Electric vehicle charging demand prediction method considering charging management strategy

Abstract

The invention discloses an electric vehicle charging demand prediction method considering a charging management strategy, belongs to the technical field of electric vehicle charging demand prediction, and solves the aggregation effect of electric vehicle charging loads and the uncertainty of high coupling between the charging loads. And the safe and stable operation of the power grid and the reliability of power supply are severely influenced by the peak load. The travel space and time characteristics of an electric vehicle user are modeled by constructing an electric vehicle travel chain model, so that the travel behavior of the electric vehicle user is analyzed; constructing an urban traffic network and a delay model, assigning an edge weight of a traffic network topology by using road traffic delay, and constructing a minimum road delay time objective function; constructing a road-power grid coupling node aggregation model; constructing an electric vehicle charging load model, and simulating electric vehicle user charging demand space-time distribution by adopting Monte Carlo; the method can achieve the accurate prediction of the charging demands of the electric vehicle users, and optimizes the planning and operation of the power distribution network.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric vehicle charging demand prediction, and relates to an electric vehicle charging demand prediction method, device and storage medium considering a charging management strategy. Background Art

[0002] In recent years, problems such as environmental pollution and energy shortage have gradually attracted public attention. As a low-carbon and environmentally friendly means of transportation, electric vehicles have developed rapidly. With the continuous progress of electric vehicle-related technologies, large-scale high-power centralized charging stations and AC charging piles widely distributed in various parking lots have been successively built, and a comprehensive electric vehicle charging network has gradually taken shape. However, due to the aggregation effect of electric vehicle charging loads and the highly coupled uncertainty between charging loads, peak loads seriously affect the safe and stable operation of the power grid and the reliability of power supply. Therefore, accurately predicting the charging demand of electric vehicles helps to reasonably plan the construction of charging infrastructure and optimize the operation of urban power grids.

[0003] The existing research methods for electric vehicle charging demand at the present stage have the following deficiencies: existing research has not simultaneously combined the spatio-temporal distribution state of electric vehicle users and the in-station charging management strategy, and considered factors such as the travel characteristics of electric vehicle users and the charging pile selection preferences, which will affect the accuracy of the prediction model application to a certain extent, resulting in unreasonable distribution of charging station resources and low operation efficiency. Summary of the Invention

[0004] The technical solution of the present invention is used to solve the problems of the aggregation effect of electric vehicle charging loads and the highly coupled uncertainty between charging loads, and the peak load seriously affecting the safe and stable operation of the power grid and the reliability of power supply.

[0005] The present invention solves the above technical problems through the following technical solutions:

[0006] The present invention provides an electric vehicle charging demand prediction method considering a charging management strategy, including:

[0007] S1 Construct an electric vehicle travel chain model to model the travel space and time characteristics of electric vehicle users, so as to analyze the travel behavior of electric vehicle users;

[0008] S2 Construct an urban traffic road network and delay model, assign weights to the edges of the traffic network topology using road traffic delays, determine the travel route with road traffic delays as the path decision factor, and construct a minimum road delay time objective function;

[0009] S3 Construct a road-power grid coupling node aggregation model;

[0010] S4 constructs an electric vehicle charging load model and uses Monte Carlo simulation to analyze the spatio-temporal distribution of electric vehicle users' charging demands.

[0011] Further, the construction of the electric vehicle trip chain model is as follows:

[0012]

[0013] Among them, G TC is the set of spatio-temporal characteristic quantities of electric vehicle trips; s i represents the starting point of the user's trip; d i represents the destination of the user's trip; t0 is the starting time of the user's trip, is the travel time for the user to travel from the starting point s i to the destination d i ; is the residence time of the user at the destination d i ; is the travel distance of the i-th trip; the starting and ending points in the trip chain are represented by H M , W M , C M , O M respectively.

[0014] Further, the method for modeling the travel space and time characteristics of electric vehicle users is as follows:

[0015] Calculate the state transition probability according to the Markov chain:

[0016]

[0017] Among them, m0, m1, …, m t-1 , m t , m t+1 respectively represent the states of the user at times 0 to t + 1; represents the transition probability for the user to move from state m t to state m t+1 ;

[0018] Construct a state transition probability matrix to represent the transition probability of electric vehicles between different locations:

[0019]

[0020] Among them:

[0021]

[0022] Among them, the matrix elements are composed of the transition probabilities between various functional locations in the city;

[0023] Construct the probability density function of the starting time of the travel chain:

[0024]

[0025] Among them, μ T is the mean value of the time of leaving home for the first time, and σ T is the variance of the time of leaving home for the first time;

[0026] Construct the probability density function of the parking time at different destination places:

[0027]

[0028] Among them, f H (x, λ, k) is the probability density function of the residence time of the electric vehicle in the residential area; H(x, μ, δ, ε) is the probability density function of the residence time of the electric vehicle in the work area and the commercial area; k is the shape parameter, λ is the scale parameter, μ is the location parameter; δ is the scale parameter; ε is the shape parameter;

[0029] Calculate the starting time of the next trip:

[0030]

[0031] Among them, J is the total number of trips in the travel chain; t j+1 is the time when the vehicle starts the (j + 1)-th trip; t0 is the time when the vehicle leaves home for the first time; is the total time consumed by the vehicle to complete the previous j trips; t stay,j is the residence time of the vehicle at the destination of the j-th trip;

[0032] Calculate the charging duration of the vehicle:

[0033]

[0034] Among them, T park represents the parking duration of the electric vehicle, C represents the battery capacity, P charge represents the charging power, and η represents the charging efficiency.

[0035] Furthermore, the method for constructing the urban traffic road network and delay model is as follows:

[0036] Construct the urban traffic road network model as follows:

[0037]

[0038] Among them, G is the road network set; O is the set of all intersections in the road network, with a total of u; L is the set of road segments in the road network; T is the set of divided time periods, and in the present invention, the whole day is divided into h time periods; W is the set of road segment weights; o i is the i-th intersection of the road network; lij is the road connecting the i-th and j-th intersections; ω ij (t) is the weight of road section v ij at time period t;

[0039] The urban traffic delay model is constructed as follows:

[0040]

[0041] where κ n (t) represents the delay time of road network intersection nodes; v m (t) represents the delay time of road sections.

[0042] Furthermore, taking the road traffic delay as a path decision factor to determine the travel route, the minimum road delay time objective function is constructed as follows:

[0043]

[0044] where f(t) represents the objective function of the minimum road delay time; U is the set of road sections of the shortest path of vehicle travel; λ m is a 0-1 flag variable representing the driving state of the vehicle on road section m.

[0045] Furthermore, the road-power grid coupled node aggregation model is constructed as follows:

[0046]

[0047] where CD k represents the coupled node after the power distribution of the road network node to the power distribution system; D m represents all power distribution network nodes, and there is D m ={m|m = 1, 2,..., M}; M is the total number of power distribution network nodes; R z is the set of road network nodes that can provide charging stations / piles; k is the number of coupled nodes, and the upper limit of the quantity is equal to the total number of power distribution network nodes; dist(i, R n ) represents that the coupled node i is the power distribution network node closest to the road network node R n ; dist(j, R n ) represents that the coupled node j is the power distribution network node closest to the road network node Rn; R n represents all road network nodes, and there is R n ={n|n = 1, 2,..., N}; represents the peak charging load of electric vehicles of the corresponding road network nodes within the power supply range of the coupled node CD k ; represents the original basic load at the coupled power distribution network node; is the upper limit of the load potential that the distribution transformer at the road network coupling node i can accept.

[0048] Furthermore, the method for constructing the electric vehicle charging load model is as follows:

[0049] Construct the battery capacity model as follows:

[0050]

[0051] where x represents the battery capacity of the electric vehicle, and a and b are the upper and lower limits of the electric vehicle capacity distribution range;

[0052] Construct the electric vehicle energy consumption model as follows:

[0053]

[0054] where is the total power consumption of the vehicle from s i traveling to d i ; e0 is the power consumption per unit mileage; is the battery power when the electric vehicle reaches the destination; E0 represents the starting battery power of the electric vehicle; is the state of charge when the electric vehicle reaches the destination; B ev is the battery capacity;

[0055] Construct the user charging decision model as follows:

[0056]

[0057] where r is the charging decision number, and S min represents the power threshold;

[0058] Construct the user's selection probability model for fast and slow charging piles as follows:

[0059]

[0060] where p hp,n and p lp,n are the selection probabilities of the nth electric vehicle user for fast and slow charging respectively; P low is the charging power of the slow charging pile; η is the charging efficiency of the charging pile; T p,n is the parking duration of the nth electric vehicle user; S start,n is the starting charging time power of the nth electric vehicle user; L n is the next driving mileage of the nth electric vehicle user; S th is the minimum allowable remaining battery power of the electric vehicle;

[0061] Simulate the total user charging demand model as follows:

[0062] W n = (S end,n - S start,n )B(24)

[0063]

[0064] where W n is the charging demand generated by the nth electric vehicle; S end,n is the SOC at the end of charging of the nth electric vehicle; u represents the number of peak periods; S n1 represents the charging amount of the nth electric vehicle during the parking duration; (t pt,start , t pt,end ) i represents the time period from the start to the end of the ith charging peak; ω ∈ {hp, lp} is the type of charging pile selected by the user; P ω is the power of the charging pile of type ω; λ is the time constant of power decay; and are respectively the constant current charging duration, parking time, starting power and total span duration of the nth electric vehicle from the end of the ith charging peak to the start of the (i + 1)th charging peak.

[0065] Furthermore, the method for adopting Monte Carlo simulation to simulate the spatio-temporal distribution of the charging demand of electric vehicle users is as follows:

[0066] 1) Model the urban road network, and input the road adjacency matrix and the road delay matrix calculated according to the traffic delay model;

[0067] 2) Initialize the basic vehicle information, and generate the battery capacity, initial SOC, and initial travel time characteristic quantities of the electric vehicle;

[0068] 3) Extract the complete actual travel chain, and plan the optimal travel path for the vehicle according to the matrix and the prior path algorithm;

[0069] 4) Determine the spatio-temporal distribution state of each user through the battery capacity model, energy consumption model and charging decision model, and screen out the electric vehicle users who need to charge;

[0070] 5) Determine the type of charging pile selected by the user to be charged according to the charging pile selection probability model;

[0071] 6) Determine the actual charging amount of the user to be charged according to the charging management strategy;

[0072] 7) Considering the coupling relationship of the road-power network, reduce the charging demand under each road network node to the grid node according to the distribution network power supply unit partition;

[0073] 8) If the cumulative mileage exceeds the extracted driving mileage length, the trip chain ends and enters the next loop until the number of vehicles reaches the vehicle ownership.

[0074] The present invention also provides an electronic device, including a memory and a processor. The memory is used to store a program that supports the processor to execute the above-mentioned electric vehicle charging demand prediction method considering the charging management strategy, and the processor is configured to execute the program stored in the memory.

[0075] The present invention also provides a storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the above-mentioned electric vehicle charging demand prediction method considering the charging management strategy.

[0076] The beneficial effects of the present invention are as follows:

[0077] Based on the trip chain model, the present invention analyzes the travel behavior of electric vehicle users: statistically analyzes the time and space characteristic quantities of the trip chain of electric vehicle users and establishes a probability model; considers traffic factors to analyze the spatio-temporal distribution of electric vehicle charging demand: establishes an urban traffic delay model based on the road resistance theory, and combines the directed graph theory to plan the prior travel full trajectory for each electric vehicle; establishes a supply division model coupled with the distribution unit: better reflects the coupling relationship between the spatio-temporal distribution characteristics of the charging load in the road network and the distribution network; calculates the electric vehicle charging demand based on Monte Carlo simulation: comprehensively considers the spatio-temporal power characteristics of electric vehicle users and the charging pile selection preferences, and proposes a charging management strategy to predict the charging demand of electric vehicle users, which can more accurately calculate the actual charging demand of electric vehicle users and provide a more accurate basis for the subsequent site selection and capacity determination of charging stations; the present invention can accurately predict the charging demand of electric vehicle users and optimize the planning and operation of the distribution network. Description of the Drawings

[0078] Figure 1 is a flowchart of the electric vehicle charging demand prediction method considering the charging management strategy of the present invention;

[0079] Figure 2 is a simulation verification diagram of the effectiveness of the charging management strategy of the electric vehicle charging demand prediction method considering the charging management strategy of the present invention. Detailed Embodiment

[0080] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0081] The technical solution of the present invention will be further described below in conjunction with the accompanying drawings of the specification and specific embodiments:

[0082] Embodiment 1

[0083] As Figure 1 shown, a method for predicting the charging demand of an electric vehicle considering a charging management strategy according to an embodiment of the present invention includes the following steps:

[0084] Step 1. Analyze the travel behavior of electric vehicle users based on the travel chain model.

[0085] Step 1.1. Construct an electric vehicle travel chain model:

[0086]

[0087] In formula (1): G TC is a set of spatio-temporal characteristic quantities of electric vehicle travel; s i represents the starting point of the user's travel; d i represents the destination of the user's travel; t0 is the starting time of the user's travel, is the travel time for the user to travel from the starting point s i to the destination d i ; is the residence time of the user at the destination d i ; is the travel distance of the i-th trip; the starting point and the end point in the travel chain are respectively represented by H M , W M , C M , O M , which are the residential area, the working area, the commercial area and other areas respectively.

[0088] Step 1.2. Model the travel space characteristics of electric vehicle users;

[0089] Step 1.2.1. The Markov chain is used to describe the transition probability between a user moving from one node to another stopping node. Calculate the state transition probability according to the Markov chain:

[0090]

[0091] In formula (2): m0, m1,..., m t-1 , m t , m t+1 respectively represent the states of the user at times 0 to t + 1; represents the transition probability for the user to move from state m t to state m t+1 .

[0092] Step 1.2.2. For the four urban functional areas divided in the present invention, an electric vehicle user can have four different states at time t. A state transition probability matrix is constructed to represent the transition probability of the electric vehicle between different places:

[0093]

[0094] Where:

[0095]

[0096] In formulas (3)-(4), The matrix elements are composed of the transition probabilities between the functional places in the city.

[0097] Step 1.3. Model the travel time characteristics of electric vehicle users;

[0098] Step 1.3.1. Construct the probability density function of the starting time of the travel chain:

[0099]

[0100] In formula (5): μ T Is the mean value of the initial leaving home time, with a value of 7.32; σ T Is the variance of the initial leaving home time, with a value of 1.34.

[0101] Step 1.3.2. Construct the probability density function of the parking time at different destination places:

[0102]

[0103] In formulas (6)-(7): f H (x, λ, k) is the probability density function of the residence time of the electric vehicle in the residential area; H(x, μ, δ, ε) is the probability density function of the residence time of the electric vehicle in the work area and the commercial area (both follow the generalized extreme value distribution); k is the shape parameter, with a value of 14.05; λ is the scale parameter, with a value of 7.16; μ is the location parameter; δ is the scale parameter; ε is the shape parameter. Among them, μ, δ ∈ R, δ > 0. When ε > 0, record 1 / ε = α; when ε < 0, record -1 / ε = α.

[0104] Step 1.3.3. Calculate the start time of the next trip:

[0105]

[0106] In formula (8): J is the total number of trips in the travel chain; t j+1 Is the time when the vehicle starts the (j + 1)-th trip; t0 is the initial leaving home time of the vehicle; is the total time consumed for the vehicle to complete the first j trips; t stay,j is the residence time of the vehicle at the destination of the j-th trip.

[0107] Step 1.3.4. Calculate the charging duration of the vehicle:

[0108]

[0109] In Equation (9): T park represents the parking duration of the electric vehicle, C represents the battery capacity, P charge represents the charging power, and η represents the charging efficiency.

[0110] Step Two. Analyze the spatio-temporal distribution of the charging demand of electric vehicles considering traffic factors.

[0111] Step 2.1. Construct a city traffic road network model:

[0112]

[0113] In Equation (10): G is the road network set; O is the set of all intersections in the road network, with a total of u; L is the set of road segments in the road network; T is the set of divided time periods. In the present invention, the whole day is divided into h time periods; W is the set of road segment weight values; o i is the i-th intersection of the road network; l ij is the road connecting the i-th and j-th intersections; ω ij (t) is the weight value of road segment v at time t ij . The connection relationship between intersections in the road network set G is described by an adjacency matrix D. The element d of matrix D ij (t) is expressed as:

[0114]

[0115] Step 2.2. Construct a city traffic delay model:

[0116]

[0117] In Equation (12): κ n (t) represents the delay time of the road network intersection node; v m (t) represents the delay time of the road segment.

[0118] By comprehensively considering the influence of road traffic flow density and traffic signals on the actual road passing time, the Webster steady-state delay model and the Akcelik instantaneous delay model can be combined to establish a traffic delay model for intersection nodes:

[0119]

[0120] In formula (13): τ is the traffic signal cycle ratio; ζ is the green light ratio of traffic signals; S is the road traffic flow; S0 is the critical saturation of road traffic flow; ρ is the road vehicle driving rate.

[0121] According to the improved time-flow road impedance model, a traffic delay model for road sections can be established as follows:

[0122]

[0123] In formula (14), t0 is the theoretical traffic time of the road; both Α and Β are factor parameters affecting the road section impedance.

[0124] Step 2.3: Use the road traffic delay function to assign weights to the edges of the traffic network topology, and the weight of each road section corresponds to an independent random variable t m , whose expected value is denoted as μ m :

[0125] E(t m ) = μ m = t free + At free E(S) B + E(κ n ) (15)

[0126]

[0127] In formulas (15)-(17): t m is the travel time of electric vehicles on each road section; t free is the free flow travel time of the road; E(S) is the expected value of the road traffic flow density; E(κ n ) is the expected value of the road intersection delay time.

[0128] Step 2.4: Use the road traffic delay as a path decision factor to determine the travel route, and construct a minimum road delay time objective function:

[0129]

[0130] In formula (18): f(t) represents the objective function with the minimum road delay time; U is the set of road sections of the shortest vehicle travel path; λ m is a 0-1 flag indicating the driving state of the vehicle on road section m.

[0131] Step three: Construct a road-power grid coupling node aggregation model as follows:

[0132]

[0133] In formula (19): CD kDenote the coupling node after the power distribution of the road network nodes to the power distribution system; D m Denote all the power distribution network nodes, with D m ={m|m = 1, 2, …M}; M is the total number of power distribution network nodes; R z Is the set of road network nodes that can provide charging stations / piles; where k is the number of coupling nodes, and the upper limit of the quantity is equal to the total number of power distribution network nodes; dist(i, R n ) represents that the coupling node i is the power distribution network node closest to the road network node R n ; dist(j, R n ) represents that the coupling node j is the power distribution network node closest to the road network node R n ; R n Denote all the road network nodes, with R n ={n|n = 1, 2, …, N}; Denote the peak charging load of electric vehicles corresponding to the road network nodes within the power supply range of the coupling node CD k ; Denote the original basic load at the coupling power distribution node; Is the upper limit of the load potential that the distribution transformer at the road network coupling node i can accept.

[0134] Step Four: Calculate the electric vehicle charging demand based on Monte Carlo simulation.

[0135] Step 4.1: Build an electric vehicle charging load model;

[0136] Step 4.1.1: Build a battery capacity model:

[0137]

[0138] In formula (20): x represents the battery capacity of the electric vehicle, and a and b are the upper and lower limits of the electric vehicle capacity distribution range, with the unit of kW·h. The present invention assumes that the battery capacity of the electric vehicle follows a uniform distribution of U~[40, 55].

[0139] Step 4.1.2: Build an electric vehicle energy consumption model:

[0140]

[0141] In formula (21), Is the total power consumption of the vehicle when driving from s i to d i ; e0 is the power consumption per unit mileage; Is the remaining battery power when the electric vehicle reaches the destination; E0 represents the initial battery power of the electric vehicle; Is the state of charge of the electric vehicle when it reaches the destination; B ev Is the battery capacity.

[0142] Step 4.1.3, Construct the user's charging decision model:

[0143] Due to the influence of the user's psychological uncertainty, users usually start considering charging when the battery level drops to a certain level and do not wait until the battery is completely depleted. In this section, this battery level threshold is called the minimum SOC, denoted by S min , and set to 15%. When the SOC of an electric vehicle user is greater than 15%, the charging decision number r is 0, indicating that charging is not considered; otherwise, it indicates starting to charge.

[0144]

[0145] Step 4.1.4, Construct the user's selection probability model for fast and slow charging piles:

[0146]

[0147] In Equation (23): p hp,n and p lp,n are respectively the selection probabilities of the nth electric vehicle user for fast and slow charging; P low is the charging power of the slow charging pile; η is the charging efficiency of the charging pile; T p,n is the parking duration of the nth electric vehicle user; S start,n is the battery level of the nth electric vehicle user at the start of charging; L n is the next driving mileage of the nth electric vehicle user; S th is the minimum allowable remaining battery level of the electric vehicle.

[0148] Step 4.1.5, According to the charging management strategy proposed in the present invention: When the charging peak arrives, the electric vehicle users charging in the station first charge in a constant current mode until the threshold. If the users have not returned to the charging station after the end of the charging peak period, they continue to charge in a constant voltage mode. Simulate the total charging demand model of users as follows:

[0149] W n =(S end,n -S start,n )B (24)

[0150]

[0151] In Equations (24)-(27): W n is the charging demand generated by the nth electric vehicle; S end,n is the SOC of the nth electric vehicle at the end of charging; u represents the number of times during the peak period (in this section, it is assumed that the number of electric vehicles generating charging demands in each period in the target area exceeds 60% of the total number); S n1represents the charging amount of the nth electric vehicle during the parking duration; (t pt,start , t pt,end ) i represents the time period from the start to the end of the ith charging peak; ω ∈ {hp, lp} is the type of charging pile selected by the user (fast charging, slow charging); P ω is the power of the charging pile of type ω; λ is the time constant of power decay; and are respectively the constant - current charging duration, parking time, starting power, and total span duration of the nth electric vehicle from the end of the ith charging peak to the start of the (i + 1)th charging peak.

[0152] Step 4.2: Adopt Monte Carlo simulation to simulate the spatio - temporal distribution of the charging demands of electric vehicle users;

[0153] Input of the urban traffic road network model in Step 4.2.1: Model the urban road network, input the road adjacency matrix and the road delay matrix calculated according to the traffic delay model;

[0154] Input of parameters of electric - vehicle - related characteristic quantities in Step 4.2.2: Initialize the basic vehicle information, and generate characteristic quantities such as the battery capacity, initial SOC, and initial travel time of the electric vehicle.

[0155] Step 4.2.3: Extract the complete actual travel chain, and plan the optimal travel path for the vehicle according to the matrix and the prior path algorithm;

[0156] Step 4.2.4: Determine the spatio - temporal distribution state of each user through the battery capacity model, energy consumption model, and charging decision model, and screen out the electric - vehicle users who need to charge.

[0157] Step 4.2.5: Determine the type of charging pile selected by the users to be charged according to the charging - pile selection probability model.

[0158] Step 4.2.6: Determine the actual charging amount of the users to be charged according to the charging management strategy.

[0159] Power reduction of road - grid nodes in Step 4.2.7: Considering the coupling relationship of the road - power network, reduce the charging demands under each road network node to the grid nodes according to the distribution network power supply units.

[0160] Step 4.2.8: If the cumulative mileage exceeds the length of the extracted driving mileage, the travel chain ends, and enter the next loop until the number of vehicles reaches the vehicle ownership.

[0161] Simulation verification

[0162] To verify the effectiveness of the electric vehicle charging demand prediction method considering the charging management strategy of the present invention, taking fast charging piles as an example, the daily load curves of electric vehicle users in the target area are calculated under the condition of having the management strategy of the present invention and without the management strategy of the present invention, as shown in the appendix. Figure 2 As can be seen Figure 2 from this, under the charging management strategy of the present invention, the daily load curve of electric vehicle users in the target area shows a significant decrease during peak hours. This is because electric vehicle users charging in the station suspend the behavior of continuing to charge in constant current mode during this period, which can reduce the impact of large-scale electric vehicle charging load on the distribution network during peak hours, verifying the effectiveness of the present invention.

[0163] Aiming at the problem that traditional electric vehicle charging demand prediction is only used to determine the spatio-temporal distribution state of users and fails to calculate the actual charging amount of users, the present invention comprehensively considers the spatio-temporal power characteristics of electric vehicle users and the charging pile selection preference, and proposes a charging management strategy to predict the charging demand of electric vehicle users. The proposed method can calculate the actual charging demand of electric vehicle users more accurately, providing a more accurate basis for the subsequent site selection and capacity determination of charging stations.

[0164] Aiming at the problem that the results of traditional electric vehicle charging demand prediction are difficult to reflect the dynamic changes of actual charging load, this study introduces an urban traffic delay model and proposes a path planning method based on traffic delay. This method can dynamically adjust the vehicle driving path, significantly reducing the time waste and energy consumption increase caused by traffic congestion, thus improving the accuracy of electric vehicle charging demand prediction.

[0165] Aiming at the problem that the research on the coupling relationship between the spatio-temporal distribution characteristics of electric vehicle charging load and the distribution network is not sufficient, the present invention proposes a collaborative operation coupling model, revealing the dynamic coupling relationship between traffic characteristics and distribution network load. By establishing a power supply area division model for distribution units, the spatio-temporal characteristics of charging load in the road network and the interaction with the distribution network are effectively reflected, thereby improving the accuracy and reliability of electric vehicle charging demand prediction.

[0166] In summary, the present invention models the urban road network, inputs the road adjacency matrix and the road delay matrix calculated according to the traffic delay model; initializes the basic vehicle information, generates characteristic quantities such as the battery capacity of electric vehicles, the initial SOC, and the initial travel time, etc., extracts the complete actual travel chain, and plans the optimal travel path for the vehicle according to the matrix and the prior path algorithm; determines the spatio-temporal distribution of each user through the energy consumption model and the charging decision model, and screens out the electric vehicle users who need to charge; determines the type of charging pile selected by the users to be charged according to the charging pile selection probability model, and adopts the charging management strategy proposed by the present invention to determine the total charging demand of the users to be charged.

[0167] Embodiment 2

[0168] An apparatus, comprising a memory and a processor, where the memory is used to store a program for supporting the processor to execute the electric vehicle charging demand prediction method considering the charging management strategy in Embodiment 1, and the processor is configured to execute the program stored in the memory.

[0169] Embodiment 3

[0170] A storage medium stores a computer program, and when the computer program is run by a processor, it executes the steps of the electric vehicle charging demand prediction method considering the charging management strategy in Embodiment 1.

[0171] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the present invention in each embodiment.

Claims

1. A method for predicting the charging demand of an electric vehicle considering a charging management strategy, characterized in that, Including: S1 Build an electric vehicle travel chain model to model the travel space and time characteristics of electric vehicle users, so as to analyze the travel behavior of electric vehicle users; S2 Build an urban traffic road network and delay model, assign edge weights to the traffic network topology using road traffic delays, determine travel routes with road traffic delays as path decision factors, and build a minimum road delay time objective function; S3 Build a road-power grid coupling node aggregation model; S4 Build an electric vehicle charging load model, and use Monte Carlo simulation to simulate the spatio-temporal distribution of electric vehicle users' charging demands.

2. The method for predicting the charging demand of an electric vehicle considering a charging management strategy according to claim 1, wherein The method for building the electric vehicle travel chain model is as follows: Among them, G TC is a set of spatio-temporal characteristic quantities of electric vehicle trips; s i represents the starting point of the user's trip; d i represents the destination of the user's trip; t0 is the starting time of the user's trip, is the travel time for the user to travel from the starting point s i to the destination d i ; is the residence time of the user at the destination d i ; is the travel distance of the i-th trip; the starting point and the ending point in the trip chain are represented by H M , W M , C M , O M respectively.

3. The electric vehicle charging demand prediction method considering the charging management strategy according to claim 1, characterized in that The method for modeling the travel space and time characteristics of electric vehicle users is as follows: Calculate the state transition probability according to the Markov chain: Among them, m0, m1, …, m t-1 , m t , m t+1 respectively represent the states of the user at times 0 to t + 1; represents the transition probability of the user from state m t to state m t+1 in between; Build a state transition probability matrix to represent the transition probability of electric vehicles between different places: Where: Among them, the matrix elements are composed of the transfer probabilities between the functional places in the city; Build a probability density function for the starting time of the travel chain: Among them, μ T is the mean value of the time of leaving home for the first time, and σ T is the variance of the time of leaving home for the first time; Build a probability density function for the parking time at different destination places: where f H (x, λ, k) is the probability density function of the residence time of the electric vehicle in the residential area; H(x, μ, δ, ε) is the probability density function of the residence time of the electric vehicle in the work area and the commercial area; k is the shape parameter, λ is the scale parameter, μ is the position parameter; δ is the scale parameter; ε is the shape parameter; Calculate the starting time of the next trip: where J is the total number of trips in the travel chain; t j+1 is the time when the vehicle starts the (j + 1)-th trip; t0 is the time when the vehicle first leaves home; is the total time taken for the vehicle to complete the first j trips; t stay,j is the residence time of the vehicle at the destination of the j-th trip; Calculate the vehicle charging duration: Among them, T park represents the parking duration of the electric vehicle, C represents the battery capacity, P charge represents the charging power, and η represents the charging efficiency.

4. The electric vehicle charging demand prediction method considering a charging management strategy according to claim 1, wherein The method for building the urban traffic road network and delay model is as follows: The method for building the urban traffic road network model is as follows: Among them, G is the set of road networks; O is the set of all intersections in the road network, with a total of u; L is the set of road segments in the road network; T is the set of divided time periods, and the present invention divides the whole day into h time periods; W is the set of road segment weights; o i is the i-th intersection of the road network; l ij is the road connecting the i-th and j-th intersections; ω ij (t) is the weight of road segment v ij at time t; The method for building the urban traffic delay model is as follows: Among them, κ n (t) represents the delay time of the road network intersection node; v m (t) represents the delay time of the road section.

5. The method for predicting the charging demand of an electric vehicle considering a charging management strategy according to claim 1, wherein The method for determining travel routes with road traffic delays as path decision factors and building a minimum road delay time objective function is as follows: Among them, f(t) represents the objective function of the minimum road delay time; U is the set of road segments of the shortest path for vehicle trips; λ m is a 0-1 flag indicating the driving state of the vehicle on road segment m.

6. The electric vehicle charging demand prediction method considering a charging management strategy according to claim 1, characterized in that The method for building the road-power grid coupling node aggregation model is as follows: Among them, CD k represents the coupling node after the power of the road network node is distributed to the distribution system; D m represents all distribution network nodes, where D m ={m|m = 1, 2, … M}; M is the total number of distribution network nodes; R z is the set of road network nodes that can provide charging stations / piles; k is the number of coupling nodes, and the upper limit of the quantity is equal to the total number of distribution network nodes; dist(i, R n ) represents that the coupling node i is the distribution network node closest to the road network node R n ; dist(j, R n ) represents that the coupling node j is the distribution network node closest to the road network node R n ; R n represents all road network nodes, where R n ={n|n = 1, 2, …, N}; represents the peak value of the electric vehicle charging load of the corresponding road network nodes within the power supply range of the coupling node CD k ; represents the original basic load at the coupling distribution network node; is the upper limit of the load potential that the distribution transformer at the road network coupling node i can accept.

7. The method for predicting the charging demand of an electric vehicle considering a charging management strategy according to claim 1, wherein The method for building the electric vehicle charging load model is as follows: Build a battery capacity model as follows: Where x represents the electric vehicle battery capacity, and a and b are the upper and lower limits of the electric vehicle capacity distribution range; Build an electric vehicle energy consumption model as follows: Among them, is the total power consumption of the vehicle from s i traveling to d i ; e0 is the power consumption per unit mileage; is the battery level when the electric vehicle reaches the destination; E0 represents the initial battery level of the electric vehicle; is the state of charge when the electric vehicle reaches the destination; B ev is the battery capacity; Build a user charging decision model as follows: where r is the charging decision number, and S min represents the power threshold; Build a probability model for users' selection of fast and slow charging piles as follows: where p hp,n and p lp,n are the selection probabilities of fast and slow charging for the user of the nth electric vehicle respectively; P low is the charging power of the slow charging pile; η is the charging efficiency of the charging pile; T p,n is the parking duration of the user of the nth electric vehicle; S start,n is the battery power at the starting charging moment of the user of the nth electric vehicle; L n is the next driving mileage of the user of the nth electric vehicle; S th is the minimum allowable remaining battery power of the electric vehicle. Simulate the total user charging demand model as follows: W n = (S end,n - S start,n )B(24) Among them, W n is the charging demand generated by the nth electric vehicle; S end,n is the SOC at the end of charging of the nth electric vehicle; u represents the number of peak periods; S n1 represents the charging amount of the nth electric vehicle during the parking duration; (t pt,start , t pt,end ) i represents the time period from the start to the end of the ith charging peak; ω ∈ {hp, lp} is the type of charging pile selected by the user; P ω is the power of the charging pile of type ω; λ is the time constant of power attenuation; and are respectively the constant-current charging duration, parking time, starting power, and total span duration of the nth electric vehicle between the end of the ith charging peak and the start of the (i + 1)th charging peak.

8. The method for predicting the charging demand of an electric vehicle considering a charging management strategy according to claim 1, wherein The method for using Monte Carlo simulation to simulate the spatio-temporal distribution of electric vehicle users' charging demands is as follows: 1) Model the urban road network, input the road adjacency matrix and the road delay matrix calculated according to the traffic delay model; 2) Initialize the basic vehicle information, generate the electric vehicle battery capacity, initial SOC, and initial travel time characteristic quantities; 3) Extract a complete actual travel chain, and plan the optimal travel path for the vehicle according to the matrix and the prior path algorithm; 4) Determine the spatio-temporal distribution status of each user through the battery capacity model, energy consumption model and charging decision model, and screen out electric vehicle users who need to charge; 5) Determine the type of charging pile selected by the user to be charged according to the charging pile selection probability model; 6) Determine the actual charging amount of the user to be charged according to the charging management strategy; 7) Considering the coupling relationship of the road-power network, sum up the charging demands under each road network node according to the distribution of the power distribution network supply units and attribute them to the power grid nodes; 8) If the cumulative mileage exceeds the extracted driving mileage length, the travel chain ends, enter the next cycle until the number of vehicles reaches the ownership.

9. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program that supports the processor to execute the electric vehicle charging demand prediction method considering the charging management strategy according to any one of claims 1 to 8, and the processor is configured to execute the program stored in the memory.

10. A storage medium, on which a computer program is stored, characterized in that, When the computer program is run by a processor, it executes the steps of the electric vehicle charging demand prediction method considering the charging management strategy according to any one of claims 1 to 8.

Images