Electric vehicle charging scheduling method and device based on public charging station

Abstract

The invention discloses an electric vehicle charging scheduling method and device based on a public charging station, and relates to the field of electric vehicle charging scheduling. The method comprises the following steps: acquiring information data of each electric vehicle in a target public charging station; based on a charging pile distribution mechanism of a sliding window mechanism, charging piles are distributed according to all the information data, and a distribution result is obtained and used for scheduling the charging power of each electric vehicle, so that the overall operation efficiency of the target public charging station is improved; an NSGA-II algorithm and an Entropy-TOPSIS method are adopted, charging scheduling optimization processing is carried out according to the distribution result on the basis of an optimization model, an optimized charging scheduling scheme is obtained, and the optimized charging scheduling scheme is used for achieving renewable energy consumption and real-time charging scheduling between V2G electric vehicles and enabling equalization optimization; according to the invention, effective charging management of the electric vehicle is realized, integration of renewable energy sources is enhanced, and equalization optimization is realized.

Description

Technical Field

[0001] The present application relates to the field of electric vehicle charging scheduling, and particularly to an electric vehicle charging scheduling method and device based on a public charging station. Background Art

[0002] In recent years, the popularization speed of electric vehicles has been accelerating rapidly globally. Therefore, in order to meet the rapidly growing charging demand, charging infrastructure, especially public charging stations, is being widely promoted. However, uncontrolled large-scale charging operations pose significant challenges to the safe operation of the power grid. Therefore, it is necessary to perform effective charging management on electric vehicles to reduce the power grid overload risk and the charging cost of electric vehicle users, and to strengthen the integration of renewable energy. Summary of the Invention

[0003] The purpose of the present application is to provide an electric vehicle charging scheduling method and device based on a public charging station, which can achieve effective charging management of electric vehicles, strengthen the integration of renewable energy, and achieve balanced optimization.

[0004] To achieve the above purpose, the present application provides the following solutions:

[0005] In a first aspect, the present application provides an electric vehicle charging scheduling method based on a public charging station, including:

[0006] Obtaining information data of each electric vehicle in a target public charging station; the information data includes: charging request data, current state of charge data, expected final state of charge data, expected departure time, and minimum acceptable state of charge data at departure; the charging request data includes: selecting a fast charging pile and selecting a slow charging pile; the maximum power of the fast charging pile is greater than a preset power; the maximum power of the slow charging pile is less than the preset power; the target public charging station is connected to a microgrid powered by renewable energy, and the distribution network is used as an auxiliary power source;

[0007] Based on a charging pile allocation mechanism of a sliding window mechanism, performing charging pile allocation according to all the information data to obtain an allocation result; the allocation result is used to schedule the charging power of each electric vehicle to improve the overall operation efficiency of the target public charging station;

[0008] The NSGA-II algorithm and the Entropy-TOPSIS method are adopted to perform optimization processing on the charging schedule based on the optimization model according to the allocation result, and an optimized charging schedule plan is obtained; the optimized charging schedule plan is used to realize real-time charging scheduling between renewable energy consumption and V2G electric vehicles, and to realize enabling balance optimization; the optimization model is determined according to the charging behavior database of electric vehicles in the target public charging station, and considering the uncertainty of charging time and charging characteristics; the optimization model includes an objective function and constraint conditions; the charging behavior database includes multiple data samples; each of the data samples includes: battery capacity, starting charging state, ending charging state, starting charging time, ending charging time, and parking time.

[0009] In a second aspect, the present application provides an electric vehicle charging scheduling device based on a public charging station, including:

[0010] An information data acquisition module, configured to acquire information data of each electric vehicle in the target public charging station; the information data includes: charging request data, current state of charge data, expected final state of charge data, expected departure time, and minimum acceptable state of charge data at departure; the charging request data includes: selecting a fast charging pile and selecting a slow charging pile; the maximum power of the fast charging pile is greater than a preset power; the maximum power of the slow charging pile is less than the preset power; the target public charging station is connected to a microgrid for renewable energy generation, and the distribution network is used as an auxiliary power source;

[0011] An allocation module, configured to perform charging pile allocation according to all the information data based on a charging pile allocation mechanism based on a sliding window mechanism, and obtain an allocation result; the allocation result is used to schedule the charging power of each electric vehicle to improve the overall operation efficiency of the target public charging station;

[0012] An optimization processing module, configured to adopt the NSGA-II algorithm and the Entropy-TOPSIS method to perform optimization processing on the charging schedule based on the optimization model according to the allocation result, and obtain an optimized charging schedule plan; the optimized charging schedule plan is used to realize real-time charging scheduling between renewable energy consumption and V2G electric vehicles, and to realize enabling balance optimization; the optimization model is determined according to the charging behavior database of electric vehicles in the target public charging station, and considering the uncertainty of charging time and charging characteristics; the optimization model includes an objective function and constraint conditions; the charging behavior database includes multiple data samples; each of the data samples includes: battery capacity, starting charging state, ending charging state, starting charging time, ending charging time, and parking time.

[0013] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0014] The present application provides an electric vehicle charging scheduling method and device based on a public charging station. By constructing an energy-saving charging behavior database of electric vehicles in a target public charging station and considering the time uncertainty and charging characteristics of charging, the electric vehicle charging requests are incorporated into the necessary process of orderly charging scheduling, while meeting the concerns of the charging station, electric vehicles, and the grid side. And based on the charging pile allocation mechanism of the sliding window mechanism, the charging power of each electric vehicle is scheduled to maximize the overall operation efficiency of the target public charging station. The NSGA-II algorithm and the Entropy-TOPSIS method are used for charging scheduling optimization to achieve the balanced optimization of multiple optimization objectives. The present application can effectively manage the charging of electric vehicles, strengthen the integration of renewable energy, and achieve balanced optimization. Description of the Drawings

[0015] Figure 1 is a flowchart of the electric vehicle charging scheduling method based on a public charging station;

[0016] Figure 2 is a schematic diagram corresponding to the information included in the data sample; where Figure 2 in (a) is the battery capacity information; Figure 2 in (b) is the SOC distribution at the start and end of electric vehicle charging; Figure 2 in (c) is the battery capacity change distribution;

[0017] Figure 3 is a schematic diagram of the electric vehicle arrival time distribution; where Figure 3 in (a) is the arrival time distribution diagram of fast-charging electric vehicles on weekdays; Figure 3 in (b) is the arrival time distribution diagram of slow-charging electric vehicles on weekdays; Figure 3 in (c) is the arrival time distribution diagram of fast-charging electric vehicles on weekends; Figure 3 in (d) is the arrival time distribution diagram of slow-charging electric vehicles on weekends;

[0018] Figure 4 is a distribution diagram of electric vehicle charging and parking time; where Figure 4 in (a) is the distribution diagram of charging and parking time of fast-charging electric vehicles; Figure 4 in (b) is the distribution diagram of charging and parking time of slow-charging electric vehicles; Figure 4 in (c) is the cumulative probability distribution diagram of fast-charging parking time and idle time; Figure 4 in (d) is the cumulative probability distribution diagram of slow-charging parking time and idle time;

[0019] Figure 5 is a schematic diagram of the microgrid structure;

[0020] Figure 6Schematic diagram for comparison of the ordered charging and disordered charging processes and segmentation of the charging process;

[0021] Figure 7 Schematic diagram for multi-stage constant current charging;

[0022] Figure 8 Schematic diagram of the flow of the electric vehicle fast charging pile allocation method;

[0023] Figure 9 Schematic diagram of the flow of the electric vehicle slow charging pile allocation method;

[0024] Figure 10 Schematic diagram of the sliding window mechanism;

[0025] Figure 11 Schematic diagram of the NSGA-II process;

[0026] Figure 12 Flow chart of the Entropy-TOPSIS implementation method. Specific implementation manner

[0027] Aiming at the problem that the charging pile may cause the instability of the microgrid, a real-time charging scheduling method for electric vehicles that realizes the consumption of renewable energy and V2G in a public charging station in the microgrid is proposed. First, a database of energy-saving charging and discharging behaviors of electric vehicles is constructed, considering the time uncertainty and charging characteristics of fast charging and slow charging on weekdays and weekends. Then, a charging pile allocation mechanism is proposed to schedule the charging power of each electric vehicle to maximize the overall operation efficiency of the charging station. Furthermore, a microgrid system model considering efficient V2G interaction and renewable energy integration is established. Finally, an enabling charging scheduling scheme is proposed to achieve the balanced optimization of multiple optimization goals.

[0028] In an exemplary embodiment, as Figure 1 shown, a charging scheduling method for electric vehicles based on a public charging station is provided. This method is executed by a computer device, and specifically can be executed alone by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiments of the present application, taking the server to which this method is applied as an example for illustration, the method includes the following steps.

[0029] Step 100: Obtain the information data of each electric vehicle in the target public charging station. The information data includes: charging request data, current state of charge data, expected final state of charge data, expected departure time, and minimum acceptable state of charge data at departure; the charging request data includes: selecting a fast charging pile and selecting a slow charging pile; the maximum power of the fast charging pile is greater than the preset power; the maximum power of the slow charging pile is less than the preset power; the target public charging station is connected to a microgrid with renewable energy generation, and the distribution network is used as an auxiliary power source.

[0030] Step 200: Based on the charging pile allocation mechanism of the sliding window mechanism, allocate charging piles according to all the information data to obtain an allocation result. The allocation result is used to schedule the charging power of each electric vehicle to improve the overall operation efficiency of the target public charging station.

[0031] Step 300: Adopt the NSGA-II algorithm and the Entropy-TOPSIS method to perform optimization processing on the charging schedule based on the optimization model according to the allocation result to obtain an optimized charging schedule plan. The optimized charging schedule plan is used to achieve real-time charging scheduling between renewable energy consumption and V2G electric vehicles, realizing enabling balance optimization; the optimization model is determined according to the charging behavior database of electric vehicles in the target public charging station, and considering the uncertainty of charging time and charging characteristics; the optimization model includes an objective function and constraint conditions; the charging behavior database includes multiple data samples; each data sample includes: battery capacity, starting charging state, ending charging state, starting charging time, ending charging time, and parking time.

[0032] The objective function includes: the load fluctuation of the distribution network, the charging cost of electric vehicles, and the energy consumption difference of renewable energy generation.

[0033] In the power fluctuation of the distribution network, the expression of the distribution network load fluctuation is:

[0034]

[0035] In the charging tariff of electric vehicle users, the expression of the charging cost of electric vehicles is:

[0036]

[0037] In the real-time energy consumption difference, the expression of the energy consumption difference of renewable energy generation is:

[0038]

[0039] Among them, DNLE is the distribution network load fluctuation; T is the scheduling period divided by the charging and discharging process of the electric vehicle; P Load (t) is the total load of the distribution network in the t time period; is the average load of the distribution network during the charging and discharging process of the electric vehicle; τ is the moment in the t time period; p is the moment serial number in the t time period; EVCC is the charging cost of the electric vehicle; C(t) is the charging cost in the t time period; RECD is the energy consumption difference; P diff (t) is the difference between the renewable energy power generation and the charging and discharging power of the electric vehicle in the t time period.

[0040] The constraints include: the capacity constraint of the power distribution equipment, the output power constraint of the microgrid, and the remaining charge constraint.

[0041] The expression of the capacity constraint of the power distribution equipment is:

[0042]

[0043] The expression of the output power constraint of the microgrid is:

[0044]

[0045] The expression of the remaining charge constraint is:

[0046] 0 ≤ SOC ev ≤ 100%.

[0047] Among them, P DN is the capacity of the power distribution equipment; is the maximum discharge power of the power distribution equipment; is the maximum charging power of the power distribution equipment; is the predicted output power of the wind power; is the rated wind power; is the predicted output power of the photovoltaic power; is the photovoltaic wind power; SOC ev is the remaining charge of the electric vehicle.

[0048] In one embodiment, the charging pile allocation mechanism includes that the energy information scheduling center adopts the minimum power allocation method and the random power allocation method for electric vehicles that select fast charging piles, and the energy information scheduling center adopts the random power allocation method for electric vehicles that select slow charging piles.

[0049] Among them, the process of the energy information scheduling center adopting the minimum power allocation method and the random power allocation method for electric vehicles that select fast charging piles to allocate charging piles specifically includes:

[0050] The energy information scheduling center judges whether there is an idle charging pile according to the usage situation of the charging pile, and obtains the first judgment result.

[0051] If the first judgment result is yes, then judge whether there is a charging pile that can meet the first preset condition, and obtain the second judgment result; the first preset condition is to achieve the predicted final state of charge data.

[0052] If the second judgment result is yes, the electric vehicle participates in orderly charging, and the charging pile with the lowest output power or a randomly selected charging pile that meets the preset conditions is allocated to the electric vehicle for charging.

[0053] If the second judgment result is negative, then it is judged whether there is a charging pile that can meet the second preset condition, and a third judgment result is obtained; the second preset condition is to reach the minimum acceptable state of charge data at the time of departure.

[0054] If the third judgment result is positive, adopt the multi-stage constant current fast charging method, charge at the maximum power under the charging power limit, do not participate in the orderly charging, and allocate the charging pile with the lowest output power or randomly select a charging pile that meets the second preset condition to the electric vehicle for charging.

[0055] If the third judgment result is negative, then identify all the charging piles in the target public charging station, and judge whether there is a charging pile that can meet the first preset condition after the electric vehicle waits, and obtain a fourth judgment result.

[0056] If the fourth judgment result is positive, then choose to participate in the orderly charging and allocate the first idle charging pile after waiting to the electric vehicle.

[0057] If the fourth judgment result is negative, then after the electric vehicle waits, judge whether there is a charging pile that can meet the second preset condition, and obtain a fifth judgment result.

[0058] If the fifth judgment result is positive, then do not participate in the orderly charging and allocate the first idle charging pile after waiting to the electric vehicle.

[0059] If the fifth judgment result is negative, then the electric vehicle gives up charging.

[0060] If the first judgment result is positive or negative, then return "If the third judgment result is negative, then identify all the charging piles in the target public charging station, and judge whether there is a charging pile that can meet the first preset condition after the electric vehicle waits, and obtain a fourth judgment result".

[0061] The energy information dispatching center adopts a random power distribution method for electric vehicles that select slow charging piles during the process of charging pile allocation, which specifically includes:

[0062] The energy information dispatching center judges whether there is an idle charging pile according to the usage situation of the charging pile, and obtains a sixth judgment result; if the sixth judgment result is positive, then the electric vehicle participates in the orderly charging, and randomly allocates the charging pile to the electric vehicle for charging; if the sixth judgment result is negative, then judge whether there is a charging pile that meets the third preset condition after the electric vehicle waits, and obtain a seventh judgment result. The third preset condition is that there is an idle charging pile before the expected departure time and it can reach the expected final state of charge data.

[0063] If the result of the seventh judgment is yes, the electric vehicle participates in orderly charging, and the first available charging pile is allocated to the electric vehicle so that the electric vehicle waits before starting to charge; if the result of the seventh judgment is no, it is judged whether there is a charging pile that meets the second preset condition after the electric vehicle waits, and the eighth judgment result is obtained.

[0064] If the result of the eighth judgment is yes, the electric vehicle does not participate in orderly charging, and the first available charging pile is allocated to the electric vehicle so that the electric vehicle waits before starting to charge; if the result of the eighth judgment is no, charging is abandoned.

[0065] In one embodiment, the NSGA-II algorithm and the Entropy-TOPSIS method are used to perform charging scheduling optimization processing based on the optimization model according to the allocation result, and an optimized charging scheduling scheme is obtained, specifically including:

[0066] Based on the optimization model, the NSGA-II algorithm is used to perform crossover and mutation processing according to the allocation result and the objective function to obtain a Pareto solution set. The Pareto solution set is the Pareto optimal solution set. The Entropy-TOPSIS method is used to perform charging scheduling optimization processing according to the Pareto solution set to obtain an optimized charging scheduling scheme.

[0067] As an optional implementation manner, based on the optimization model, the NSGA-II algorithm is used to perform crossover and mutation processing according to the allocation result and the objective function based on the optimization model to obtain a Pareto solution set, specifically including:

[0068] Determine the initial parameters; the initial parameters include: the size μ of the population, the set number of iterations Gen, the crossover probability, and the mutation probability; the population includes multiple individuals; the individuals are the respective charging piles in the allocation result; the size of the population is the total number of individuals that make up the individuals.

[0069] Determine the fitness function according to the objective function; use the fitness function to determine the fitness value of each individual, and select and determine the parents in the population based on the fitness value.

[0070] Based on the parents, perform crossover and mutation according to the crossover probability and the mutation probability, and use the fitness function to perform iterative optimization based on the non-dominated sorting and crowding distance assignment criteria to obtain a Pareto solution set.

[0071] Among them, the gen-th iteration process is:

[0072] Use the fitness function to determine the fitness value of each individual in the population in the gen-th iteration, and select and determine the parents in the population in the gen-th iteration based on the fitness value.

[0073] Perform crossover on the parents in the gen-th iteration using the crossover probability in the gen-th iteration to obtain the offspring λ in the gen-th iteration.

[0074] Modify the offspring in the gen-th iteration using the mutation probability in the gen-th iteration to obtain the modified offspring in the gen-th iteration.

[0075] Calculate the fitness value in the gen-th iteration according to the fitness function and the modified offspring in the gen-th iteration.

[0076] Determine whether the gen-th iteration reaches the set number of iterations Gen; if so, use the modified offspring in the gen-th iteration as the Pareto solution set; if not, update and combine the modified offspring in the gen-th iteration based on the non-dominated sorting and crowding distance assignment criteria according to the fitness value in the gen-th iteration, and use the updated and combined modified offspring as the population in the (gen + 1)-th iteration.

[0077] As an alternative implementation, use the Entropy-TOPSIS method to perform charging scheduling optimization processing according to the Pareto solution set to obtain an optimized charging scheduling plan, specifically including:

[0078] Perform logarithmic transformation and normalization transformation on the Pareto solution set to obtain the pre-converted Pareto value; perform normalization processing on the pre-converted Pareto value and calculate the entropy value; determine the weight according to the calculated entropy, and perform reverse logarithmic transformation and normalization processing on the weight value to obtain the normalized processing data.

[0079] Determine the normalized matrix of TOPSIS according to the normalized processing data and the pre-Pareto value; respectively determine the Euclidean distances corresponding to the ideal solution and the negative ideal solution based on the normalized matrix of TOPSIS.

[0080] Calculate the closeness based on the Euclidean distance; the calculation formula for the closeness is:

[0081]

[0082] where closeness is the closeness; distance worst is the Euclidean distance of the negative ideal solution; distance best is the Euclidean distance of the ideal solution.

[0083] Determine the optimized charging scheduling plan according to the closeness.

[0084] In practical applications, the technical concept of the method mentioned in this application includes the following steps.

[0085] Build an energy-saving charging and discharging behavior database for electric vehicles.

[0086] Based on the real charging data of some charging stations in a certain city, a charging behavior database for electric vehicles was established. The charging behavior database includes multiple data samples. Each data sample contains battery capacity, starting state of charge (SOC), ending state of charge (end SOC), start charging time, end charging time, and parking time. As Figure 2 shown.

[0087] The battery capacity of energy-saving electric vehicles is mainly distributed between 45 kWh and 62 kWh, as Figure 2 (a) shown. The starting SOC, ending SOC, and SOC change are as Figure 2 (b) and Figure 2 (c) shown. It can be seen that even when the remaining SOC is still relatively high, electric vehicles show a charging trend, and the final SOC of the vast majority of electric vehicles is 100%. The SOC change during a single charge is mainly in the range of 20% to 60%.

[0088] Start the day at 4 am and divide the day into 96 segments, each segment with an interval of 15 minutes. In the case of having a charging pile, assume the arrival time at the charging station as the start charging time. The arrival time distribution of electric vehicles is as Figure 3 shown. Among them, Figure 3 (a) is the arrival time distribution of fast-charging electric vehicles on weekdays; Figure 3 (b) is the arrival time distribution of slow-charging electric vehicles on weekdays; Figure 3 (c) is the arrival time distribution of fast-charging electric vehicles on weekends; Figure 3 (d) is the arrival time distribution of slow-charging electric vehicles on weekends.

[0089] From Figure 3 it can be seen that the arrival time of fast-charging electric vehicles shows a unimodal trend. The arrival time of slow-charging electric vehicles on weekdays shows two peaks in the morning and evening; on weekends, there is only one peak at night, and private cars dominate the charging vehicles.

[0090] The statistical results reveal the randomness of the arrival time of electric vehicles at the charging station and the differences in the charging behaviors of electric vehicles on weekdays and weekends. To reduce the influence of outliers, the Gaussian Mixture Model (GMM) is used for curve fitting and residual analysis. It can be seen that the fitting curve has a good approximation effect. Normalize the values obtained by GMM fitting to obtain the probability distribution of the arrival of electric vehicles (EV) within a day.

[0091] Electric vehicles usually start charging immediately after arriving at the charging station and continue to park for some time after charging is completed. The parking time after charging is defined as idle time here. The distributions of the charging and parking times of electric vehicles are as Figure 4 shown. Among them, Figure 4 (a) is the distribution of the charging and parking times of fast-charging electric vehicles; Figure 4 (b) is the distribution of the charging and parking times of slow-charging electric vehicles; Figure 4 (c) is the cumulative probability distribution of the fast-charging parking time and idle time; Figure 4 (d) is the cumulative probability distribution of the slow-charging parking time and idle time.

[0092] It can be seen that most fast-charging electric vehicles can reach the target SOC within 3 hours and stay for more than 1 hour after charging is completed. Most slow-charging electric vehicles can reach the target SOC within 8 hours and stay for more than 10 hours after charging is completed.

[0093] The idle time after the electric vehicle's charging is completed provides the possibility for charging scheduling. By using the idle time, the charging load of electric vehicles can be transferred to a certain extent, reducing the impact of electric vehicle charging on the power grid.

[0094] An optimization model is proposed.

[0095] The charging station studied in this application is connected to a microgrid powered by renewable energy, as Figure 5 shown. The Energy Information Deploy Centre (EIDC) is responsible for regulating the power flow between units, and the distribution network serves as an auxiliary power source. When the power supply exceeds the power demand, the excess electrical energy is consumed by the basic power consumption of the charging station or transferred to other nodes of the distribution network. As Figure 6 shown, the charging and discharging process of electric vehicles can be divided into T scheduling cycles, that is, T is the scheduling cycle divided for the charging and discharging process of electric vehicles, with an interval of Δt. The power balance equation is composed of:

[0096]

[0097] Among them, P DN is the total power generation; P WT is the actual wind power generation; P PV is the actual photovoltaic power generation; P load is the basic load within the same distribution network; refers to the electric vehicle charging of the total load, that is, the rated power of charging; P dis refers to the total load of electric vehicle discharging, that is, the actual discharging power.

[0098] During the charging process, the actual charging power on the power supply side is the rated power During the discharging process, the actual discharging power P dis is variable.

[0099] Wind turbine (WT) power prediction model:

[0100]

[0101] Wherein, is the predicted output power of wind power; v in , v out , v * are the incoming wind speed, the outgoing wind speed, and the rated wind speed respectively; is the rated wind power.

[0102] Photovoltaic (PV) power prediction model:

[0103]

[0104] Wherein, is the predicted output power of PV; η PV is the PV conversion efficiency; G is the solar radiation intensity; A is the contact area.

[0105] In the electric vehicle charging and discharging load, P EV represents the total electric vehicle charging and discharging load at time t:

[0106]

[0107] The relationship between the actual charging and discharging power and the rated power of the electric vehicle within the t time period is:

[0108]

[0109] Wherein, η cha is the charging efficiency; η dis is the discharging efficiency. P cha is the actual charging power; is the rated power of discharging.

[0110]

[0111] Wherein, P ev，cha , P ev，dis are the actual charging and discharging powers of the nth electric vehicle respectively. Similarly, each EV satisfies:

[0112]

[0113] Wherein, respectively represent the rated charging and discharging powers of a single electric vehicle.

[0114] The main control objectives of this application are to mitigate the power fluctuations in the distribution network, reduce the charging costs of electric vehicle users, and accommodate more renewable energy generation.

[0115] In the load fluctuations of the distribution network system:

[0116] The total load P of the distribution network within the time period t Load (t), abbreviated as P Load :

[0117] P load = P load + P EV (11)

[0118] The distribution network load fluctuation (DNLF) during the scheduling period is:

[0119]

[0120] In the charging tariffs of electric vehicle users:

[0121] The charging cost C(t) of an electric vehicle within the time period t, abbreviated as C:

[0122]

[0123] Among them, γ cha 、γ dis are the real-time charge and discharge prices respectively. Since the electric vehicle is in the charging or discharging mode at a specific time, it is necessary to maintain:

[0124]

[0125] Then the charging cost of the electric vehicle within the scheduling period T divided by the charging and discharging process of the electric vehicle, that is, the electric vehicle charging cost EVCC can be expressed as:

[0126]

[0127] In the real-time energy consumption difference:

[0128] The power generated by wind power WT and photovoltaic PV cannot be accurately predicted. P diff (t) can be abbreviated as P diff , defined as the difference between the predicted renewable energy power generation and the total load (power) of the electric vehicle charging and discharging within the time period t:

[0129]

[0130] To maximize the utilization of renewable energy generation, the charging and discharging load curve of the electric vehicle should closely match the predicted renewable energy power generation. It can be characterized by minimizing the real-time energy consumption difference (RECD).

[0131]

[0132] In the optimization model, there are many constraints, which are reflected in the capacity of the distribution equipment, the output power of the Micro-Grid (MG), and the remaining charge (i.e., the state of charge SOC).

[0133] In the capacity of the distribution equipment:

[0134] During the time period t, the total load of the distribution network should fall within the capacity range of the distribution equipment:

[0135]

[0136] Among them, and describe the maximum charge-discharge power distribution that the equipment can withstand, that is, is the maximum discharge power of the distribution equipment; is the maximum charging power of the distribution equipment; P DN is the capacity of the distribution equipment.

[0137] In the output power of the MG:

[0138] The output power limits of WT and PV can be described as:

[0139]

[0140] Among them, is the predicted output power of the wind power; is the rated wind power; is the predicted output power of the photovoltaic power; is the photovoltaic wind power.

[0141] The relationship between the charge-discharge power of electric vehicles and the charge-discharge power of charging piles:

[0142] P ev,cha = η cha P ile,cha (21)

[0143]

[0144] P pile,cha is the charging power of the charging pile; P pile,dis is the discharging power of the charging pile.

[0145] The fast charging control corresponding to the fast charging pile often adopts a multi-stage constant current charging method, as shown in Figure 7 shown.

[0146] The power boundary of fast charging within the time period t:

[0147]

[0148] For slow charging, the AC charging rate is generally less than 0.2C, and it can be obtained from:

[0149]

[0150]

[0151] wherein, and are the minimum charge and discharge powers to maintain the charge and discharge cycles, and 0.2kW is adopted in this application; and are the maximum charge and discharge powers of the charging pile, and respectively represent the upper bounds of the charge and discharge powers during multi-stage constant current charging. It is assumed that:

[0152]

[0153] In the SOC:

[0154] During the entire process of the electric vehicle charge and discharge operation, it should be satisfied that:

[0155] 0 ≤ SOC ev ≤ 100% (28)

[0156] SOC ev is the remaining charge of the electric vehicle.

[0157] During adjacent time periods, the change in the SOC of a single electric vehicle within the t time period, SOC(t), can be described as:

[0158]

[0159] wherein, E ev is the battery capacity.

[0160] The overall SOC change can be obtained from:

[0161]

[0162] wherein, SOC ev,start and SOC ev,end are the starting and ending SOCs respectively; t start and t end are the starting time and ending time respectively. The powers during the start and end periods satisfy:

[0163]

[0164] 0 < t start , t end≤Δt (33)

[0165] wherein, the actual powers at the start and end time periods are respectively and P ev,start and P ev,end are the average powers at the start and end respectively.

[0166] SOC ev,accepted = SOC ev,start + 0.8×(SOC ev,expected - SOC ev,start ) (34)

[0167] wherein, SOC ev,expected is the expected final SOC after charging completion; SOC ev,accepted is the SOC acceptable to the electric vehicle user. The NSGA-II algorithm used inevitably results in the difference SOC ev,end , SOC ev,end from SOC ev,expected being: error SOC

[0168] SOC error = ∣SOC ev,end - SOC ev,expected ∣ (35)

[0169] The error should be kept within a certain range, and the range is:

[0170] SOC error ≤ 0.1% (36)

[0171] A charging pile configuration is proposed.

[0172] The charging station is configured with fast charging piles of different maximum powers, and all slow charging piles have the same maximum power. In order to effectively allocate the charging piles to the arriving electric vehicles, a minimum power allocation method MAM and a random power allocation method RAM are proposed for fast charging electric vehicles, and a random power allocation method S-RAM is proposed for slow charging electric vehicles. The following assumptions are made in this application:

[0173] An electric vehicle selected for fast charging will not be allocated to a slow charging pile, and an electric vehicle selected for slow charging will not be allocated to a fast charging pile.

[0174] When all the charging piles in the station are available, there is always a charging pile that can meet the charging demand.

[0175] The actual departure time is the same as the expected departure time.

[0176] The waiting time for an available charging pile is not too long, satisfying:

[0177] t wait = min(0.3 × t park , 60) (37)

[0178] where t wait is the waiting time; and t park is the estimated stay time.

[0179] It is assumed that the electric vehicle does not participate in the charging plan under the following circumstances:

[0180] When the expected terminal SOC of the EV is lower than 80%, it will not participate in the charging plan.

[0181] The expected parking time is less than Δt, and the parking cycle is within the same time period.

[0182] Figure 8 For the MAM and RAM methods proposed in this application, the specific steps are as follows:

[0183] Step 1: After the vehicle enters the charging station, the electric vehicle interacts with the EIDC to obtain information such as the current state of charge, the expected final state of charge, the minimum acceptable state of charge at departure, and the expected departure time.

[0184] Step 2: If there are charging piles, determine all the charging piles that can achieve the expected terminal SOC.

[0185] Step 3: If there are available charging piles, the electric vehicle will participate in the charging scheduling. The charging pile with the lowest output power is assigned to the EV (MAM), and a charging pile that meets the conditions is randomly selected and assigned to the EV (RAM), and charging starts.

[0186] Step 4: If there are no such charging piles, determine all the charging piles that can reach an acceptable terminal SOC.

[0187] Step 5: If there are such charging piles, the electric vehicle selects the maximum power charging throughout the charging process and does not participate in the charging scheduling. The charging pile with the lowest output power is assigned to the electric vehicle, and a charging pile that meets the conditions is randomly selected and assigned to the electric vehicle, and charging starts.

[0188] Step 6: If there are no such piles, identify all the piles where the EV can still reach the target terminal SOC after waiting for a period of time.

[0189] Step 7: If there are such piles, the electric vehicle selects to participate in the charging scheduling. The first available charging pile is assigned to the electric vehicle.

[0190] Step 8: If there are no such piles, find all the piles where the EV can reach an acceptable SOC after waiting.

[0191] Step 9: If there are such charging piles, the electric vehicle does not participate in the orderly charging scheduling. The earliest released charging pile is allocated to the electric vehicle, and charging starts after waiting.

[0192] Step 10: If there are no such charging piles, the electric vehicle can only abandon charging.

[0193] Step 11: If there are no charging piles, continue to execute Steps 6 - 10.

[0194] In the case of not allocating charging piles (NA), if the electric vehicle does not know when the charging pile will be released and there are no available charging piles, the electric vehicle immediately abandons charging. After completing Steps 1 - 5, directly abandon charging. In addition, in Steps 3 and 5, the electric vehicle randomly selects a charging pile for allocation.

[0195] Figure 9 The S - RAM proposed in this application is shown. The specific steps are as follows:

[0196] Step 1: Select slow - charging electric vehicles to enter the charging station. The EVO and EIDC interact to obtain the current SOC of the electric vehicle, the expected SOC at departure, the minimum acceptable SOC at departure, and the expected time to leave the charging station.

[0197] Step 2: If there are available charging piles, the electric vehicle participates in the orderly charging plan. The charging piles are randomly allocated to the electric vehicle, and the electric vehicle starts charging.

[0198] Step 22: If there are no available charging piles, obtain all available charging piles before the departure time of the electric vehicle, and the electric vehicle can be charged to the expected SOC after they become available.

[0199] Step 4: If there are such charging piles, the electric vehicle participates in the orderly charging plan. The earliest available charging pile is allocated to the electric vehicle, and the electric vehicle waits before starting to charge.

[0200] Step 5: If there are no such charging piles, obtain all the piles that can charge the EV to the minimum acceptable SOC after they become available.

[0201] Step 6: If there are such charging piles, the electric vehicle does not participate in the orderly charging plan. The earliest available charging pile is allocated to the electric vehicle, and the electric vehicle waits before starting to charge.

[0202] Step 7: If there are no such charging piles, the electric vehicle must abandon charging.

[0203] When no charging pile allocation (NA) is carried out, since the electric vehicle does not know when there will be available charging piles, if there are no available charging piles, the electric vehicle immediately abandons charging.

[0204] An ordered charging sliding window mechanism is proposed.

[0205] As Soon As Possible (ASAP) charging: After the electric vehicle is connected to the charging pile, a multi-stage constant current fast charging method is immediately adopted to meet the charging demand under the charging power limits of formulas (23) and (24).

[0206] Ordered charging: After the electric vehicle is connected to the charging pile, the EIDC formulates a charging plan for the electric vehicle according to the input information of the electric vehicle's charging demand. Electric vehicles in the charging state will not be affected.

[0207] A general charging scheduling procedure is as follows:

[0208] Step 1: The electric vehicle starts the charging operation and determines the length of the scheduling window.

[0209] Step 2: The electric vehicle decides whether to participate in the charging scheduling.

[0210] Step 3: If the vehicle chooses to participate in the charging plan, the charging plan is exported.

[0211] Step 4: If the vehicle chooses not to participate in the charging plan, ASAP is activated.

[0212] Step 5: Slide the window and repeat steps 1 to 4.

[0213] As Figure 10 shown, Electric Vehicle 1 chooses not to participate in the charging scheduling, and Electric Vehicles 2 and 3 choose to participate in the charging scheduling. Since Electric Vehicle 2 has a longer idle time and the length of the scheduling window is less than the parking time of the EV, this application assumes that for fast charging behavior, the scheduling time satisfies the following conditions:

[0214] t dispatch = min(t park , t ASAP + 120) (38)

[0215] For slow charging behavior, the scheduling duration satisfies:

[0216] t dispatch = min(t park , t ASAP + 240) (39)

[0217] Where, t dispatch is the length of the scheduling window, t park is the parking duration of the electric vehicle, and t ASAP is the time required for the electric vehicle to charge to the expected SOC under the ASAP charging strategy. The unit is minutes, and the calculation method is:

[0218]

[0219] Wherein, I is the number of times of SOC stage change, and the maximum value is 4 times; is the maximum charging power in the i-th stage, and satisfies:

[0220] SOC0 = SOC ev,start (41)

[0221] SOC I+1 = SOC ev,expected (42)

[0222]

[0223] Wherein, SOC0 is the initial SOC of the electric vehicle at the start of orderly charging; SOC I+1 is the SOC of the electric vehicle during orderly charging.

[0224] Setting the scheduling time limit can reduce the number of decision variables, thereby reducing the computing power requirements; reduce the fire hazard caused by the electric vehicle being connected to the charging pile for a long time; prevent the SOC from being insufficient due to the electric vehicle leaving early.

[0225] Combined with the scheduling method, the decision variable array can be described as:

[0226]

[0227] Wherein, represents the charging and discharging power within the scheduling period T divided during the charging and discharging process of the electric vehicle, and its value should satisfy the constraints of formula (23) - formula (26).

[0228] Regarding the real-time charging scheduling method for electric vehicles that realizes renewable energy consumption and V2G in a public charging station in a microgrid, this application uses the NSGA-II algorithm to obtain the Pareto solution set and uses the Entropy-TOPSIS method to determine the optimal solution. Their roles in the optimization derivation process are as Figure 11 and Figure 12 shown.

[0229] Regarding the improved NSGA-II algorithm: Step 1: Initialize the population consisting of ≥0 individuals. Step 2: Evaluate each individual using the fitness function. Step 3: Select the individuals with higher fitness from the current population as parents. Step 4: Perform crossover and mutation to generate offspring. Step 5: Evaluate the offspring using the fitness function. Step 6: Use the non-dominated sorting and crowding distance assignment of NSGA-II to select the next generation population from the combined population. Step 7: Repeat steps 2 to 6 until the maximum number of iterations is reached. Step 8: Obtain the Pareto solution set.

[0230] Regarding the Entropy-TOPSIS method. Step 1: Adopt pfln = ln(pareto_front) performs a logarithmic transformation on the Pareto front value (Pareto solution set) to obtain the transformed information pf ln , to eliminate the influence of the magnitude difference between objectives. Here, pareto_front is the Pareto solution set.

[0231] Step 2: Adopt to normalize the transformed Pareto front value (Pareto front value) to ensure consistent weights across dimensions.

[0232] Step 3: Calculate the entropy entropy using the normalized Pareto front value. The calculation formula is:[[]]

[0233] entropy = -∑(pf In,normalized ·ln(pf In,normolized ))

[0234] Step 4: Calculate the weights weights based on the entropy ln .

[0235]

[0236] Step 5: Reverse the logarithmic transformation of the weights weights = exp(weights ln ), and normalize it to obtain the normalized data weights normolized .

[0237] Step 6: Calculate the normalized matrix matrix of TOPSIS normallzed .

[0238]

[0239] Step 7: Calculate the ideal solution ideal best and the negative ideal solution ideal worst .

[0240]

[0241] Step 8: Calculate the Euclidean distances to the ideal solution and the negative ideal solution.

[0242]

[0243] Step 9: Calculate the closeness degree.

[0244]

[0245] Step 10: Select the best solution, i.e., optimize the charging scheduling solution best 。

[0246] sloution best =pareto f ront[arg max(closeness)]。

[0247] Based on the charging data samples collected by the electric vehicle big data platform and combined with the charging behaviors of electric vehicles, this application proposes a real-time charging scheduling method for electric vehicles to achieve renewable energy consumption and V2G in public charging stations in a microgrid. Using actual data for verification, it is found that the optimal vehicle-to-pile ratio is 7 in the fast charging mode and 1.5 in the slow charging mode. Subsequently, this application proposes a real-time scheduling and orderly charging strategy for electric vehicles based on a sliding window mechanism, which affects the overall optimization goal of the charging station by adjusting the charging and discharging power of individual electric vehicles. The charging pile allocation mechanism improves the utilization rate of charging piles and reduces the waiting time, waiting rate, and abandonment rate of electric vehicles. Numerical results show that compared with disorderly charging, the scheduling strategy proposed in this application reduces the load fluctuation of the distribution network by 3.94%, reduces the average charging cost of electric vehicles by 1.18 CNY, and reduces the average difference in real-time energy consumption by 0.47%.

[0248] In an exemplary embodiment, an electric vehicle charging scheduling device based on a public charging station is provided, including: an information data acquisition module, configured to acquire information data of each electric vehicle in a target public charging station. The information data includes: charging request data, current state of charge data, expected final state of charge data, expected departure time, and minimum acceptable state of charge data at departure; the charging request data includes: selecting a fast charging pile and selecting a slow charging pile; the maximum power of the fast charging pile is greater than a preset power; the maximum power of the slow charging pile is less than the preset power; the target public charging station is connected to a microgrid powered by renewable energy, and the distribution network serves as an auxiliary power source. A distribution module, configured to perform charging pile allocation according to all the information data based on a charging pile allocation mechanism of a sliding window mechanism, and obtain an allocation result. The allocation result is used to schedule the charging power of each electric vehicle to improve the overall operation efficiency of the target public charging station. An optimization processing module, configured to perform optimization processing on the charging scheduling based on the allocation result according to an optimization model by using the NSGA-II algorithm and the Entropy-TOPSIS method, and obtain an optimized charging scheduling scheme. The optimized charging scheduling scheme is used to achieve real-time charging scheduling between electric vehicles for renewable energy consumption and V2G, and realize enabling balance optimization; the optimization model is determined according to a charging behavior database of electric vehicles in the target public charging station, and considering the uncertainty of charging time and charging characteristics; the optimization model includes an objective function and constraint conditions; the charging behavior database includes multiple data samples; each data sample includes: battery capacity, starting charging state, ending charging state, starting charging time, ending charging time, and parking time.

Claims

1. An electric vehicle charging scheduling method based on a public charging station, characterized in that, The electric vehicle charging scheduling method based on a public charging station includes: Obtaining information data of each electric vehicle in the target public charging station; the information data includes: charging request data, current state of charge data, expected final state of charge data, expected departure time, and minimum acceptable state of charge data at departure; the charging request data includes: selecting a fast charging pile and selecting a slow charging pile; the maximum power of the fast charging pile is greater than a preset power; the maximum power of the slow charging pile is less than the preset power; the target public charging station is connected to a microgrid with renewable energy generation, and the distribution network is used as an auxiliary power source; Based on the charging pile allocation mechanism of the sliding window mechanism, allocating charging piles according to all the information data to obtain an allocation result; the allocation result is used to schedule the charging power of each electric vehicle to improve the overall operation efficiency of the target public charging station; Using the NSGA-II algorithm and the Entropy-TOPSIS method, performing charging scheduling optimization processing based on the optimization model according to the allocation result to obtain an optimized charging scheduling plan; the optimized charging scheduling plan is used to achieve real-time charging scheduling between renewable energy consumption and V2G electric vehicles, and to achieve enabling balance optimization; the optimization model is determined according to the charging behavior database of electric vehicles in the target public charging station, and considering the uncertainty of charging time and charging characteristics; the optimization model includes an objective function and constraint conditions; the charging behavior database includes multiple data samples; each data sample includes: battery capacity, starting charging state, ending charging state, starting charging time, ending charging time, and parking time.

2. The electric vehicle charging scheduling method based on a public charging station according to claim 1, wherein, The objective function includes: distribution network load fluctuation, electric vehicle charging cost, and energy consumption difference of renewable energy generation; In the power fluctuation of the distribution network, the expression of the distribution network load fluctuation is: In the charging tariff of electric vehicle users, the expression of the electric vehicle charging cost is: In the real-time energy consumption difference, the expression of the energy consumption difference of renewable energy generation is: Among them, DNLF is the load fluctuation of the distribution network; T is the scheduling period divided by the charging and discharging process of electric vehicles; P Load (t) is the total load of the distribution network in the t time period; is the average load of the distribution network during the charging and discharging process of electric vehicles; τ is the moment in the t time period; p is the moment serial number in the t time period; EVCC is the charging cost of electric vehicles; C(t) is the charging cost in the t time period; RECD is the energy consumption difference; P diff (t) is the difference between the renewable energy power generation and the charging and discharging power of electric vehicles in the t time period.

3. The electric vehicle charging scheduling method based on a public charging station according to claim 1, wherein The constraint conditions include: distribution equipment capacity constraint, microgrid output power constraint, and remaining state of charge constraint; The expression of the distribution equipment capacity constraint is: The expression of the microgrid output power constraint is: The expression of the remaining state of charge constraint is: 0 ≤ State of Charge (SOC) ev ≤ 100%; Among them, P DN is the capacity of the power distribution equipment; is the maximum discharge power of the power distribution equipment; is the maximum charging power of the power distribution equipment; is the predicted output power of wind power; is the rated wind power; is the predicted output power of photovoltaic power; is the photovoltaic wind power; SOC ev is the remaining charge of the electric vehicle.

4. The electric vehicle charging scheduling method based on a public charging station according to claim 1, wherein, The charging pile allocation mechanism includes that the energy information scheduling center adopts a minimum power allocation method and a random power allocation method for electric vehicles that select fast charging piles, and the energy information scheduling center adopts a random power allocation method for electric vehicles that select slow charging piles.

5. The electric vehicle charging scheduling method based on a public charging station according to claim 4, wherein The process of the energy information scheduling center adopting the minimum power allocation method and the random power allocation method for electric vehicles that select fast charging piles to allocate charging piles specifically includes: The energy information scheduling center judges whether there is an idle charging pile according to the usage situation of the charging pile to obtain a first judgment result; If the first judgment result is yes, then judge whether there is a charging pile that can meet the first preset condition to obtain a second judgment result; the first preset condition is to achieve the expected final state of charge data; If the second judgment result is yes, the electric vehicle participates in orderly charging, and the charging pile with the lowest output power or a charging pile randomly selected to meet the preset conditions is allocated to the electric vehicle for charging; If the second judgment result is no, it is judged whether there is a charging pile that can meet the second preset condition, and a third judgment result is obtained; the second preset condition is to reach the minimum acceptable state of charge data at departure; If the third judgment result is yes, a multi-stage constant current fast charging method is adopted to charge at the maximum power under the charging power limit, does not participate in orderly charging, and the charging pile with the lowest output power or a charging pile randomly selected to meet the second preset condition is allocated to the electric vehicle for charging; If the third judgment result is no, all the charging piles in the target public charging station are identified, and it is judged whether there is a charging pile that can meet the first preset condition after the electric vehicle waits, and a fourth judgment result is obtained; If the fourth judgment result is yes, it is selected to participate in orderly charging, and the first idle charging pile after waiting is allocated to the electric vehicle; If the fourth judgment result is no, after the electric vehicle waits, it is judged whether there is a charging pile that can meet the second preset condition, and a fifth judgment result is obtained; If the fifth judgment result is yes, it does not participate in orderly charging, and the first idle charging pile after waiting is allocated to the electric vehicle; If the fifth judgment result is no, the electric vehicle gives up charging; If the first judgment result is yes or no, return "If the third judgment result is no, all the charging piles in the target public charging station are identified, and it is judged whether there is a charging pile that can meet the first preset condition after the electric vehicle waits, and a fourth judgment result is obtained".

6. The electric vehicle charging scheduling method based on a public charging station according to claim 5, characterized in that, The energy information scheduling center adopts a random power distribution method for electric vehicles that select slow charging piles during the process of charging pile allocation, specifically including: The energy information scheduling center judges whether there is an idle charging pile according to the usage situation of the charging pile, and obtains a sixth judgment result; If the sixth judgment result is yes, the electric vehicle participates in orderly charging, and the charging piles are randomly allocated to the electric vehicle for charging; If the sixth judgment result is no, it is judged whether there is a charging pile that meets the third preset condition after the electric vehicle waits, and a seventh judgment result is obtained; the third preset condition is that there is an idle charging pile before the expected departure time and it can reach the expected final state of charge data; If the seventh judgment result is yes, the electric vehicle participates in orderly charging, and the first idle charging pile is allocated to the electric vehicle so that the electric vehicle waits before starting to charge; If the seventh judgment result is no, it is judged whether there is a charging pile that meets the second preset condition after the electric vehicle waits, and an eighth judgment result is obtained; If the eighth judgment result is yes, the electric vehicle does not participate in orderly charging, and the first idle charging pile is allocated to the electric vehicle so that the electric vehicle waits before starting to charge; If the eighth judgment result is no, give up charging.

7. The electric vehicle charging scheduling method based on a public charging station according to claim 1, characterized in that, Using the NSGA-II algorithm and the Entropy-TOPSIS method, based on the optimization model, charging scheduling optimization is performed according to the allocation result to obtain an optimized charging scheduling scheme, specifically including: Based on the optimization model, the NSGA-II algorithm is used to perform crossover and mutation processing according to the allocation result and the objective function to obtain a Pareto solution set; The Entropy-TOPSIS method is used to perform charging scheduling optimization processing according to the Pareto solution set to obtain an optimized charging scheduling scheme.

8. The electric vehicle charging scheduling method based on a public charging station according to claim 7, wherein Based on the optimization model, the NSGA-II algorithm is used to perform crossover and mutation processing according to the allocation result and the objective function to obtain a Pareto solution set, specifically including: Determine the initial parameters; the initial parameters include: the size of the population, the set number of iterations, the crossover probability, and the mutation probability; the population includes multiple individuals; the individuals are each charging pile in the allocation result; the size of the population is the total number of individuals; Determine the fitness function according to the objective function; Use the fitness function to determine the fitness value of each individual, and select and determine the parents in the population based on the fitness value; Based on the parents, perform crossover and mutation according to the crossover probability and the mutation probability, and use the fitness function to perform iterative optimization based on the non-dominated sorting and crowding distance assignment criteria to obtain a Pareto solution set; Among them, the gen-th iteration process is: Use the fitness function to determine the fitness value of each individual in the population in the gen-th iteration, and select and determine the parents in the population in the gen-th iteration based on the fitness value; Perform crossover processing on the parents in the gen-th iteration using the crossover probability in the gen-th iteration to obtain the offspring in the gen-th iteration; Use the mutation probability in the gen-th iteration to correct the offspring in the gen-th iteration to obtain the corrected offspring in the gen-th iteration; According to the fitness function and the corrected offspring in the gen-th iteration, calculate the fitness value in the gen-th iteration; Judge whether the gen-th iteration reaches the set number of iterations; If so, use the corrected offspring in the gen-th iteration as the Pareto solution set; If not, based on the fitness value in the gen-th iteration, update and combine the corrected offspring in the gen-th iteration based on the non-dominated sorting and crowding distance assignment criteria, and use the updated and combined corrected offspring as the population in the (gen + 1)-th iteration.

9. The electric vehicle charging scheduling method based on a public charging station according to claim 7, wherein Using the Entropy-TOPSIS method, charging scheduling optimization is performed according to the Pareto solution set to obtain an optimized charging scheduling scheme, specifically including: Perform logarithmic transformation and normalization transformation on the Pareto solution set to obtain the pre-converted Pareto value; Perform normalization processing on the pre-converted Pareto value and calculate the entropy value; Determine the weight according to the calculated entropy, and perform reverse logarithmic transformation and normalization processing on the weight value to obtain the normalized processing data; Determine the normalized matrix of TOPSIS according to the normalized processed data and the Pareto front value; Based on the normalized matrix of TOPSIS, respectively determine the Euclidean distances corresponding to the ideal solution and the negative ideal solution; Calculate the intimacy based on the Euclidean distance; the calculation formula of the intimacy is: Among them, closeness is the intimacy; distance worst is the Euclidean distance to the negative ideal solution; distance best is the Euclidean distance to the ideal solution; Determine the optimized charging scheduling scheme according to the intimacy; 10. Electric vehicle charging scheduling device based on a public charging station, characterized in that, The electric vehicle charging scheduling device based on the public charging station includes: An information data acquisition module, configured to acquire the information data of each electric vehicle in the target public charging station; the information data includes: charging request data, current state of charge data, expected final state of charge data, expected departure time, and minimum acceptable state of charge data at departure; the charging request data includes: selecting a fast charging pile and selecting a slow charging pile; the maximum power of the fast charging pile is greater than a preset power; the maximum power of the slow charging pile is less than the preset power; the target public charging station is connected to a microgrid powered by renewable energy, and the distribution network is used as an auxiliary power source; A distribution module, configured to perform charging pile allocation according to all the information data based on a charging pile allocation mechanism based on a sliding window mechanism, and obtain an allocation result; the allocation result is used to schedule the charging power of each electric vehicle to improve the overall operation efficiency of the target public charging station; An optimization processing module, configured to adopt the NSGA-II algorithm and the Entropy-TOPSIS method, and perform charging scheduling optimization processing based on the optimization model according to the allocation result to obtain an optimized charging scheduling scheme; the optimized charging scheduling scheme is used to realize real-time charging scheduling between renewable energy consumption and V2G electric vehicles, and realize enabling balance optimization; the optimization model is determined according to the charging behavior database of electric vehicles in the target public charging station, and considering the uncertainty of charging time and charging characteristics; the optimization model includes an objective function and constraint conditions; the charging behavior database includes a plurality of data samples; each data sample includes: battery capacity, starting charging state, ending charging state, starting charging time, ending charging time, and parking time.

Images