into charging station No. 4, divides road network traffic nodes 1, 2, 3, 4, 8, 31 and 32 into charging station No. 5, and divides road network traffic nodes 16, 21, 22, 25, 26, 29 and 33 into charging station No. 6.

[0489] 1) Analysis of the effectiveness of considering load clustering method in charging station planning.

[0490] This patent analyzes EV charging station planning from both the user and charging station perspectives. Table 4 shows the impact of optimizing charging station layouts using methods 1 and 2 on EV users' time costs, taking into account real-time traffic conditions.

[0491] Table 4 Time cost of electric vehicle users in methods 1 and 2

[0492]

[0493] As shown in Table 4, compared with method 1, the total road travel time and total queuing time of electric vehicle users in one day in method 2 are reduced by 19.24% and 90.98% respectively. The main reason is that: the method of dividing the service area using the Voronoi diagram in method 1 is to calculate the Euclidean distance between the edge node and the center point through geometric principles, and take this minimum as the goal, without considering the actual distance and real-time road conditions of the complex traffic network. Unlike method 1, method 2 divides the service area with the goal of minimizing the actual road time spent by each edge node to the center node, so that the road travel time spent by electric vehicles to the charging station for charging is reduced. The total queuing time is greatly reduced. The reason is that if the service area is divided based on Euclidean distance only from a geometric perspective, it may lead to the marginalization of nodes farther away after the node-dense areas are clustered into service areas, reducing the balance of charging loads contained in each service area. Further analysis shows that the mean square deviation of the number of electric vehicles arriving at the charging station each hour is as follows: Figure 6 As shown, the charging service area division diagram of the charging station of Method 1 and Method 2 is as follows Figure 7 and 8 shown.

[0494] Depend on Figure 6 It can be seen that in terms of the mean square error of the number of electric vehicles arriving at the charging station every hour, the mean square error of method 2 is smaller than that of method 1, which means that the charging demand distribution in the charging service area of ​​the charging station divided by method 2 is more uniform. Figure 7 and8 It can be seen that the area divided by No. 2 and No. 4 in Method 1 is the largest, containing a total of 18 traffic nodes, accounting for 54.54% of the total number of nodes, and their load demand accounts for 58.85% of the total demand. In Method 2, the largest charging service area is No. 3 and No. 4 charging stations, and their number of nodes and load demand account for 42.42% and 39.04% of the total number of nodes and charging demand respectively. If the high proportion of charging demand cannot be converted into charging service volume, it will inevitably lead to long queues and the loss of electric vehicle users. The comparison of the number of queues in Methods 1 and 2 is as follows: Figure 9 shown.

[0495] Depend on Figure 9 It can be seen that the longest queue time is concentrated between 18:00 and 23:00. Among them, the charging stations with the longest charging queues in method 1 are concentrated in charging stations 2 and 4. This shows that the charging demand in the designated charging service areas 2 and 4 accounts for too much, far exceeding their maximum charging service capacity. The charging service areas in method 2 are divided more evenly, effectively avoiding the situation where users queue too concentratedly, and are more conducive to charging services at charging stations.

[0496] The impact of using methods 1 and 2 to optimize the layout of charging stations on the calculation of various costs and benefits of charging stations is shown in Table 5.

[0497] Table 5 Calculation of various costs and benefits of methods 1 and 2

[0498]

[0499] Table 5 shows that compared to Method 1, Method 2's annual total revenue increased by 27.06%. This is because Method 2's more evenly divided charging service area results in greater utilization of charging equipment and a larger number of electric vehicles receiving charging services daily, which in turn increases total revenue. In contrast, Method 1, due to queue time constraints, reduces the number of electric vehicles receiving services per day by 17.14% compared to Method 2, resulting in a decrease in annual total revenue of 1.8822 million yuan. Method 2's annual investment and operating costs increased by 1.1832 million yuan and 94,700 yuan, respectively, compared to Method 1. This is because Method 2, taking into account charging demand within its own charging service area, plans a 27% increase in the number of charging stations. Furthermore, Method 2 plans three charging stations in work areas and only two in residential areas, while Method 1 plans four charging stations, mostly concentrated in residential areas where land prices are relatively affordable. High investment costs mean high operating and maintenance costs, making Method 2's annual investment and operating costs higher than those of Method 1. However, the increase in the annual total income of Method 2 is greater than the increase in its cost, so the annual total income of Method 2 increases by 16.36% compared with Method 1.

[0500] It can be seen that the load clustering method used in Method 2 can effectively divide the service area of ​​the charging station, which can not only significantly reduce the time spent on the road and in queues by electric vehicle users to reach the charging station, but also effectively improve the overall service capacity of the charging station and improve the economic efficiency of its operation.

[0501] 2) Analysis of the effectiveness of considering the balancing effect of electric vehicle charging load in charging station planning.

[0502] After considering the balancing effect of charging loads, charging stations must plan and construct supporting facilities to accommodate the capacity configuration of the charging equipment. This patent only considers the supporting investment and construction of transformers during the charging station planning process, thus requiring consideration of network loss costs as well as the investment, operating, and loss costs of transformers. The impact of Methods 2 and 3 on the time cost of electric vehicle users after considering the balancing effect of electric vehicle charging loads is shown in Table 6.

[0503] Table 6 Time cost of electric vehicle users in methods 2 and 3

[0504]

[0505] Table 6 shows that, compared with Method 2, Method 3 increases and decreases the total waiting time and travel time for electric vehicle users by 61.14 and 22.57 hours, respectively. This is primarily due to the fact that, considering the impact of centralized electric vehicle charging on the overall operation of the power grid, Method 3 reduces the total number of charging piles by 5.51% compared to Method 2 in order to maintain load balancing between regions and reduce network losses. This reduced number of charging piles leads to supply shortages during certain periods, thus increasing the average and total waiting time. However, because the charging pile layout in Method 3 reduces the travel time for users to charge, the total time spent in Method 3 is only 6.91% higher than that in Method 2.

[0506] The impact of using methods 2 and 3 to optimize the layout of charging stations on the calculation of various costs and benefits of charging stations is shown in Table 7.

[0507] Table 7 Calculation of various costs and benefits of Methods 2 and 3

[0508]

[0509]

[0510] As can be seen from Table 7, the annual total income of Method 3 is not much different from that of Method 2, with a difference of only 72,270 yuan. This is because the number of electric vehicles served by the two methods is almost the same. Compared with Method 2, the annual investment cost and operating cost of the charging station and the investment cost and operating cost of the transformer of Method 3 are reduced by 7.47%, 7%, 1.96% and 1.9% respectively. The main reason is that Method 2 does not plan and decide the number of charging piles from the perspective of the overall operation of the power grid, but only optimizes the time spent by users on charging and the maximum profit of the charging station. More charging facilities are planned at some nodes, which leads to an increase in the annual investment cost and operating cost of its charging stations. The large-scale investment and construction of charging facilities also requires the planning of larger capacity transformers, which in turn increases the investment cost and operating cost of the transformer. The impact of the two methods on the voltage fluctuation rate of the power grid is shown in the figure. Figure 10 shown.

[0511] like Figure 10 As shown, Method 3 reduces node voltage fluctuations compared to Method 2, with an average voltage fluctuation reduction of 45.9%. This is primarily due to the fact that Method 2 ignores the impact of centralized EV charging on the overall power flow distribution of the grid. It plans and constructs charging stations with increased charging capacity at nodes with high voltage fluctuations, resulting in severe voltage overshoots at some nodes. This increases Method 2's annual total network loss cost by 45.82%, further increasing Method 3's total revenue by 1.1221 million yuan compared to Method 2.

[0512] This patent calculates the total fitness of the results obtained by the two methods based on formula (40), using ω1 = 0.5 and the maximum benefit F 1,max =600, the total charging time consumed by the user is F 2,max =3000, the total fitness value of method 2 is 0.832, and that of method 3 is 0.486. This shows that although method 3 slightly increases the charging time of electric vehicle users, from the perspective of the overall operation of the power grid, planning and decision-making can reduce planning investment costs and network losses while improving the economy and safety of its planning.

Claims

1. A charging station planning method considering the "road-network-station" network and load, characterized by: It includes the following steps: Step 1: Consider the impact of traffic conditions, overall grid operation, and charging station queues on planning results and construct a road-grid-station coupled network topology; Step 2: Build a charging decision model for electric vehicle users based on the speed-flow model and charging load distribution prediction results; Step 3: Construct a multi-objective planning model for charging stations considering the load balancing effect; Step 4: Use the improved immune genetic algorithm to solve the constructed multi-objective programming model of the charging station; In step 2, when building the charging decision model for electric vehicle users, the following steps are adopted: Step 2-1: Simplify the distribution of electric vehicles to dispersed at the intersection nodes of the road network. Calculate the total time it takes for each node to reach each charging station in a day. With the goal of minimizing the total time, decide which charging station service area the charging load of the node belongs to. Step 2-2: Nodes with the same decision-making plan are assigned to the same charging station service area; In step 2-1, the objective function is as shown in formula (7): （7）； Where, For nodes i To k The total time spent on the road at each charging station; is the decision variable; for t Time Node i Electric vehicles from k Time spent at each charging station; for t Time Node i the number of electric vehicles at the site; is the time collection; is a set of road nodes; （8）； Where, For nodes i Electric vehicles from k Road sections passed by charging stations q ; For electric vehicles t Road sections passed at the time q Driving speed at 100 km / h; is the set of paths passed by the electric vehicle; It can be calculated by formula (9): （9）； Where, Indicates road section q average driving speed; For the q road sections in t Traffic flow density at the time; Indicates the q The maximum traffic flow of each road section; , b and is the adaptive coefficient of the road; Indicates the q road sections in t The number of electric vehicles passing by at any given moment; The market penetration rate of electric vehicles; In step 2-2, nodes with the same decision-making plan are assigned to the same charging station service area. The specific charging demand distribution of charging stations is shown in formulas (10)-(11): （10）； （11）； in, Indicates the k Charging stations t Constant charging needs; The constraints of the electric vehicle user charging decision model can be expressed as: 1) Driving distance constraint: （12）； Where, is the maximum distance between the electric vehicle and the charging station; 2) Queue waiting time constraints: （13）； Where, The maximum waiting time for electric vehicles at charging stations.

2. The method according to claim 1, characterized in that In step 1, when constructing the road-network-station coupling network topology, the details are as follows: The topology of the traffic network is shown in formula (1): （1）； Where, is the road network topology; It is a traffic node in the road network structure, that is, an intersection node of the road; Indicates the road network edge; is the adjacency matrix of the road network, which is used to describe the connection relationship between traffic nodes and the road length; Indicates the total number of traffic nodes; and Respectively represent i and j Traffic nodes; Representation node and connected roads; is the length of the road section, and the specific expression is shown in formula (2); （2）； Where, Indicates road section The length of ; inf means infinity; It can be expressed by formula (3): （3）； Considering only the connection edges and nodes of the power grid, the topological structure of the power grid is obtained, and its topological structure can be expressed by formula (4): （4）； Where, is the grid topology; is the node set of the power grid; is the set of grid edges; It is the parameter matrix of the relevant information of the power grid branch; Indicates the total number of grid nodes; Representation node and The connected power grid line; , , and Represents the power grid branches x - y The resistance, reactance, susceptance and the transmission power limit of the branch; The topology of the power station layer is shown in formula (5): （5）； Where, is the charging station topology; A set of nodes representing charging stations; Represents the relevant information parameter matrix of the charging station; Indicates the k The number of the charging station; Indicates the total number of charging stations; For the k The number of charging piles in each charging station; Indicates the capacity of the newly built supporting transformer; Indicates the k The land price and labor costs of each charging station; The road-network-station coupling network topology also includes the interconnection relationship between the road-network-station levels, and its specific expression is shown in formula (6): （6）； Where, Represents the "Route-Network-Station" route-network-station coupled network topology; is the coupling network edge, which represents the coupling between the road network node, the power grid node and the charging station access node; is the coupling network edge that is not connected to the charging station, that is, the coupling between the road network node and the power grid node; It is the coupling network edge connected to the charging station, that is, the coupling of the three nodes "road-network-station".

3. The method according to claim 1, characterized in that In step 3, the multi-objective planning model of the charging station considering the load balancing effect is constructed as follows: The objective function of the multi-objective planning model of the charging station is: （14）； Where, The planned total revenue for the charging station; The total planned investment cost for the charging station; is the total operation and maintenance cost of the charging station; The investment cost of the transformer for the charging station; Operation and maintenance costs of the supporting transformer; is the network loss cost of the distribution network; is the loss cost of the transformer; The total revenue from the charging station plan is: （15）； Where, The time-of-use electricity price for charging stations; The time-of-use electricity purchase price for charging stations from the power grid company; For the k Charging station No. t The number of charging stations that are busy at any given time; is the rated power of the charging pile; The total planned investment cost of the charging station is: （16）； Where, is the area occupied by the charging station, which is directly proportional to the number of charging piles in the charging station; The land price; Planning capacity for charging stations; is the unit price of the charging pile; is the discount rate; T To plan the service life; The total operation and maintenance cost of the charging station is: （17）； Where, The unit charging equipment operation and maintenance cost; The investment cost of the transformer for the charging station is: （18）； （19）； Where, The planned capacity of the transformer supporting the charging station; is the unit price of its corresponding capacity; The transformer capacity planning should be adapted to the number of charging equipment in the charging station. , , , and Plan the capacity of the transformer for the number of charging piles; The operation and maintenance cost of the supporting transformer is: （20）； Where, is the unit operation and maintenance cost of the transformer; is the maximum annual utilization hours; The loss cost of the transformer is: （21）； Where, is the unit loss cost of the transformer; and are the rated iron loss and copper loss of the transformer respectively; The network loss cost of the distribution network is: （22）； Where, is the unit network loss cost of the distribution network; For the power grid l Line loss of the line; is a collection of lines; The number of charging piles at a charging station is configured as shown in formula (23): （23）； Where, Plan the maximum number of charging piles for the charging station. k The number of charging piles per charging station The minimum value of the maximum number of electric vehicles arriving at the service area in one day; exist t Moment k The number of electric vehicles queuing for charging at a charging station can be obtained by formula (24): （24）； Where, for t -1 moment in k The number of electric vehicles waiting in line at each charging station; Indicates the k Charging stations t The number of charging piles that are not busy at any time can be calculated by formula (25); （25）； Where, and Respectively t At the moment k The number of charging piles in the busy state and the charging piles in the state of being about to be charged; and It can be calculated by formula (26) and (27) respectively: （26）； （27）； Where, and Respectively represent k Charging station No. t -1 The number of charging piles that are always busy and about to be charged; and Respectively represent k Charging station No. t -1 and t -2 The number of newly charged electric vehicles at any given moment; exist t Moment k The number of newly charged electric vehicles in a charging station can be expressed as: （28）； According to formulas (24)-(28), the electric vehicle k Charging stations t Queueing time at the moment: （29）； Where, is the queue impact coefficient; After taking into account the queuing time, the total time the user spends traveling to the charging station is: （30）； Where, For electric vehicles k Charging stations and t The waiting time in queue is very short.

4. The method according to claim 3, characterized in that In step 3, the constraints of the multi-objective programming model of the charging station can be expressed as: 1) Upper and lower limit constraints of node amplitude: （31）； Where, For nodes x The node voltage at and are its upper and lower limits respectively; 2) Branch flow constraints: （32）； Where: and They are t Time Node x Active power and reactive power; and They are t Time Node x and nodes y The voltage amplitude; and Branch xy The conductance and susceptance of For nodes x and nodes y Phase angle difference between voltages; is the node set of the power grid; 3) Planning the number of charging piles and transformer capacity constraints in the charging station: （33）； Where, and To plan the upper and lower limits of the number of charging piles in the charging station; 4) Planning the number of charging stations: （34）； （35）； Where, and The planned number of charging stations upper and lower limits; is the total number of electric vehicles in the planning area; is the battery capacity; is the average charging time; Indicates rounding down; 5) Voltage deviation constraint: （36）； Where, is the reference voltage value; is the voltage fluctuation rate; is the maximum volatility constraint; 6) Distance constraints between charging stations: （37）； Where, and Represents the distance between any two charging stations Maximum and minimum constraints; 7) Substation main transformer load factor constraints: （38）； Where, For the k The grid nodes connected to the charging stations x Basic electricity load at the location; Indicates the k The grid nodes connected to the charging stations x Electric vehicle charging load at For the k The rated capacity of the transformer configured for each charging station; To plan the upper limit of transformer load rate; 8) Transformer capacity constraints: （39）； Where, and To plan the upper and lower limits of transformer capacity.

5. The method according to claim 1, wherein In step 4, when solving the constructed multi-objective programming model of the charging station, the following steps are adopted: Step 4-1: Initialize parameters; Step 4-2: Generate initial antibody population and memory library; Step 4-3: Calculate individual fitness; substitute the corresponding parameters into the objective function to obtain the corresponding and Considering that the revenue calculation of charging stations is based on maximizing the revenue, while the goal of electric vehicle users is to minimize the time spent on charging; in order to facilitate the comparison of the advantages and disadvantages of different planning schemes, a global variable operator is defined and , used to normalize multiple target values; Will and After normalization, the global fitness can be calculated using formula (40): （40）； Where, and is the weight; Step 4-4: Calculate the affinity and antibody concentration between individual antibodies; Affinity between individual antibodies Indicates antibody u and v The similarity between the two groups, the antibody concentration Indicates the overall similarity of the antibody population; （41）； （42）； Where, Indicates antibody u and v Length of the same value; represents the total length of the antibody individual; m is the total amount of antibody population; Step 4-5: Determine whether the antibody population meets the conditions; determine whether the antibody concentration of the antibody population is equal to 1. If so, obtain the planning result and terminate the cycle; if not, continue with steps 6-8; Steps 4-6: Selection operation: select excellent individuals from the exchanged individual antibodies, and use the "roulette wheel" method to select individuals from the antibody population to generate a new population; Steps 4-7: Crossover operation: select two antibody individuals from the antibody population according to a certain exchange probability and randomly exchange them to generate new individuals; Step 4-8: Select antibody individuals according to a certain probability, and randomly change the value of a gene in the string structure data of the selected individuals to generate new individuals; And return to step 3 for a new round of iteration.

Images