Charging load prediction method and device, medium and program product

Abstract

The invention provides a charging load prediction method and device, a medium and a program product. The method comprises the steps that a single electric vehicle model and a dynamic traffic road network model are built; based on the charging cost, the battery replacement cost and the charging / battery replacement waiting duration, calculating a comprehensive charging cost; calculating the driving cost based on the driving distance, the driving energy consumption and the driving time consumption; according to the driving cost and the comprehensive charging cost, predicting a charging decision result of each electric vehicle; and according to the charging and swapping decision result, calculating the charging and swapping load. According to the charging and swapping load prediction method, the possible charging and swapping condition of each electric vehicle is predicted according to the driving cost and the comprehensive charging cost, the space-time distribution of the charging and swapping load of the electric vehicle can be effectively predicted, and finally the electrical load of each charging / swapping station is calculated according to the space-time distribution. Accurate prediction of the charging load in one day in one region is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of interaction between electric vehicles and power grids, and specifically relates to a charging and swapping load prediction method, device, medium, and program product. Background Art

[0002] With the rapid expansion of the electric vehicle market, the impact of charging and swapping loads on transportation and the power grid is becoming increasingly significant. The large-scale integration of electric vehicles not only increases traffic pressure but also exacerbates peak-to-valley load variations in the power grid, reducing the economic efficiency of grid operation. Charging and swapping load forecasting provides strong data support for stable grid operation, the rational layout of charging and swapping infrastructure, and the formulation of appropriate charging and swapping pricing strategies.

[0003] Existing research, when predicting charging and swapping load, mostly considers only users' charging or swapping needs, ignoring the impact of users' simultaneous charging and swapping needs on the forecast of replenishment demand. Furthermore, while most existing research considers users' decision-making methods when selecting charging and swapping stations and driving routes, most of these methods fail to account for the differences in preferences between different types of electric vehicles, resulting in a relatively simplistic decision-making approach. Furthermore, existing research fails to fully consider the impact of changes in station selection and route planning during actual operation of multiple types of vehicles due to factors such as traffic conditions and electricity price fluctuations. Summary of the Invention

[0004] In response to the deficiencies in the prior art, the present invention provides a charging and swapping load prediction method, device, medium, and program product to solve the problem of inaccurate prediction results in charging and swapping load prediction due to differences in electric vehicle charging and swapping preferences and decision-making update behaviors.

[0005] The present invention achieves the above technical objectives through the following technical means.

[0006] A charging and swapping load prediction method, characterized by:

[0007] Build a single electric vehicle model and a dynamic traffic network model;

[0008] Calculate the comprehensive charging cost based on charging cost, battery replacement cost, and charging / battery replacement waiting time;

[0009] Calculate driving costs based on driving distance, driving energy consumption, and driving time;

[0010] Predict the charging and swapping decision results of each electric vehicle based on driving cost and comprehensive charging cost;

[0011] Based on the charging and swapping decision results, the charging and swapping load is calculated.

[0012] Furthermore, the method for predicting the charging and swapping decision results of each electric vehicle based on the driving cost and the comprehensive charging cost is as follows:

[0013] Calculate the comprehensive value of driving to each charging / swapping station for recharging:

[0014] V com =ω d V d +ω c V c

[0015]

[0016] Where V com is the comprehensive value, V d and V c are respectively the driving cost value and the comprehensive charging cost value, ω d and ω c are the driving cost weight coefficient and the comprehensive charging cost weight coefficient, ω d +ω c =1;D ev is the driving cost to the charging / swapping station, C ev is the comprehensive electricity supplement cost, D ref is the maximum driving cost acceptable to the user, C ref is the maximum comprehensive electricity replenishment cost acceptable to the user, α1 and β1 are the risk preference coefficients of the user under profit and loss conditions, respectively, where α1>0, β1<1, λ1 is the loss aversion coefficient, and λ1≥1;

[0017] Select V com The highest one is taken as the charging and swapping decision result.

[0018] Furthermore, the method for predicting the charging and swapping decision results of each electric vehicle based on the driving cost and the comprehensive charging cost is as follows:

[0019] Calculate the cumulative prospect value of driving to each charging / swapping station for recharging:

[0020]

[0021] Where, is the cumulative prospect value, ω d and ω c are the driving cost weight coefficient and the comprehensive charging cost weight coefficient, ω d +ω c =1,n d and m d is the number of gain and loss cases in driving cost, f d1 and f d2is the number of the profit and loss in the driving cost, where the profit is D ev <D ref , the loss situation is D ev ≥D ref , D ev D is the driving cost to the charging / swapping station, ref is the maximum driving cost acceptable to the user; n c and m c is the number of benefits and losses in the comprehensive electricity supplement cost, f c1 and f c2 is the number of the profit and loss situations in the driving cost, where the profit situation is C ev <C ref , the loss situation is C ev ≥C ref , C ev is the comprehensive electricity supplement cost, C ref The highest comprehensive electricity supplement cost acceptable to users; D ref and C ref Obtain possible values and corresponding probability distribution through past data statistics;

[0022] and Respectively represent the f d1 The cost-benefit situation of the first driving and the d2 The driving cost value V under the condition of driving cost loss d , and Respectively represent the f c1 The comprehensive cost-benefit of electricity supplement and the c2 The comprehensive electricity supplement cost value V under the condition of comprehensive electricity supplement cost loss c , V d and V c By calculating as follows:

[0023]

[0024] Where α1 and β1 are the risk preference coefficients of users in the case of gains and losses, respectively, where α1>0, β1<1, λ1 is the loss aversion coefficient, and λ1≥1;

[0025] By calculating as follows:

[0026]

[0027] In the formula, n, m, f1, and f2 are substituted into n d 、m d 、f d1 、fd2 Calculated and Substitute n, m, f1, and f2 into n respectively c 、m c 、f c1 、f c2 Calculated and p is the probability variable, p f1 represents the probability of the f1th profit situation occurring, represents the probability of the f2th loss situation occurring, W + (p) is the user’s subjective perception probability of the profit situation, W - (p) is the user's subjective perceived probability of loss, is the decision weight when the user is in the f1th profit situation, is the decision weight when the user is in the f2th loss situation, γ and δ are the risk-return attitude coefficient and risk-loss attitude coefficient respectively;

[0028] and According to D ref The probability distribution of possible values is obtained, and According to C ref The probability distribution of possible values is obtained;

[0029] choose The highest one is taken as the charging and swapping decision result.

[0030] Furthermore, the comprehensive electricity supplement cost is:

[0031]

[0032] Where C cb C is the electricity supplement fee. cb,max and C cb,min are the maximum and minimum charging costs in all charging / swapping stations, T w T is the waiting time for charging and swapping. w,max and T w,min The maximum and minimum waiting times in all charging / swapping stations respectively;

[0033] Electricity replenishment fee C cb The charging cost C cs or battery replacement cost C bs ,in:

[0034] Charging fee C cs for:

[0035] C cs =E0(e0,t +e cs,t )(S end -S0)

[0036] Where E0 is the battery capacity of the electric vehicle, e 0,t is the electricity price during period t, e cs,t is the charging service fee for period t, S end is the SOC of the electric vehicle after charging is completed, and S0 is the SOC of the electric vehicle before charging;

[0037] Battery replacement cost C bs for:

[0038] C bs =e b,t ΔP bs -C d

[0039] e b,t =e 0,t +e bs,t

[0040] Where, e b,t is the battery replacement price during period t, e bs,t is the battery replacement service fee during period t, C d Compensation for battery replacement for partially charged batteries is calculated as follows:

[0041] C d =e b,t (1-ρ b )

[0042] Where, ρ b is the discount rate for battery replacement, which is calculated as follows:

[0043]

[0044] Where S b,ρ =1-S chg S is the power gap between a partially charged battery and a fully charged battery. chg is the SOC of the battery, r b is the discount rate coefficient, S′ b,ρ is the inflection point of the discount rate curve;

[0045] Charging and swapping waiting time T w The calculation is as follows:

[0046] T w =T queue +T cb

[0047]

[0048] Where, Tqueue is the queue time, T cb is the charging time, k cb is the number of electric vehicles that need to be recharged in the charging / swapping station, c cb is the number of charging facilities in the charging / swapping station, T avr is the average charging time of electric vehicles, T cb,min is the shortest remaining charging time of the electric vehicles being served in the charging and swapping station, η ch and P ch are the charging efficiency and charging power of the charging pile respectively, T b Time for battery replacement.

[0049] Furthermore, the driving cost is calculated as follows:

[0050] The travel cost for a path is:

[0051]

[0052] Where R is the total number of roads in the route, ω1, ω2, and ω3 are the weighted scores of driving distance, driving energy consumption, and driving time, respectively. are the normalized values of the length, energy consumption, and time consumption of road r, respectively;

[0053] Use Dijkstra algorithm to minimize the travel cost P total As the goal, the path of each electric vehicle to each charging / swapping station is planned, and the driving cost of the planned path is used as the driving cost D of the electric vehicle to the charging / swapping station. ev .

[0054] Furthermore, the single electric vehicle model includes: probability distribution of the first trip time of each type of electric vehicle, probability distribution of the stay time, trip chain, and energy consumption model, wherein the energy consumption model is:

[0055]

[0056] Where, E com is the comprehensive energy consumption per unit mileage of electric vehicles, E T (T en ) is the electric vehicle at ambient temperature T en Energy consumption per mile under T (20℃) is the energy consumption per unit mileage of electric vehicles at an ambient temperature of 20℃, E v is the energy consumption per unit mileage of the electric vehicle at speed v;

[0057] Speed and energy consumption of electric vehicles:

[0058]

[0059] Where, E v is the energy consumption per unit mileage of the electric vehicle at speed v, α, β, ω, λ are speed energy consumption coefficients;

[0060] Between ambient temperature and electric vehicle energy consumption:

[0061]

[0062] Where, E T The electric vehicle is at an ambient temperature T en Energy consumption per mile under a n is the temperature energy consumption coefficient.

[0063] Furthermore, the dynamic traffic network model is:

[0064]

[0065] Where G t Represents the traffic network topology diagram under time period t, including N, E, W t Three elements, N is the set of all road nodes in the traffic network, N node is the total number of nodes, E is the set of all roads in the traffic network, (i, j) represents the road between node i and node j, W t is the dynamic road network information set in time period t, w t,ij is the dynamic information of road ij in time period t, including the length of road ij and the time and energy consumption required to complete the journey;

[0066] The time consumption is calculated based on the road length and the road speed, and the energy consumption is calculated based on the road length, the road speed and the energy consumption model; the road speed is calculated as follows:

[0067]

[0068] Where, v ij (t) represents the driving speed of road ij during time period t, i and j both represent road nodes in the traffic network, road ij represents the road between node i and node j, v ij,max represents the zero flow speed of road ij, Q ij (t) represents the traffic volume of road ij in period t, μ v is the speed coefficient, C ij is the traffic capacity of road ij, k1, k2, k3 are the adaptive coefficients of the road.

[0069] A computer device comprising a memory and a processor;

[0070] The memory is used to store computer programs;

[0071] The processor is used to execute the computer program and implement the above-mentioned charging and swapping load prediction method when executing the computer program.

[0072] A computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to execute the above-mentioned charging and swapping load prediction method.

[0073] A computer program product includes a computer program, which implements the above-mentioned charging and swapping load prediction method when executed by a processor.

[0074] The beneficial effects of the present invention are:

[0075] (1) The present invention provides a method, device, medium, and program product for predicting charging and swapping loads, wherein a single electric vehicle model and a dynamic traffic network model are established, and the possible charging and swapping conditions of each electric vehicle are predicted based on the driving cost and the comprehensive charging cost. This can effectively predict the spatiotemporal distribution of the charging and swapping loads of electric vehicles, and ultimately calculate the power load of each charging / swapping station, thereby achieving an accurate prediction of the charging and swapping load in a region during a day.

[0076] (2) The prediction method of the present invention estimates the comprehensive value of each user's charging and swapping decisions based on the cumulative prospect theory, thereby solving the problem of each user's acceptable maximum driving cost and unknown comprehensive charging cost in actual prediction situations, thereby providing technical possibilities for the smooth promotion and application of the prediction method of the present invention.

[0077] (3) In the present invention, when calculating the comprehensive charging cost, the charging / battery replacement cost and the waiting time are comprehensively considered, which can more comprehensively simulate the factors that influence the user's decision to choose charging / battery replacement, thereby being closer to the actual situation.

[0078] (4) When calculating the driving cost, the present invention comprehensively considers the length, time consumption and energy consumption of the path, and uses the Dijkstra algorithm to plan the path with the goal of minimizing the driving cost, so as to obtain the driving cost of each electric vehicle to each alternative charging / battery swap station. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 This is a flow chart of the prediction method of the present invention;

[0080] Figure 2 This is the road network topology diagram in the test case;

[0081] Figure 3 This is the functional area distribution diagram in the test case;

[0082] Figure 4This is a curve chart of time-of-use electricity prices and charging / battery swapping prices in the test case;

[0083] Figure 5 The power load curves of each charging / swapping station obtained in the test case. DETAILED DESCRIPTION

[0084] The embodiments of the present invention are described in detail below. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be understood as limiting the present invention.

[0085] 1. Technical Solution

[0086] Reference Figure 1 As shown in Figure 2, the charging and swapping load prediction method includes the following parts:

[0087] 1. Single electric vehicle model and dynamic traffic network model for multiple types of electric vehicles

[0088] 1.1 Single Electric Vehicle Model

[0089] (1) First trip time

[0090] For electric private cars, electric taxis, and electric official cars, Gaussian distribution is used to fit the first travel time of the three types respectively, and the probability density functions of the first travel time of the three types are obtained:

[0091]

[0092] Where t is the time period of the day, exp[·] represents the natural exponential function, and the first trip time refers to the start time of the electric vehicle's first trip of the day. The specific parameters in the probability density function of the first trip time are obtained by fitting data from the NHTS database and the Xi'an Resident Travel Survey Report in this embodiment. Therefore, the fitted probability density function may vary depending on the actual data participating in the survey.

[0093] (2) Length of stay

[0094] Regarding the length of time electric vehicles stay in each functional area:

[0095] 1) Due to operational requirements, electric taxis will not stay at a certain location for too long, so their stay time is set to follow the uniform distribution U(3,15);

[0096] 2) The dwell time of electric private cars and electric official cars varies according to the vehicle type and travel purpose. Their probability distribution can be approximately considered to follow a normal distribution. The probability density function of the corresponding dwell time is:

[0097]

[0098] Where σ is the standard deviation of the normal distribution, μ is the mean of the normal distribution; the length of stay of electric private cars in residential areas, work areas, commercial areas or leisure and entertainment areas follows the normal distribution N(8.54, 1.15 2 )、N(7.58,1.79 2 )、N(3.84,2.13 2 ); For electric official vehicles, their stay time all obeys the normal distribution N(3.31,1.75 2 ).

[0099] Note: The parameters in the above uniform distribution and normal distribution are obtained based on survey data in this embodiment.

[0100] (3) Travel chain type

[0101] 1) For private electric vehicles, their travel chains often start and end at home, forming a closed loop. These chains can be categorized into three types: HWH, H-W+S / LH, and HS / LH. Based on actual survey data, these three types of travel chains accounted for 49.9%, 23.1%, and 27%, respectively. H represents residential areas, W represents work areas, S represents commercial areas, and L represents leisure and entertainment venues.

[0102] 2) For electric taxis, since their travel chains are relatively complex and the transfer behavior between functional areas is highly random, their travel chains are characterized based on the OD matrix. The travel OD matrix is as follows:

[0103]

[0104] Where, OD t represents the OD matrix of the electric taxi travel chain during period t, It represents the probability of an electric taxi traveling from functional area P1 to functional area P2, and the values of P1 and P2 are H, W, S, and L.

[0105] 3) For electric official vehicles, their routes are relatively fixed and concentrated in specific areas. There are two types of travel chains: W1-W2-W1 and WS / LW, accounting for 45.6% and 54.4% respectively. Where W1 and W2 represent the first and second work zones, respectively.

[0106] (4) Energy consumption model

[0107] 1) The relationship between speed and energy consumption of electric vehicles is:

[0108]

[0109] Where, E vis the energy consumption per unit mileage of the electric vehicle at speed v, α, β, ω, and λ are speed energy consumption coefficients, which depend on the road grade. The speed energy consumption coefficients are different for different road grades.

[0110] 2) The relationship between ambient temperature and electric vehicle energy consumption is:

[0111]

[0112] Where, E T The electric vehicle is at an ambient temperature T en Energy consumption per mile under a n is the temperature energy consumption coefficient, and the specific coefficient value can be obtained by fitting the actual data.

[0113] 3) Based on the above relationship between electric vehicle energy consumption, speed and ambient temperature, the following comprehensive energy consumption is taken:

[0114]

[0115] Where, E com is the comprehensive energy consumption per unit mileage of electric vehicles, E T (T en ) is the electric vehicle at ambient temperature T en Energy consumption per mile under T (20℃) is the energy consumption per unit mileage of electric vehicles at an ambient temperature of 20℃.

[0116] 1.2 Dynamic Traffic Network Model

[0117] The traffic flow of the actual road network at different times will affect the driving speed of electric vehicles, thereby changing the driving energy consumption and driving time of electric vehicles. Therefore, the present invention introduces the speed-flow model to build a dynamic traffic network model to simulate the actual operation of the traffic network. Its expression is:

[0118]

[0119] Where, v ij (t) represents the driving speed of road ij during time period t, i and j both represent road nodes in the traffic network, road ij represents the road between node i and node j, v ij,max represents the zero flow speed of road ij (the driving speed when the traffic flow is zero), Q ij (t) represents the traffic volume of road ij in period t, μ v is the speed coefficient, C ij is the traffic capacity of road ij, k1, k2, k3 are the adaptive coefficients of the road.

[0120] Based on the inherent length of the road and the road speed obtained from the above expression, the time and energy consumption required for an electric vehicle to travel the road can be calculated. The time and energy consumption of each node, the roads between nodes, and each road are summarized to obtain a dynamic traffic network model:

[0121]

[0122] Where G t Represents the traffic network topology diagram under time period t, including N, E, W t Three elements, N is the set of all road nodes in the traffic network, N node is the total number of nodes, E is the set of all roads in the traffic network, (i, j) represents the road between node i and node j, W t is the dynamic road network information set in time period t, w t,ij is the dynamic information of road ij in time period t, including the length of road ij and the time and energy consumption required to complete the journey.

[0123] 2. Power replenishment cost

[0124] 2.1 Charging Fees

[0125] The charging cost of electric vehicles is calculated based on the charging price of the corresponding period and the user's charging degree. The charging cost is calculated as follows:

[0126] C cs =E0(e 0,t +e cs,t )(S end -S0)

[0127] Where C cs represents the charging cost, E0 is the battery capacity of the electric vehicle, e 0,t is the electricity price during period t, e cs,t is the charging service fee for period t, S end S0 is the state of charge (SOC) of the electric vehicle after charging is completed, and S0 is the SOC of the electric vehicle before recharging (charging or replacing the battery).

[0128] 2.2. Battery replacement costs

[0129] (1) Battery dispatch strategy for battery swap stations

[0130] 1) In any time period, the total number of batteries in the battery swap station is constant N bssThe battery is divided into three categories according to its working status: waiting to be charged, charging, and fully charged. Assuming that the battery charging process at the battery swap station is regarded as constant power charging, the SOC change of the charging battery per unit time (the duration of each period) is constant; therefore, based on the above-mentioned SOC change per unit time as the basis for equal division, the SOC is divided into K intervals from 0 to 1, with the SOC of the first interval being the smallest and the SOC of the Kth interval being the largest. The transition relationship between the three types of battery states is as follows:

[0131]

[0132] Where, They represent the number of batteries to be charged, batteries being charged, and fully charged batteries in the t+1 period, and the corresponding subscript t represents the number in the corresponding t period. Indicates the number of batteries under charge whose charge is in the first SOC interval in the t+1 period. Indicates the number of batteries under charge whose charge is in the Kth SOC interval during the tth period, Indicates the number of batteries replaced during the t+1 period.

[0133] In the above formula: Based on the following battery scheduling strategy, the batteries to be charged are charged in order of SOC from low to high, so It must be the number of newly added charging batteries in the t+1 period, so the first row This part of the quantity needs to be subtracted, the second line This part of the quantity needs to be added; based on the above SOC interval division basis, the battery in the Kth SOC interval during the t period can be fully charged and converted to a fully charged battery in the next period, so This batch of batteries will be fully charged by the time t+1, so the second row This part needs to be subtracted, the third line This part needs to be added.

[0134] Battery scheduling should meet the constraints of the total number of batteries and the number of batteries charged at the battery swap station:

[0135]

[0136] Where Z b The number of battery charging bays in the battery swap station.

[0137] 2) Assuming that all batteries in the battery swap station are of the same specification and that the actual performance differences between batteries are not considered, the battery scheduling strategy is formulated as follows:

[0138] ① For batteries to be charged, charge them in order of SOC from low to high, and when there is a vacant position in the charging compartment, immediately add the battery to be charged;

[0139] ② For battery replacement service, if the number of fully charged batteries in the t+1 period Greater than or equal to the battery replacement demand D n,t+1 , then D b,t+1 A fully charged battery is used for battery replacement service; otherwise, After a fully charged battery is used for battery replacement service, the number of The battery is filled with rechargeable batteries in descending order of SOC, and the rechargeable batteries used for filling must meet the user's acceptable SOC lower limit;

[0140] ③ Compensation mechanism for battery replacement with a partially charged battery: Since using a partially charged battery to replace a battery is a loss for the user, a certain amount of compensation must be given to the user. The compensation fee is calculated as follows:

[0141] C d =e b,t (1-ρ b )

[0142] Where C d Compensation for battery replacement for partially charged batteries, e b,t is the battery replacement price in period t, ρ b is the discount rate for battery replacement, and its value depends on the power gap S between the partially charged battery and the fully charged battery. b,ρ ,Right now:

[0143] S b,ρ =1-S chg

[0144] Where S chg The SOC of the battery after the battery swap is completed;

[0145] S b,ρ The larger the ρ b The smaller it is, the Richards model is used to determine the size of the discount rate:

[0146]

[0147] Where r b is the discount rate coefficient, S′ b,ρ is the inflection point of the discount rate curve.

[0148] 3) The user must ultimately pay electricity charges and service fees based on the number of exchanged kilowatt-hours, where the number of exchanged kilowatt-hours ΔP bs Refers to the difference in power between the two batteries before and after battery replacement, that is:

[0149] ΔP bs =E0(S chg -S0)

[0150] The actual battery replacement cost of electric vehicles can be calculated as follows:

[0151] C bs =e b,t ΔP bs -C d

[0152] e b,t =e 0,t +e bs,t

[0153] Where C bs is the battery replacement fee, e bs,t It is the battery replacement service fee during period t.

[0154] 2.3 Comprehensive power supply cost

[0155] 1) Waiting time T of electric vehicle users after arriving at the charging / swapping station w Including the queue time T queue The duration of charging (charging or replacing) T cb , which is calculated as follows:

[0156] T w =T queue +T cb

[0157]

[0158]

[0159] Where k cb is the number of electric vehicles that need to be recharged in the charging / swapping station, c cb is the number of charging facilities (charging piles and battery swap stations) in the charging / battery swap station, T avr is the average charging time of electric vehicles, T cb,min is the shortest remaining charging time of the electric vehicles being served in the charging and swapping station, η ch and P ch are the charging efficiency and charging power of the charging pile respectively, T b The battery replacement time is basically the same for each vehicle, so it is a fixed value.

[0160] 2) The comprehensive charging cost includes the charging / swapping fee and the waiting time. Since the charging / swapping fee and the waiting time have different dimensions, they are normalized and then added together to obtain the comprehensive charging cost as shown below:

[0161]

[0162] Where C ev is the comprehensive charging cost of electric vehicles, c cb The charging fee (charging fee Ccs or battery replacement cost C bs ), C cb,max and C cb,min are the maximum and minimum charging costs in all charging / swapping stations, T w,max and T w,min They are the maximum and minimum waiting times in all charging / swapping stations respectively.

[0163] 3. Driving costs

[0164] 1) The analytic hierarchy process is used to calculate the weight scores ω1, ω2, and ω3 of different types of electric vehicles (electric private cars, electric taxis, and electric official vehicles) for driving distance, driving energy consumption, and driving time.

[0165] 2) For each road in the entire traffic path, the dynamic information w of each road t,ij The length, energy consumption, and time consumption are normalized separately (specifically, the maximum-minimum value normalization method can be used) to obtain the normalized length, energy consumption, and time consumption of each road section:

[0166]

[0167] Where, l r 、e r , t r are the actual values of the length, energy consumption and time consumption of road r, are the normalized values of the length, energy consumption, and time consumption of road r, respectively. min and l max are the minimum and maximum lengths of all roads in the traffic network, e min and e max are the minimum and maximum energy consumption of all roads in the traffic network, t min and t max are the minimum and maximum time consumptions for all roads in the traffic network respectively.

[0168] 3) In summary, for a certain route of an electric vehicle traveling from one node to another, its driving cost is:

[0169]

[0170] Where R is the total number of roads in the path, P total is the travel cost of the path.

[0171] 4) Path Planning

[0172] Based on the dynamic traffic network model, the Dijkstra algorithm is used to minimize the travel cost P totalAs the goal, the path of each electric vehicle to each charging / swapping station is planned, and the driving cost under the corresponding driving path is obtained as the driving cost D of each electric vehicle to each charging / swapping station. ev .

[0173] 4. Predict user charging / battery swapping decisions

[0174] Based on the established single-electric vehicle model and dynamic traffic network model, each electric vehicle in the traffic network is determined to require recharging based on its SOC. In this embodiment, a SOC below 25% is used to determine whether the corresponding electric vehicle requires recharging. For electric vehicles requiring recharging, such as private electric vehicles and official electric vehicles, the next destination is first determined to be home (residential area) or work (work area) based on their respective travel chains. If it is home or work, slow charging can be performed at the corresponding location, eliminating the need for charging / swapping decision prediction.

[0175] In addition to the above situations, the charging / replacement decision prediction is made for the electric vehicles that need to be recharged. In this embodiment, according to the comprehensive recharge cost C ev and driving cost D ev ,By simulating the user’s decision-making process by considering the cumulative prospect theory of the user’s bounded rationality, the charging / swapping decision results are obtained.

[0176] (1) Comprehensive value function of electric vehicle users’ decision-making

[0177] The comprehensive value function of electric vehicle users' charging and swapping decisions is as follows:

[0178] V com =ω d V d +ω c V c

[0179] Where V com is the comprehensive value, V d and V c are respectively the driving cost value and the comprehensive charging cost value, ω d and ω c are the driving cost weight coefficient and the comprehensive charging cost weight coefficient, ω d +ω c =1;

[0180] V d and V c The value functions of the two are as follows:

[0181]

[0182] Where Dref is the maximum driving cost acceptable to the user, C ref is the highest comprehensive electricity supplement cost acceptable to the user, α1 and β1 are the user’s ev <D ref or C ev <C ref ) and loss (D ev ≥D ref or C ev ≥C ref ) is the risk preference coefficient under the condition of , where α1>0, β1<1, λ1 is the loss aversion coefficient, and λ1≥1.

[0183] (2) Cumulative prospect value

[0184] For an electric car, the comprehensive value V of its driving to each charging / swapping station to replenish electricity is calculated. com After that, you can follow V com For size, select V com The highest one is used as the decision-making scheme for charging / swapping (i.e., which charging / swapping station to go to for charging). However, due to the specific D ref and C ref Unknown, so this embodiment uses the cumulative prospect theory to estimate the comprehensive value of each charging and swapping decision of the user:

[0185] 1) Obtain D through big data statistics ref and C ref the probability distribution of possible values;

[0186] 2) According to the cumulative prospect theory, the probability weight function is as follows:

[0187]

[0188] Where n is the number of gain situations and m is the number of loss situations, for example, D ref There are 10 possible values, 3 of which are greater than D ev , 7 values less than or equal to D ev , then for the driving cost, n is 3 and m is 7; f1 is the number of the profit situation, f2 is the number of the loss situation, p is the probability variable, p f1 Indicates the probability of the f1th profit situation occurring (corresponding to D ref or C ref The probability distribution of each value), It represents the probability of the f2th loss situation. When f1 and f2 are 0, p0=0, W + (p) is the user’s subjective perception probability of the profit situation, W -(p) is the user's subjective perceived probability of loss, is the decision weight when the user is in the f1th profit situation, is the decision weight when the user is in the f2th loss situation, γ and δ are the risk-return attitude coefficient and risk-loss attitude coefficient, respectively.

[0189] 3) Based on the above 1) and 2), calculate the cumulative prospect value of each alternative charging solution (which charging / swapping station the electric vehicle should go to for charging):

[0190]

[0191] Where, is the cumulative prospect value, and Respectively represent the f d1 The cost-benefit situation of the first driving and the d2 V under the condition of driving cost loss d , similarly and Respectively represent the f c1 The comprehensive cost-benefit of electricity supplement and the c2 V under the condition of comprehensive power supply cost loss c , n d and m d is the number of gain and loss cases in driving cost, f d1 and f d2 is the number of the profit and loss situations in the driving cost, n c and m c is the number of benefits and losses in the comprehensive electricity supplement cost, f c1 and f c2 The number of the profit and loss in the driving cost; the decision weight is calculated by the probability weight function When n d and n c Substitute n, m accordingly d and m c Substitute m, f accordingly d1 and f c1 Substitute f1, f d2 and f c2 Substitute into f2 accordingly.

[0192] In summary, when predicting the charging plan for each electric vehicle, based on the cumulative prospect value Size, select The largest charging / swapping station is used as the prediction result of the charging plan for each electric vehicle.

[0193] (3) Update of prediction results

[0194] For an electric vehicle that needs to be recharged, after the possible recharge options (which charging / battery swap station to drive to) are predicted using the above method, the charging / battery swap station or route selection may be updated due to changes in road conditions or charging / battery swap prices during the process of the electric vehicle driving to the charging / battery swap station. Therefore, this embodiment sets the following prediction restart conditions:

[0195] ① In the planned path, when there is a road with a driving speed v ij (t) reduced to or below the threshold;

[0196] ② Changes in charging service fees or battery replacement service fees;

[0197] Before the electric vehicle drives to the charging / battery swapping station, if any of the above conditions is met, the prediction of the user's charging / battery swapping decision will be restarted.

[0198] 5. Charging and swapping load prediction

[0199] After predicting the charging scheme for each electric vehicle in the traffic network, the corresponding charging / swapping load prediction results are obtained based on the predicted results and the power consumption of each charging / swapping station.

[0200] 1) For the charging load of each node in each time period, based on the prediction results of the above-mentioned charging scheme, the number of electric vehicles charging in the charging station at each node is counted, and the power consumption of the charging station is calculated by combining the power of the charging pile and the charging time.

[0201] 2) For the battery swap load at each node in each time period, based on the prediction results of the above-mentioned power replenishment plan, the number of batteries being charged in the battery swap station at each node is counted, and the power load of the battery swap station is calculated in combination with the battery charging power and charging time.

[0202] 2. Test and Verification

[0203] The aforementioned charging and swapping load prediction method uses a Monte Carlo method to sample electric vehicle travel data and determine the daily charging and swapping load within a specific region. Parameters for the single electric vehicle model and the dynamic traffic network model, including the type and number of electric vehicles, are set based on the test data used for simulation. The Monte Carlo method is used to extract information such as the time of the electric vehicle's first trip, the starting SOC, the SOC at which the charging demand arises, and the starting node.

[0204] like Figure 2 、 Figure 3As shown in the figure, they are the road network topology map and functional area distribution map of the study area respectively. In the figure, there are 3 charging stations and 1 battery swap station marked A, B, C, and D respectively. The number of charging piles at charging stations A, B, and C are 16, 20, and 20 respectively. They are all fast charging piles with a charging power of 60kW. The charging piles for households and units are all slow charging piles with a charging power of 21kW, and the charging efficiency of all charging piles is 0.95. Assume that the battery swap station is large, with three battery swap positions, 100 batteries available for replacement, and 60 battery charging compartments with a rated charging power of 25kW. The total number of electric vehicles is set to 800, with the proportions of electric private cars, taxis, and official vehicles being 50%, 40%, and 10% respectively. Among them, 70% of users are charging electric vehicles, and the remaining 30% are charging and swapping electric vehicles. The battery capacity of all electric vehicle users is 100kWh. The weights of the three types of electric vehicles during route planning are shown in Table 1. The charging and swapping prices of each station are set as follows Figure 4 shown.

[0205] Table 1: Path planning weight table

[0206]

[0207] The load forecast results of each charging / swapping station in the study area are as follows: Figure 5 As shown in the figure, charging station A's peak charging times are between 12:00 PM and 2:00 PM and 8:00 PM and 11:00 PM; charging station B's peak charging times are between 11:00 AM and 2:00 PM and 6:00 PM and 8:00 PM; and charging station C's peak charging times are between 4:00 PM and 8:00 PM and 10:00 PM and 11:00 PM. These time periods correspond to the charging behavior of office workers during rush hour and the charging behavior of electric taxis during breaks and after operations, thus aligning with the peak charging times for electric vehicle users in real life. In addition to these load peaks, charging loads at each station are also high in the early morning hours. This is because most taxi drivers leave work late and generally choose to charge immediately after get off work to meet their driving needs the next day. During this time, a large amount of charging load enters the grid, causing an increase in nighttime load. From the above analysis, it can be seen that the charging load of each charging station shows the characteristics of multiple load peaks, but the time periods of the charging peaks are different. On the one hand, this is because the functional areas where the charging stations are located are different, and the travel characteristics of different types of electric vehicles in each functional area are also different, which leads to differences in charging demand. On the other hand, the service fees of each charging station are priced differently. The price difference makes users more inclined to choose charging stations with lower fees, which leads to different peak charging load periods for each charging station.

[0208] Furthermore, the load trends at battery swap stations in the early morning are similar to those at charging stations. However, during the daytime hours of 10:00 AM to 11:00 PM, the load remains high and shows a slowly increasing trend. This is due to the different operating models of battery swap stations. When users choose to recharge at a battery swap station, it takes only five minutes to complete the recharge process, while at a charging station, it takes several times longer to obtain the same amount of power as a battery swap. This higher service speed leads to a gradual increase in the number of recharged batteries at battery swap stations, ultimately making their charging load higher than that of charging stations.

[0209] III. Devices, Storage Media, and Program Products

[0210] 1. Based on the same inventive concept as the above-mentioned charging and swapping load prediction method, the present application also provides an electronic device, which includes a processor and a memory, and the memory stores computer-readable code, wherein when the computer-readable code is executed by the processor, the charging and swapping load prediction method of the present invention is implemented.

[0211] The memory includes a non-volatile storage medium and internal memory; the non-volatile storage medium can store an operating system and computer-readable code. The computer-readable code includes program instructions that, when executed, enable the processor to execute the charging and swapping load prediction method. The processor is used to provide computing and control capabilities to support the operation of the entire electronic device. The memory provides an environment for the operation of the computer-readable code in the non-volatile storage medium. When executed by the processor, the computer-readable code enables the processor to execute the charging and swapping load prediction method.

[0212] It should be understood that the processor may be a central processing unit, other general-purpose processors, digital signal processors, application-specific integrated circuits, field programmable gate arrays or other programmable logic devices, transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or any conventional processor.

[0213] 2. This application also provides a readable storage medium, which can be the internal storage unit of the electronic device described in the aforementioned embodiment, such as the hard disk or memory of the computer device. The readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart memory card, secure digital card, etc. equipped with the electronic device.

[0214] 3. The present application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the charging and swapping load prediction method of the present invention.

[0215] The present invention is not limited to the above-mentioned embodiments. Any obvious improvement, replacement or modification that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the scope of protection of the present invention.

Claims

1. A charging and swapping load prediction method, characterized by: Build a single electric vehicle model and a dynamic traffic network model; Calculate the comprehensive charging cost based on charging cost, battery replacement cost, and charging / battery replacement waiting time; Calculate driving costs based on driving distance, driving energy consumption, and driving time; Predict the charging and swapping decision results of each electric vehicle based on driving cost and comprehensive charging cost; Based on the charging and swapping decision results, the charging and swapping load is calculated.

2. The charging and swapping load prediction method according to claim 1, characterized in that: The method for predicting the charging and swapping decision results of each electric vehicle based on the driving cost and the comprehensive charging cost is as follows: Calculate the comprehensive value of driving to each charging / swapping station for recharging: V com =ω d V d +ω c V c Where V com is the comprehensive value, V d and V c are respectively the driving cost value and the comprehensive charging cost value, ω d and ω c are the driving cost weight coefficient and the comprehensive charging cost weight coefficient, ω d +ω c =1;D ev is the driving cost to the charging / swapping station, C ev is the comprehensive electricity supplement cost, D ref is the maximum driving cost acceptable to the user, C ref is the maximum comprehensive electricity replenishment cost acceptable to the user, α1 and β1 are the risk preference coefficients of the user under profit and loss conditions, respectively, where α1>0, β1<1, λ1 is the loss aversion coefficient, and λ1≥1; Select V com The highest one is taken as the charging and swapping decision result.

3. The charging and swapping load prediction method according to claim 1, characterized in that: The method for predicting the charging and swapping decision results of each electric vehicle based on the driving cost and the comprehensive charging cost is as follows: Calculate the cumulative prospect value of driving to each charging / swapping station for recharging: Where, is the cumulative prospect value, ω d and ω c are the driving cost weight coefficient and the comprehensive charging cost weight coefficient, ω d +ω c =1,n d and m d is the number of gain and loss cases in driving cost, f d1 and f d2 is the number of the profit and loss in the driving cost, where the profit is D ev <D ref , the loss situation is D ev ≥D ref , D ev D is the driving cost to the charging / swapping station, ref is the maximum driving cost acceptable to the user; n c and m c is the number of benefits and losses in the comprehensive electricity supplement cost, f c1 and f c2 is the number of the profit and loss situations in the driving cost, where the profit situation is C ev <C ref , the loss situation is C ev ≥C ref , C ev is the comprehensive electricity supplement cost, C ref The highest comprehensive electricity supplement cost acceptable to users; D ref and C ref Obtain possible values and corresponding probability distribution through past data statistics; and Respectively represent the f d1 The cost-benefit situation of the first driving and the d2 The driving cost value V under the condition of driving cost loss d , and Respectively represent the f c1 The comprehensive cost-benefit of electricity supplement and the c2 The comprehensive electricity supplement cost value V under the condition of comprehensive electricity supplement cost loss c , V d and V c By calculating as follows: Where α1 and β1 are the risk preference coefficients of users in the case of gains and losses, respectively, where α1>0, β1<1, λ1 is the loss aversion coefficient, and λ1≥1; By calculating as follows: In the formula, n, m, f1, and f2 are substituted into n d 、m d 、f d1 、f d2 Calculated and Substitute n, m, f1, and f2 into n respectively c 、m c 、f c1 、f c2 Calculated and p is the probability variable, represents the probability of the f1th profit situation occurring, represents the probability of the f2th loss situation occurring, W + (p) is the user’s subjective perception probability of the profit situation, W - (p) is the user's subjective perceived probability of loss, is the decision weight when the user is in the f1th profit situation, is the decision weight when the user is in the f2th loss situation, γ and δ are the risk-return attitude coefficient and risk-loss attitude coefficient respectively; and According to D ref The probability distribution of possible values is obtained, and According to C ref The probability distribution of possible values is obtained; choose The highest one is taken as the charging and swapping decision result.

4. The charging and swapping load prediction method according to claim 2 or 3, characterized in that: The comprehensive electricity supplement cost is: Where C cb C is the electricity supplement fee. cb,max and C cb,min are the maximum and minimum charging costs in all charging / swapping stations, T w T is the waiting time for charging and swapping. w,max and T w,min The maximum and minimum waiting times in all charging / swapping stations respectively; Electricity replenishment fee C cb The charging cost C cs or battery replacement cost C bs ,in: Charging fee C cs for: C cs =E0(e 0,t +e cs,t )(S end -S0) Where E0 is the battery capacity of the electric vehicle, e 0,t is the electricity price during period t, e cs,t is the charging service fee for period t, S end is the SOC of the electric vehicle after charging is completed, and S0 is the SOC of the electric vehicle before charging; Battery replacement cost C bs for: C bs =e b,t ΔP bs -C d And b,t =and 0,t +e bs,t Where, e b,t is the battery replacement price during period t, e bs,t is the battery replacement service fee during period t, C d Compensation for battery replacement for partially charged batteries is calculated as follows: C d =e b,t (1-p b ) Where, ρ b is the discount rate for battery replacement, which is calculated as follows: Where S b,ρ =1-S chg S is the power gap between a partially charged battery and a fully charged battery. chg is the SOC of the battery replacement, r b is the discount rate coefficient, S′ b,ρ is the inflection point of the discount rate curve; Charging and swapping waiting time T w The calculation is as follows: T w =T queue +T cb Where, T queue is the queue time, T cb is the charging time, k cb is the number of electric vehicles that need to be recharged in the charging / swapping station, c cb is the number of charging facilities in the charging / swapping station, T avr is the average charging time of electric vehicles, T cb,min is the shortest remaining charging time of the electric vehicles being served in the charging and swapping station, η ch and P ch are the charging efficiency and charging power of the charging pile respectively, T b Time for battery replacement.

5. The charging and swapping load prediction method according to claim 2 or 3, characterized in that: The driving cost is calculated as follows: The travel cost for a path is: Where R is the total number of roads in the route, ω1, ω2, and ω3 are the weighted scores of driving distance, driving energy consumption, and driving time, respectively. are the normalized values of the length, energy consumption, and time consumption of road r, respectively; Use Dijkstra algorithm to minimize the travel cost P total As the goal, the path of each electric vehicle to each charging / swapping station is planned, and the driving cost of the planned path is used as the driving cost D of the electric vehicle to the charging / swapping station. ev .

6. The charging and swapping load prediction method according to claim 1, characterized in that: The single electric vehicle model includes: the probability distribution of the first trip time of each type of electric vehicle, the probability distribution of the length of stay, the trip chain, and the energy consumption model, wherein the energy consumption model is: Where, E com is the comprehensive energy consumption per unit mileage of electric vehicles, E T (T en ) is the ambient temperature of the electric vehicle en Energy consumption per mile under T (20℃) is the energy consumption per unit mileage of electric vehicles at an ambient temperature of 20℃, E v is the energy consumption per unit mileage of the electric vehicle at speed v; Speed and energy consumption of electric vehicles: Where, E v is the energy consumption per unit mileage of the electric vehicle at speed v, α, β, ω, λ are speed energy consumption coefficients; Between ambient temperature and electric vehicle energy consumption: Where, E T The electric vehicle is at an ambient temperature T en Energy consumption per mile under a n is the temperature energy consumption coefficient.

7. The charging and swapping load prediction method according to claim 6, characterized in that: The dynamic traffic network model is: Where G t Represents the traffic network topology diagram under time period t, including N, E, W t Three elements, N is the set of all road nodes in the traffic network, N node is the total number of nodes, E is the set of all roads in the traffic network, (i, j) represents the road between node i and node j, W t is the dynamic road network information set in time period t, w t,ij is the dynamic information of road ij in time period t, including the length of road ij and the time and energy consumption required to complete the journey; The time consumption is calculated based on the road length and the road speed, and the energy consumption is calculated based on the road length, the road speed and the energy consumption model; the road speed is calculated as follows: Where, v ij (t) represents the driving speed of road ij during time period t, i and j both represent road nodes in the traffic network, road ij represents the road between node i and node j, v ij,max represents the zero flow speed of road ij, Q ij (t) represents the traffic volume of road ij in period t, μ v is the speed coefficient, C ij is the traffic capacity of road ij, k1, k2, k3 are the adaptive coefficients of the road.

8. A computer device, characterized in that: including memory and processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the charging and swapping load prediction method according to any one of claims 1 to 7 when executing the computer program.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor executes the charging and swapping load prediction method according to any one of claims 1 to 7.

10. A computer program product, characterized in that: It includes a computer program, which, when executed by a processor, implements the charging and swapping load prediction method according to any one of claims 1 to 7.

Images