Electric vehicle coordinated charging scheduling method based on micro-grid

Abstract

The electric vehicle coordinated charging scheduling method based on the micro-grid comprises the following steps: based on collected charging information, simulating randomness of electric vehicles in combination with a Monte Carlo simulation (MCS) method, and obtaining a charging demand of each electric vehicle; establishing a coordinated charging scheduling optimization model by taking minimization of the overall peak-valley load difference as a target function, and considering various limitations of slow charging of the electric vehicle, quick charging of the electric vehicle and operation of the micro-grid; solving the coordinated charging scheduling optimization model, generating a scheduling strategy and transmitting the scheduling strategy to an electric vehicle aggregator; and the electric vehicle aggregator charges different electricity demands according to the scheduling instruction. The result shows that the charging mode selection of the electric vehicle owner and the peak load shifting of the micro-grid operation are realized, and the peak value, the range and the variance of the load are reduced, so that the safety and the reliability of the micro-grid are improved.

Description

Technical Field

[0001] The present invention belongs to the field of power system microgrid optimization scheduling, and in particular relates to a microgrid-based electric vehicle coordinated charging scheduling method. Background Art

[0002] As an emerging means of transportation, electric vehicles (EVs) have developed rapidly with the advantages of zero emissions and high industrial added value. When the number of EVs charged exceeds a certain scale, it will bring certain risks to the planning and operation of the power system. To this end, it is necessary to reasonably optimize the charging behavior of large-scale EVs. The optimal scheduling method to solve the EV charging scheduling problem can also adopt the decentralized optimization strategy of the Lagrangian relaxation method, which has high computational efficiency and is basically unaffected by the scale of EV scheduling; or introduce a training learning mechanism to reduce the peak-valley difference load, which has high computational efficiency; through the spatiotemporal dual-layer optimization scheduling strategy of EV charging load, it can also meet the interests of the power grid, charging station operators and EV users.

[0003] However, through the above analysis, it is found that existing research mainly focuses on the power supply side to achieve the lowest operating cost or the smallest peak-valley load difference. Secondly, EV charging plans usually do not take into account the specific charging needs of EV owners, but only treat the charging load as a simple whole. Affected by my country's policies and social habits, many electric vehicle owners have similar travel patterns, which may lead to a large number of electric vehicles being charged in the same time period. In extreme cases, the user's centralized charging time happens to be during the peak load period, which increases the burden on the power grid and threatens the safe and stable operation of the power grid.

[0004] In addition, the dispatching and control difficulty of the power grid will also increase. Since electric vehicle users have subjective initiative, their charging behavior is uncertain, which leads to randomness in users' charging, affecting the prediction of electric vehicle charging load in the region and the operation and dispatching of the power grid. Furthermore, large-scale centralized charging of electric vehicles may also affect the quality of electric energy. This centralized charging behavior may not only cause system voltage deviation, but also increase energy loss and cause negative consequences such as harmonic distortion, accelerating the wear and aging of power equipment.

[0005] Therefore, it is of great significance to study how to guide users' charging behavior to charge electric vehicles in an orderly manner and ensure the safe and stable operation of the power system. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a coordinated charging scheduling method for electric vehicles based on a microgrid. The method takes into account the slow and fast charging of electric vehicles and various limitations of microgrid operation, and selects the charging mode of the electric vehicle through the charging urgency index, thereby reflecting different charging needs and further reasonably meeting the electricity needs of each charging car owner.

[0007] The technical solution adopted by the present invention is:

[0008] The coordinated charging scheduling method of electric vehicles based on microgrid includes the following steps:

[0009] Step 1: Based on the collected charging information, the Monte Carlo simulation (MCS) method is used to simulate the randomness of electric vehicles and obtain the charging demand of each electric vehicle;

[0010] Step 2: Taking minimizing the overall peak-valley load difference as the objective function, a coordinated charging scheduling optimization model is established, and various restrictions on electric vehicle slow charging, electric vehicle fast charging, and microgrid operation are considered;

[0011] Step 3: Solve the coordinated charging scheduling optimization model, generate a scheduling strategy and transmit it to the electric vehicle aggregator;

[0012] Step 4: EV aggregators will charge the vehicles according to the dispatch instructions according to different electricity demands.

[0013] In step 1, during the electric vehicle (EV) charging scheduling process, the electric vehicle aggregator (EVA) collects charging information and performs EV charging scheduling. When an electric vehicle is connected to a microgrid, the electric vehicle owner sets the charging information and sends it to the electric vehicle aggregator. The charging information includes the arrival time, the departure time, and the state of charge (SOC) when the EV is connected to the microgrid, where the lower SOC is the EV charging demand and the higher SOC is the safety of the EV battery.

[0014] The control center can implement a coordinated charging scheduling method based on the collected charging information;

[0015] First, set the length and number of time slots;

[0016] Second, the arrival and departure times of electric vehicles are normalized into time gaps.

[0017] Then, the charging urgency index (CUI) of electric vehicles is defined to distinguish whether the charging demand of electric vehicles is urgent, and two charging modes are provided for car owners;

[0018] Finally, the control center generates a dispatch strategy and transmits it to the EV aggregator, who then sends the dispatch instructions to the EV owners. Figure 2 The framework diagram of the proposed EV charging scheduling model is given.

[0019] In the step 1,

[0020] 1) Charge and discharge time:

[0021] Considering the charging habits of EV owners, the conventional mode of EV charging can be divided into two charging modes: private charging mode and shared charging mode;

[0022] In the private charging mode, the EV owner starts charging after getting home from get off work and ends charging when preparing to go to work. The arrival time f(t 1c ), departure time f(t 1dis ) conforms to the normal distribution function and can be expressed as follows:

[0023]

[0024] Where: t 1c and t 1dis Respectively represent the time when the owner arrives home and the time when he leaves home; and are the expectation and variance of the time when EV is connected to the grid, respectively; and are the expectation and variance of the time when EV leaves the grid, respectively. and

[0025] In the shared charging mode, the EV owner starts charging when he arrives at the workplace in the morning and completes charging after get off work. The arrival time f(t 2c ), departure time f(t 2dis ) conforms to the normal distribution function and can be expressed as follows:

[0026]

[0027] Where: t 2c and t 2dis They represent the time when the owner arrives at the workplace and the time when he leaves the workplace respectively; and are the expectation and variance of the time when EV is connected to the grid, respectively; and are the expectation and variance of the time when EV leaves the grid, respectively. and

[0028] According to the above definition, the probability distribution of arrival and departure time of different numbers of EVs under two charging modes can be obtained, such as Figure 3 and Figure 4 As shown. Figure 3 and Figure 4 In the figure, the red curve is the probability distribution function of EV arrival and departure time, and the histogram represents the data generated according to the probability distribution function in the Monte Carlo method to simulate the randomness of EV charging time behavior.

[0029] from Figure 3 It can be seen that in private charging mode, EV owners get home in about 18 hours and go to work in about 8 hours. Figure 4 The data show that in the shared charging mode, EV owners usually connect the EV to the microgrid when they arrive at the workplace in about 8.5 hours, and disconnect the EV from the microgrid after getting off work in about 17.5 hours. Therefore, the generated data conforms to people's life rhythm and work characteristics under the two charging modes.

[0030] 2) Time slot division:

[0031] In the scheduling process, in order to improve execution efficiency, the scheduling plan is usually executed by time period; in the proposed EV charging scheduling model, a day is discretized into 96 time slots, each time slot is 15 minutes long. Therefore, the arrival time and departure time of each EV can be expressed as:

[0032]

[0033] Where: N represents the number of EVs; and represents the time slot number when the i-th EV is connected to and disconnected from the microgrid; and They represent the arrival time and departure time of the i-th EV respectively; △T represents the length of the time slot; Represents the next integer greater than the result of the division operation; Represents the previous integer that is smaller than the result of a division operation.

[0034] Formula (5) indicates that when the EV is connected to the microgrid at the middle time point of a time slot, the charging behavior can be scheduled after the next integer time slot. At the same time, when the EV is disconnected from the microgrid at the middle time point of a time slot; Formula (6) ensures that the EV can only be scheduled before the end of the previous integer time slot.

[0035] 3) Charging Emergency Indicator (CUI):

[0036] In the proposed EV charging scheduling model, all EVs are scheduled according to their charging demand, which is defined as reflecting the EV charging urgency index (CUI).

[0037] Based on arrival time slot and leave time slot The entire time slot that the EV is connected to the microgrid can be calculated by formula (7):

[0038]

[0039] Where: T i Rem It represents the number of remaining time periods during which the i-th EV can continue to access the microgrid. The charging behavior and charging scheduling strategy should be arranged within these remaining time periods. The schematic diagram of the charging control period of the i-th EV is shown in Figure 5 As shown:

[0040] In the present invention, CUI is expressed as:

[0041]

[0042] Where: is the slow charging power of EV, η EV Indicates the charging efficiency of EV, It represents the lower limit of the state of charge when the i-th EV ends its charging behavior. represents the state of charge of the i-th EV when it is connected to the microgrid, Indicates EV battery capacity. i <0, it means that the i-th EV needs emergency charging; when CUI i ≥0, indicating that the charging demand of the i-th EV is not urgent.

[0043] 4) Charging demand mode selection:

[0044] In the proposed charging scheduling method, two charging demand modes of EV are considered, namely, fast charging mode and slow charging mode. Although fast charging shortens the charging time, the current and voltage of fast charging are relatively high, which affects the battery life. Therefore, the fast charging model is usually used in certain emergency situations. Here, the choice of charging mode is determined by the charging urgency index CUI, which can be expressed as:

[0045]

[0046] Where: represents the fast charging power, and Indicates slow charging power; CUI i Indicates charging urgency indicator.

[0047] When CUI i <0, the i-th EV adopts fast charging mode. i ≥0, the i-th EV adopts slow charging mode.

[0048] The step 2 comprises:

[0049] 2.1: Establish a coordinated charging scheduling optimization model:

[0050] After parameter preparation and charging mode selection, the optimal coordinated charging scheduling optimization model is established. The model consists of three parts: model variables, objective functions and constraints for fast-charging electric vehicles and slow-charging electric vehicles; in addition, constraints for microgrids are added to the model to ensure the operation of microgrids;

[0051] 2.1.1: Charging state variables:

[0052] In the coordinated charging scheduling optimization model, the state variable is x i,j is a binary variable. If x i,j = 0, indicating that the i-th EV does not charge in the j-th time slot; if x i,j =1, indicating that the i-th EV is charged at the j-th time slot, as shown below:

[0053]

[0054] 2..1..2: Objective function:

[0055] The optimization goal of the present invention is to transfer part of the load demand from the peak period to the valley period, thereby alleviating the peak load pressure of the microgrid. The total load of the microgrid includes the base load and the EV charging load, and the EV charging load includes the load of fast charging EV and slow charging EV. Therefore, the total load of the jth time slot of the coordinated charging scheduling optimization model can be expressed as:

[0056]

[0057] Where: represents the basic load of the jth time slot, represents the power provided to all EVs in the jth time slot. If the i-th EV has an urgent charging demand P EV,i equal If the i-th EV has no urgent charging demand, then P EV,i equal T represents a time period.

[0058] The goal of charging scheduling is to reduce the peak load difference of the microgrid, which can be expressed as:

[0059]

[0060] Where: and Represent the maximum load demand and the minimum load demand respectively.

[0061] 2.1.3: Restrictions on fast charging EVs:

[0062] When the i-th EV is regarded as an emergency charging EV, it is assumed that at each arrival time slot x i,j are all equal to 1. Therefore, the EV can be charged from the arrival time to the departure time, ensuring that the emergency EV can obtain more power at the charging time.

[0063] In addition, when the EV is disconnected from the microgrid, x i,j =0; however, when the EV charging urgency is low, continuous charging will result in a charging power Higher overcharging phenomenon. To avoid overcharging of fast-charging EVs, a charging demand limit constraint should be set, that is, the charging demand must be lower than the maximum SOC demand.

[0064] Considering the above two constraints, when EV is connected to the microgrid, x i,j is equal to 1, and maintains the charging state until the EV stops charging. The time slot when the EV stops charging is determined by the departure time and the time slot that meets the maximum SOC requirement. The constraints for emergency charging EV can be expressed as:

[0065]

[0066] Where: The time slot for stopping charging for fast charging EV, It can be expressed as:

[0067]

[0068] Where: is the maximum SOC demand; represents the EV charging time slot that needs to meet the maximum SOC demand; Indicates the previous integer that is smaller than the result of the division operation, thus avoiding overcharging of an EV in the next integer.

[0069] and represents the time slot number when the i-th EV is connected to and disconnected from the microgrid; represents the state of charge of the i-th EV when it is connected to the microgrid; Indicates EV battery capacity; represents the fast charging power of EV; η EV represents the charging efficiency of the EV, and △T represents the length of the time slot.

[0070] 2.1.4: Limitations of slow charging EVs:

[0071] Due to the diversity of car owners' charging requirements, each EV has a minimum SOC requirement and a maximum SOC requirement to avoid overcharging. For slow-charging EVs, when the EV is disconnected from the microgrid, its minimum SOC requirement must be met, which can be expressed as:

[0072]

[0073] Where: is the SOC when the i-th EV is disconnected from the microgrid, which can be calculated by the following formula:

[0074]

[0075] Where: x i,j Indicates the total number of slow-charging EVs, is the SOC of the i-th EV when it is connected to the microgrid; Indicates the slow charging power of EV;

[0076] The second constraint of the slow-charging EV is related to the scheduling time; charging behavior can only be scheduled during the time period when the EV is connected to the microgrid; therefore, when the slow-charging EV is not connected to the microgrid, the charging state x i,j It is equal to 0. It should be noted that when the EV is disconnected from the microgrid, the dispatch plan cannot be executed.

[0077] The step 2 comprises:

[0078] 2.2: Constraints:

[0079] 2.2.1: Constraints of Microgrids:

[0080] The objective function is to reduce the peak-to-valley difference of the total load, but it cannot guarantee that the peak coordination result is lower than the peak of the non-coordinated charging method that meets the maximum SOC requirement of the EV. In order to avoid the emergence of new charging peak loads in the microgrid in the coordination plan, the constraint shown in formula (17) is imposed on the microgrid to limit the increase of the peak.

[0081]

[0082] Where: It is the maximum value of the total load that satisfies the maximum SOC demand of EV in 96 time slots under the uncoordinated charging mode;

[0083] 2.2.2: Total load of microgrid:

[0084] The charging scheduling results can be obtained by solving the coordinated charging scheduling optimization model. In addition, the effect of the coordinated charging scheduling method can be verified by comparing the coordinated charging method with the non-coordinated charging method.

[0085] Figure 6 and Figure 7 The calculation process of the total load under uncoordinated and coordinated charging methods is shown in equation (15). In equation (15), the SOC of the normally charged EV when disconnected from the microgrid should be between the minimum and maximum SOC requirements in the coordinated charging method. In the uncoordinated charging method, two experiments will be conducted to meet the minimum and maximum SOC requirements of the EV. The total load in each time slot under the three charging conditions can be obtained.

[0086] In step 3, the control center solves the problem through the CPLEX solver, generates a scheduling strategy and transmits it to the electric vehicle aggregator;

[0087] The coordinated charging scheduling optimization model is established in MATLAB / YALMIP and solved by the CPLEX solver. CPLEX is a software package that provides high-performance solvers for mathematical programming problems, including linear programming, mixed integer programming, and quadratic programming. CPLEX has the characteristics of strong robustness, high optimization efficiency, and fast calculation. CPLEX has been widely used in academia, industry, and research institutions.

[0088] In step 4, the aggregator will charge the car owners with different electricity demands according to the dispatch instructions. By coordinating and dispatching the charging of electric vehicles, the results reveal the charging mode selection of electric vehicle owners and the peak-shaving and valley-filling of microgrid operation, reducing the peak, extreme and variance of the load, thereby helping to improve the safety and reliability of the microgrid.

[0089] The present invention provides a coordinated charging scheduling method for electric vehicles based on a microgrid, and the beneficial effects are as follows:

[0090] 1) The present invention proposes a coordinated charging scheduling method for EVs by considering the urgent charging needs of EV owners. During the charging scheduling process, all EVs are scheduled according to different groups rather than as a whole. Two charging modes are considered to illustrate that the method has the effect of peak shaving and valley filling.

[0091] 2) The method of the present invention can adapt to EV charging modes with different charging requirements, and can also have the effect of peak shaving and valley filling, which can improve the operation of the microgrid and ensure the stable operation of the power system.

[0092] 3) While considering the different charging needs of EV owners, the present invention proposes an EV coordinated charging scheduling method that takes both the power supply side and the demand side into consideration. First, the urgency of EV charging needs is fully considered. Second, all EVs are arranged according to different charging needs rather than as a whole. Finally, the uncertainty of EV charging behavior is considered in the simulation, and two different EV charging modes are used. The results show that the proposed method can reduce the peak, extreme and variance of the load by coordinating the scheduling of EV charging, which helps to improve the safety and reliability of the microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0094] Figure 1 This is a flow chart of the scheduling method of the present invention.

[0095] Figure 2 The framework diagram of the proposed EV charging scheduling model.

[0096] Figure 3 Simulation results of EV charging behavior in private mode.

[0097] Figure 4 The simulation results of EV charging behavior in sharing mode.

[0098] Figure 5 Schematic diagram of charging control period.

[0099] Figure 6 This is the flow chart for calculating the total load in the uncoordinated charging method.

[0100] Figure 7 This is the flow chart for calculating the total load in the coordinated charging method.

[0101] Figure 8 This is a typical base load curve for a day cycle.

[0102] Figure 9(a) shows the results of uncoordinated charging and the proposed coordinated charging in private charging mode (200EVs);

[0103] Figure 9(b) shows the result 2 of uncoordinated charging and the proposed coordinated charging in private charging mode (500EVs);

[0104] Figure 9(c) shows the result of uncoordinated charging and the proposed coordinated charging in private charging mode (800EVs).

[0105] Figure 10(a) shows the results of uncoordinated charging and the proposed coordinated charging in the shared charging mode (200EVs);

[0106] Figure 10(b) shows the result 2 of uncoordinated charging and the proposed coordinated charging in the shared charging mode (500EVs);

[0107] Figure 10(c) shows the result of uncoordinated charging and the proposed coordinated charging in the shared charging mode (800EVs). DETAILED DESCRIPTION

[0108] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0109] The coordinated charging scheduling method for electric vehicles based on microgrids can realize the transfer of load demand from peak period to valley period. It is characterized by comprising the following steps:

[0110] Step 1: The control center obtains the charging demand of each electric vehicle based on the collected charging information and the Monte Carlo simulation (MCS) to simulate the randomness of electric vehicles;

[0111] Step 2: Taking minimizing the overall peak-valley load difference as the objective function, a coordinated charging scheduling optimization model is established, and various restrictions on slow-charging EVs, fast-charging EVs, and microgrid operations are considered;

[0112] Step 3: The control center solves the problem through the CPLEX solver, generates a dispatch strategy and transmits it to the electric vehicle aggregator;

[0113] Step 4: The aggregator will charge the car owners with different electricity needs according to the dispatch instructions.

[0114] In order to verify the coordinated charging scheduling method of electric vehicles based on microgrids, realize the transfer of load demand from peak period to trough period, and enable those skilled in the art to better understand the present invention and the superiority and feasibility of the present invention, the input data is randomly generated according to the probability density function to simulate the actual EV charging situation. However, for the sake of simplicity, some assumptions must be established. First, a comparative experiment was carried out using different numbers of EVs (i.e., 200, 500, and 800). At the same time, the experiment assumes that there are different numbers of EVs charging in private charging and shared charging modes. In the simulation, statistical data on the charging and discharging time of different numbers of EVs in the two charging modes were obtained, such as Figure 3 , Figure 4 shown.

[0115] Secondly, assuming It obeys the continuous uniform distribution between 0.1 and 0.3. Evenly distributed between 0.4 and 0.6, Evenly distributed between 0.8 and 1.0, and are set to 3.5kW and 10kW respectively, ηEV Set to 0.9. In addition, all EV battery capacities Set to 32kWh.

[0116] Finally, the basic load data was simulated according to the power consumption pattern. The simulation results are shown in Figure 8 As shown, the typical basic load curve of the regional microgrid in the daily cycle scenario is displayed, with peak times occurring at approximately 6h, 10h, and 18h, with a maximum load of 710.62kW and a minimum load of 445.69kW.

[0117] 1. Scheduling results and comparative analysis:

[0118] (1.1): Private charging mode:

[0119] In the private charging mode, EV owners are accustomed to charging at night and driving the next day. In the experiment, a day cycle consists of two half days. In order to reflect the entire scheduling process of EV charging, 12 hours in one day to 12 hours in the next day are selected as the scheduling time period, and 15 minutes is defined as a time slot, dividing the day cycle into 96 equal parts.

[0120] The simulation results of EV optimal charging scheduling under three different charging methods in private charging mode are as follows: Figure 9(a) to Figure 9(c) As shown in the figure. In each figure, three different charging scheduling strategies are adopted, and 96 time slots are given in the form of histograms in each figure. For these uncoordinated charging scheduling plans, EV charging starts from the arrival time slot, and the EV is always in the charging state until the departure time slot or the time slot when the owner's maximum SOC demand is met. Since the model considers both the minimum SOC and the maximum SOC requirements, the two SOCs in the uncoordinated charging scheduling method should be considered in the uncoordinated charging scheduling method. Four indicators in the private charging mode are compared in Table 1, namely, peak difference, valley difference, extreme difference and variance. Load variance is used to characterize the fluctuation of microgrid load. The smaller the variance, the smoother the load change. In Table 1, A and B represent the EV disorderly charging scheduling methods that meet the maximum and minimum SOC requirements, respectively. C is a coordinated EV charging scheduling method.

[0121] Table 1 Comparison of private charging modes

[0122]

[0123] from Figure 9(a) to Figure 9(c)It can be seen that the coordinated charging scheduling method can reduce the peak load compared with the two uncoordinated charging scheduling methods. The charging curve of the coordinated method is smoother than the two uncoordinated charging curves, which is conducive to the smooth operation of the microgrid. For EVs with urgent charging needs, the charging time of some EVs is too short, and the uncoordinated charging strategy cannot guarantee that the EV meets the minimum SOC requirement under such urgent circumstances. However, when the coordinated charging scheduling method is adopted, emergency charging situations can be coped with. For slow-charging EVs, the conditions ensure that the SOC when the EV is disconnected from the microgrid should be between the upper and lower SOC requirements. Therefore, the overall charging power of EVs from the microgrid in the coordinated case is lower than the uncoordinated charging scheduling that meets the maximum SOC requirement, but higher than the uncoordinated charging scheduling that meets the minimum SOC requirement.

[0124] From Table 1, it can be obtained that compared with the non-coordinated charging that meets the maximum SOC demand, the coordinated charging method can reduce the peak load share by 38.57%, 54.71% and 63.32% in the cases of 200, 500 and 800 EVs, respectively. In addition, compared with the uncoordinated charging that meets the minimum SOC requirement, the coordinated charging scheduling method can reduce the proportion of peak load by 29.78%, 36.74% and 44.88%, respectively; compared with the uncoordinated charging method that meets the maximum SOC requirement, the coordinated charging scheduling method can reduce the proportion of extreme load by 85.44%, 93.43% and 88.32%, and compared with the range of uncoordinated charging that meets the minimum SOC requirement, the coordinated charging scheduling method can reduce it by 84.72%, 93.34% and 81.48%, respectively; compared with the uncoordinated charging method that meets the maximum SOC requirement, the coordinated charging scheduling method can reduce the proportion of load variance by 97.31%, 99.79% and 99.58%, and compared with the range of uncoordinated charging that meets the minimum SOC requirement, the coordinated charging scheduling method can reduce it by 95.45%, 99.79% and 98.65%, respectively.

[0125] In summary, compared with the two non-coordinated charging scheduling methods, the coordinated charging scheduling method can reduce the peak, extreme and variance of the load, which makes the total load distribution of the microgrid smoother and makes the operation of the microgrid more stable.

[0126] (1.1): Shared charging mode:

[0127] In the shared charging mode, the cycle is defined as 0 to 24 hours a day. The simulation results are shown in Figure 10(a) to Figure 10(c) As shown, the four statistical indicators of the public charging mode are shown in Table 2;

[0128] Table 2 Comparison of shared charging modes

[0129]

[0130] from Figure 10(a) to Figure 10(c) It can be seen that for all involved EVs, when the uncoordinated charging method is adopted, the charging demand surges at about 10 hours. The charging curve of the coordinated method is flatter than the two uncoordinated charging curves. Compared with the two uncoordinated charging methods, the coordinated charging scheduling method can reduce the peak of the total load, making the total load distribution of the microgrid more balanced, similar to the private charging mode.

[0131] From Table 2, it can be seen that compared with the uncoordinated charging method that meets the maximum SOC demand, the coordinated charging scheduling method can reduce the peak load by 32.43%, 46.38% and 54.01% in the case of 200, 500 and 800 EVs, respectively. Compared with the uncoordinated charging method that meets the minimum SOC demand, the coordinated charging scheduling method can reduce the peak load by 17.77%, 26.83% and 33.02%, respectively; compared with the uncoordinated charging method that meets the maximum SOC demand, the coordinated charging scheduling method can reduce the extreme load by 53.85%, 6.64% and 1.30%, respectively. 2.07% and 65.96%. Compared with the uncoordinated charging method that meets the minimum SOC requirement, the coordinated charging scheduling method can reduce the extreme load by 34.90%, 41.43% and 45.16%; compared with the uncoordinated charging method that meets the maximum SOC requirement, the coordinated charging scheduling method can reduce the variance of the total load by 71.46%, 84.60% and 88.04% respectively. Compared with the uncoordinated charging method that meets the minimum SOC requirement, the coordinated charging scheduling method can reduce the variance of the load by 35.39%, 52.00% and 59.61% respectively.

[0132] In summary, in the shared charging mode, the coordinated charging scheduling method can reduce the peak, range, and variance of the total load, making the total load curve of the microgrid smoother. In addition, when more EVs are involved, the function of reducing the peak-valley load difference is more obvious in the shared charging mode.

[0133] 2. Impact of charging mode:

[0134] It can be seen from Table 1 and Table 2 above. In the private charging mode, the uncoordinated EV charging scheduling method will increase the variance of the microgrid's base load, and the coordinated charging scheduling method can reduce the variance of the base load of three different EV numbers by 85%, 98% and 68% respectively. In the public charging mode, all three charging methods can increase the variance of the microgrid's base load. In addition, Figure 9(a) to Figure 9(c) The coordinated charging curves of the three charging scales are better than Figure 10(a) to Figure 10(c) The coordinated charging curve is gentle.

[0135] The following is a discussion and analysis of the causes of the above phenomenon:

[0136] First, the average residence time of EV in home charging mode is longer than that in public charging mode. In home charging mode, EV is connected to the microgrid at about 18 hours and disconnected from the microgrid at about 8 hours. However, in public charging mode, EV tends to be connected to the microgrid at about 8.5 hours and disconnected at 17.5 hours. In addition, in private charging mode, the peak of the base load occurs at 18h, 5h and 10h, and the valley occurs at 23-24h. In this model, most EVs are connected to the microgrid at about 18h and disconnected at about 8h, so the minimum value of the base load can be filled during most of the EV charging time. In addition, the peak load pressure can be reduced, and the peak shaving and valley filling effect is obvious.

Claims

1. A coordinated charging scheduling method for electric vehicles based on microgrids, characterized in that The following steps are involved: Step 1: Based on the collected charging information, the Monte Carlo simulation (MCS) method is used to simulate the randomness of electric vehicles and obtain the charging demand of each electric vehicle; Step 2: Taking minimizing the overall peak-valley load difference as the objective function, a coordinated charging scheduling optimization model is established, and various restrictions on electric vehicle slow charging, electric vehicle fast charging, and microgrid operation are considered; Step 3: Solve the coordinated charging scheduling optimization model, generate a scheduling strategy and transmit it to the electric vehicle aggregator; Step 4: EV aggregators will charge the vehicles according to the dispatch instructions according to different electricity demands.

2. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 1 is characterized in that: In step 1, during the charging scheduling process of electric vehicles (EVs), the electric vehicle aggregator (EV When an electric vehicle is connected to a microgrid, the owner of the electric vehicle sets the charging information and sends it to the electric vehicle aggregator. The charging information includes the arrival time, the departure time, and the state of charge (SOC) when the EV is connected to the microgrid, where the lower SOC is the EV charging demand and the higher SOC is the safety of the EV battery.

3. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 2 is characterized in that: In the step 1, 1) Charge and discharge time: Considering the charging habits of EV owners, the conventional mode of EV charging is divided into two charging modes: private charging mode and shared charging mode; In the private charging mode, the EV owner starts charging after getting home from get off work and ends charging when preparing to go to work. The arrival time f(t 1c ), departure time f(t 1dis ) conforms to the normal distribution function, which is expressed as follows: Where: t 1c and t 1dis Respectively represent the time when the owner arrives home and the time when he leaves home; and are the expectation and variance of the time when EV is connected to the grid, respectively; and are the expectation and variance of the time when EV leaves the grid, respectively; In the shared charging mode, the EV owner starts charging when he arrives at the workplace in the morning and completes charging after get off work. The arrival time of the EV f(t 2c ), departure time f(t 2dis ) conforms to the normal distribution function, which is expressed as follows: Where: t 2c and t 2dis They represent the time when the owner arrives at the workplace and the time when he leaves the workplace respectively; and are the expectation and variance of the time when EV is connected to the grid, respectively; and are the expectation and variance of the time when EV leaves the grid, respectively; According to the above definition, the probability distribution of arrival and departure time of different numbers of EVs under two charging modes can be obtained; the data generated by the probability distribution function in the Monte Carlo method is used to simulate the randomness of EV charging time behavior; 2) Time slot division: In the scheduling process, in order to improve the execution efficiency, the scheduling plan is usually executed by time period; a day is discretized into 96 time slots, each time slot is 15 minutes long; therefore, the arrival time and departure time of each EV are expressed as: Where: N represents the number of EVs; and represents the time slot number when the i-th EV is connected to and disconnected from the microgrid; and They represent the arrival time and departure time of the i-th EV respectively; △T represents the length of the time slot; Represents the next integer greater than the result of the division operation; Represents the previous integer that is smaller than the result of the division operation; Formula (5) indicates that when the EV is connected to the microgrid at the middle time point of a time slot, the charging behavior can be scheduled after the next integer time slot; at the same time, when the EV is disconnected from the microgrid at the middle time point of a time slot; Formula (6) ensures that the EV can only be scheduled before the end of the previous integer time slot; 3) Charging Emergency Indicator (CUI): EVs are dispatched according to their charging needs, which is defined as reflecting the EV charging urgency index (CUI); Based on arrival time slot and leave time slot The entire time slot that the EV is connected to the microgrid can be calculated by formula (7): Where: T i Rem It represents the number of remaining time periods during which the i-th EV can continue to access the microgrid, and the charging behavior and charging scheduling strategy should be arranged within these remaining time periods; CUI is expressed as: Where: is the slow charging power of EV, η EV Indicates the charging efficiency of EV, represents the lower limit of the state of charge when the i-th EV ends its charging behavior; represents the state of charge of the i-th EV when it is connected to the microgrid, Indicates EV battery capacity; when CUI i <0, it means that the i-th EV needs emergency charging; when CUI i ≥0, indicating that the charging demand of the i-th EV is not urgent; 4) Charging demand mode selection: Two charging demand modes of EV are considered, namely fast charging mode and slow charging mode; the choice of charging mode is determined by the charging urgency index CUI, which is expressed as: Where: represents the fast charging power, and Indicates slow charging power; CUI i Indicates charging urgency indicator; When CUI i <0, the i-th EV adopts fast charging mode. i ≥0, the i-th EV adopts slow charging mode.

4. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 1 is characterized in that: The step 2 comprises: 2.1: Establish a coordinated charging scheduling optimization model: After parameter preparation and charging mode selection, the optimal coordinated charging scheduling optimization model is established; the model consists of three parts: model variables, objective functions and constraints for fast-charging electric vehicles and slow-charging electric vehicles; in addition, constraints for microgrids are added to the model to ensure the operation of microgrids; 2.1.1: Charging state variables: In the coordinated charging scheduling optimization model, the state variable is x i,j is a binary variable; if x i,j = 0, indicating that the i-th EV does not charge in the j-th time slot; if x i,j =1, indicating that the i-th EV is charged at the j-th time slot, as shown below: 2.1.2: Objective function: The optimization goal is to transfer part of the load demand from peak hours to valley hours, thereby alleviating the peak load pressure of the microgrid; the total load of the microgrid includes the basic load and EV charging load, and the EV charging load includes the load of fast-charging EV and slow-charging EV; therefore, the total load of the jth time slot of the coordinated charging scheduling optimization model is expressed as: Where: represents the basic load of the jth time slot, represents the power provided to all EVs in the jth time slot; if the i-th EV has an urgent charging demand P EV,i equal If the i-th EV has no urgent charging demand, then P EV,i equal T represents the time period; The goal of charging scheduling is to reduce the peak load difference of the microgrid, which can be expressed as: Where: and Represent the maximum load demand and the minimum load demand respectively; 2.1.3: Restrictions on fast charging EVs: When the i-th EV is regarded as an emergency charging EV, it is assumed that at each arrival time slot x i,j are all equal to 1; therefore, the EV can be charged from the time of arrival to the time of departure, ensuring that the emergency EV can obtain more power at the time of charging; In addition, when the EV is disconnected from the microgrid, x i,j =0; however, when the EV charging urgency is low, continuous charging will result in a charging power Higher overcharging phenomenon; To avoid overcharging of fast-charging EVs, a charging demand limit constraint should be set, that is, the charging demand must be lower than the maximum SOC demand; Considering the above two constraints, when EV is connected to the microgrid, x i,j =1, and maintain the charging state until the EV stops charging; the time slot when the EV stops charging is determined by the departure time and the time slot that meets the maximum SOC requirement; the constraint of emergency charging EV is expressed as: Where: The time slot for stopping charging for fast charging EV, It is expressed as: Where: is the maximum SOC demand; represents the EV charging time slot that needs to meet the maximum SOC demand; Indicates the previous integer that is smaller than the result of the division operation, thus avoiding overcharging of an EV in the next integer; and represents the time slot number when the i-th EV is connected to and disconnected from the microgrid; represents the state of charge of the i-th EV when it is connected to the microgrid; Indicates EV battery capacity; represents the fast charging power of EV; η EV represents the charging efficiency of the EV, △T represents the length of the time slot; 2.1.4: Limitations of slow charging EVs: Due to the diversity of car owners' charging requirements, each EV has a minimum SOC requirement and a maximum SOC requirement to avoid overcharging; for slow-charging EVs, when the EV is disconnected from the microgrid, its minimum SOC requirement must be met, which can be expressed as: Where: is the SOC when the i-th EV is disconnected from the microgrid, which can be calculated by the following formula: Where: x i,j Indicates the total number of slow-charging EVs, is the SOC of the i-th EV when it is connected to the microgrid; Indicates the slow charging power of EV; The second constraint of the slow-charging EV is related to the scheduling time; charging behavior can only be scheduled during the time period when the EV is connected to the microgrid; therefore, when the slow-charging EV is not connected to the microgrid, the charging state x i, j is equal to 0; when the EV is disconnected from the microgrid, the dispatch plan cannot be executed.

5. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 4 is characterized in that: The step 2 comprises: 2.2: Constraints: 2.2.1: Constraints of Microgrids: In order to avoid the emergence of new charging peak loads in the microgrid during the coordination plan, the constraint shown in formula (17) is imposed on the microgrid to limit the increase of the peak; Where: It is the maximum value of the total load that satisfies the maximum SOC demand of EV in 96 time slots under the uncoordinated charging mode; 2.2.2: Total load of microgrid: The charging scheduling results can be obtained by solving the coordinated charging scheduling optimization model. In formula (15), the SOC of the normally charged EV when it is disconnected from the microgrid should be between the minimum and maximum SOC requirements in the coordinated charging method. In the uncoordinated charging method, two experiments will be conducted to meet the minimum and maximum SOC requirements of the EV. The total load in each time slot under the three charging conditions can be obtained.

6. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 1 is characterized in that: In step 3, the control center solves the problem through the CPLEX solver, generates a scheduling strategy and transmits it to the electric vehicle aggregator; the coordinated charging scheduling optimization model is established in MATLAB / YALMIP and solved through the CPLEX solver.

7. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 6 is characterized in that: In step 4, the aggregator will charge the car owners with different electricity demands according to the dispatch instructions; it reveals the charging mode selection of electric car owners and the peak shaving and valley filling of microgrid operation, reducing the peak value, extreme difference and variance of the load.

8. The coordinated charging scheduling method for electric vehicles based on microgrid according to claim 7 is characterized in that: The control center can implement a coordinated charging scheduling method based on the collected charging information; First, set the length and number of time slots; Second, the arrival and departure times of electric vehicles are normalized into time gaps; Then, the charging urgency index (CUI) of electric vehicles is defined to distinguish whether the charging demand of electric vehicles is urgent, and two charging modes are provided for car owners; Finally, the control center generates a dispatch strategy and transmits it to the EV aggregator, who then sends the dispatch instructions to the EV owners.

Images