Electric vehicle charging station recommendation method based on multi-microgrid load balancing

Abstract

The application discloses a kind of electric vehicle charging station recommendation methods based on multi-microgrid load balancing, comprising the following steps: S1, according to the residual capacity of electric vehicle, the position of electric vehicle and the position of charging station determine the candidate solution set of electric vehicle charging station;S2, in the candidate solution set, the charging station with the lowest average load in the charging period after electric vehicle reaches is selected as the optimal charging station matched by the algorithm for electric vehicle;S3, the method described in S2 is compared with other three scheduling methods.The above technical scheme can optimize the charging behavior of electric vehicle, effectively balance the load distribution in the system to reduce the influence of a large number of electric vehicle charging on the safe operation of the system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric vehicle charging, and particularly refers to a method for recommending electric vehicle charging stations based on load balancing of multiple microgrids. BACKGROUND

[0002] A microgrid is a comprehensive power system composed of distributed small generators. Compared with a traditional power grid, the microgrid has stronger independence and higher stability. Based on the above characteristics, the microgrid is widely welcomed as a solution for the power grid to deal with major power accidents. Due to differences in power demand in industrial areas and residential areas, the load distribution in different areas is not balanced at the same time. The uncertainty of energy supply and load demand in the power grid may exacerbate the unbalanced load distribution.

[0003] The improvement of battery technology and charging facilities has improved the charging speed of electrical vehicles. The construction of charging stations has greatly alleviated the range anxiety of users. These factors have led to a rapid increase in the number of electrical vehicles. However, the charging behavior of electrical vehicles is uncertain. If the charging activities of a large number of electrical vehicles are not coordinated, the power grid may need to inject a large amount of energy into electrical vehicles in a short time, which will increase the burden of the power grid during peak operation and is not conducive to the safe operation of the power grid. Unlike other uncertain loads in the power grid, the impact of a large number of electrical vehicles on the safe operation of the power grid can be alleviated by coordinating the charging locations and charging times of electrical vehicles. Therefore, the guiding strategy for electrical vehicle charging can be roughly divided into two categories. One is to coordinate the charging start time of electrical vehicles to achieve peak shaving. The other is to allocate electrical vehicles to appropriate charging stations to balance the load balance in the entire power grid.

[0004] In the second type of method, the load balancing matching strategy is based on the load distribution of the entire power grid at the time when the electric vehicle sends a charging request to select a suitable charging station for the electric vehicle. This recommendation method does not take into account the load change in the system during the period from when the electric vehicle sends a charging request to when the charging is completed. Therefore, we propose a method for recommending electric vehicle charging stations based on multi-microgrid load balancing, namely the improved load balancing matching strategy. This method fully considers the load distribution of the power grid during the period from when the electric vehicle sends a charging request to when the charging is completed, and guides the electric vehicle to a suitable charging station for charging by considering the electric vehicle as a tool for relieving the load imbalance in the system. Compared with the load balancing matching strategy, the shortest distance matching strategy and the random matching strategy, the method (ILBMS) proposed by us can optimize the charging behavior of electric vehicles, effectively balance the load distribution in the system to reduce the impact of electric vehicle charging on the safe operation of the system. SUMMARY

[0005] The present application proposes an algorithm for recommending electric vehicle charging stations based on multi-microgrid load balancing, namely the improved load balancing matching strategy, according to the deficiencies of the prior art. The candidate charging station solution set of the electric vehicle can be determined by the location of the electric vehicle and the charging station. In the candidate solution set, a suitable charging station is recommended for the electric vehicle according to the load distribution in the multi-microgrid system. Compared with the load balancing matching strategy, the shortest distance matching strategy and the random matching strategy, ILBMS can more effectively relieve the load distribution imbalance in the system, and greatly avoid the coincidence of electric vehicle charging activities and the peak of the power grid, which is conducive to the safe operation of the multi-microgrid system.

[0006] To solve the above technical problems, the technical scheme of the present application is as follows:

[0007] An algorithm for recommending electric vehicle charging stations based on multi-microgrid load balancing, comprising the following steps:

[0008] S1, determining a candidate charging station solution set of the electric vehicle according to an initial electric quantity of the electric vehicle, a position of the electric vehicle and a position of the charging station;

[0009] S1-1, obtaining a spatial distance matrix of the electric vehicle and the charging station according to position information of the electric vehicle at a time when the electric vehicle sends a charging request and spatial positions of the charging stations.

[0010] S1-2, calculating a farthest driving distance R of the i-th electric vehicle according to an initial electric quantity B of the electric vehicle ini，i and a distance from the charging station. i Thus, a charging station candidate solution set Sol of the i-th electric vehicle is obtained. i .

[0011] S2, recommending an optimal charging station for the electric vehicle in the charging station candidate solution set.

[0012] S2-1, assuming that the electric vehicle drives to the charging station at a constant speed v, and the consumption process of the electric quantity on the way is linear. Thus, a driving time on the way of the electric vehicle and a consumed electric quantity on the way of the electric vehicle

[0013] S2-2, obtaining a residual electric quantity of the electric vehicle after arriving at the charging station based on the consumed electric quantity on the way of the electric vehicle and the initial electric quantity at the time when the electric vehicle sends the charging request. Therefore, a time ΔT required for the electric vehicle to be fully charged i，j which can also be obtained according to a battery charging model of the electric vehicle.

[0014] S2-3, in order to select the most suitable charging station in the candidate solution set, selecting a charging station with the lowest average load in a charging period of the electric vehicle as the recommended charging station of the electric vehicle.

[0015] S3, comparing the ILBMS with other three charging strategies.

[0016] S3-1, simulation setting description

[0017] S3-2, through simulation result analysis, the charging recommendation algorithm described in S2 can better optimize the charging behavior of the electric vehicle and relieve the influence of the charging of the electric vehicle on the safe operation of the power grid.

[0018] The application has the following characteristics and beneficial effects:

[0019] By the technical scheme, the load distribution of the power grid during charging of the electric vehicle can be fully considered, the electric vehicle is used as a tool for relieving uneven load distribution in the system and is guided to charge at a suitable charging station. Compared with the LBMS, the SDMS and the RMS, the ILBMS can more effectively relieve uneven load distribution in the system and greatly avoid coincidence of the electric vehicle charging activity and the peak of the power grid, which is beneficial to safe operation of the multi-micro grid system. Moreover, the more the vehicles in the system, the better the performance of the method. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only show some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without any creative labor.

[0021] Figure 1 The method flowchart of the embodiments of the present application.

[0022] Figure 2 The comparison of the valley-peak ratios in different periods when N=16000 in the embodiments of the present application.

[0023] Figure 3 The comparison of the average valley-peak ratios under different vehicle quantities in the embodiments of the present application.

[0024] Figure 4 The comparison of the load standard deviations under different vehicle quantities in the embodiments of the present application.

[0025] Fig. 5 is the load distribution of each MG in the system at 18:00 when N=16000 in the embodiments of the present application. The red solid points in the figure represent the positions of the charging stations. Fig. 5(a) is the initial load distribution of the multi-micro grid system (without charging of the electric vehicle in the multi-micro grid system), and Fig. 5(b) is the load distribution of the multi-micro grid system after using the four methods.

[0026] Figure 6 The box plot of the total load of 100 MGs under the four algorithms when N=16000 in the embodiments of the present application. DETAILED DESCRIPTION

[0027] A method for recommending a charging station for an electric vehicle based on load balancing of a multi-micro grid, as shown in Fig. 1, comprises the following steps: Figure 1

[0028] S1, determining a candidate charging station solution set of the electric vehicle according to a remaining electric quantity of the electric vehicle, a position of the electric vehicle and positions of the charging stations;​

[0029] S1-1, according to the location information of the charging request time of the electric vehicle and the spatial position of the charging station, the distance matrix R of the electric vehicle and the charging station can be obtained, which is specifically represented as follows:

[0030]

[0031] Wherein, R i，j represents the distance between the ith electric vehicle and the jth charging station, N E is the number of electric vehicles in the system, N G is the number of charging stations.

[0032] S1-2, according to the initial electric quantity B ini，i of the electric vehicle, the maximum driving distance R i of the ith electric vehicle is calculated. i The following conditions need to be met,

[0033] R i <R max B ini，i

[0034] Wherein, R max is the maximum driving distance of the electric vehicle in the full electric state. Thus the charging station candidate solution set Sol i of the ith electric vehicle can be obtained, and is represented by the following formula,

[0035] Sol i ={V G |R i ≤R max B ini，i}

[0036] Wherein, V G represents the candidate charging station solution set that meets the conditions. In this candidate charging station solution set, the appropriate charging station is recommended to the electric vehicle user.

[0037] S2, in the candidate charging station solution set, the charging station with the lowest average load during the charging period after the electric vehicle arrives at the charging station is selected as the optimal charging station matched by the algorithm for the electric vehicle;

[0038] S2-1, assuming that the electric vehicle drives to the charging station at a constant speed v, the time consumed by the ith electric vehicle to the jth charging station is

[0039]

[0040] The electric quantity consumed on the road can be defined as

[0041] ​

[0042] So, the start time of EV charging is

[0043]

[0044] where t i is the time when EV sends the charging request.

[0045] S2-2, based on the amount of electricity consumed on the road by the electric vehicle and the initial amount of electricity at the time of sending the charging request, we can get the remaining amount of electricity of the electric vehicle after reaching the charging station is defined as follows,

[0046]

[0047] The charging process of the electric vehicle battery is not linear, and a double exponential function model is used to describe the change of the battery power with time t in the charging process of the electric vehicle.

[0048] SOC(t) = 1.0 + ae -bt -(1 + a)e -ct

[0049] where a, b, c are proportional constants in the battery charging process. Therefore, the electric vehicle needs to be charged from 0% to the required charging time can be solved as follows,

[0050]

[0051] So, the time required for the electric vehicle to be fully charged ΔT i，j is specifically expressed as follows,

[0052]

[0053] where t full is the charging time required for the electric vehicle battery to be charged from 0% to 100%, which is a constant by default.

[0054] S2-3, in order to select the most suitable charging station in the candidate charging station solution set, this embodiment introduces the average load P i，j This parameter to quantitatively describe the load of each MG in the charging period of the i-th electric vehicle after reaching the j-th charging station. P i，j is defined as follows,

[0055]

[0056] where P j (t) is the real-time load of the j-th MG at time t. Pj (t) is defined as follows,

[0057] P j (t) = P 0，j (t) + σ t，j p

[0058] where P 0，j (t) is the initial load of the jthMG at time t, i.e., the base load when no electric vehicle is charging in the MG. σ t，j is the number of electric vehicles that are charging in the jthMG at time t. p is the constant maximum power of electric vehicle charging.

[0059] It can be understood that the smaller P i，j is, the lower the average load level of the jthMG during the charging period of the ith electric vehicle, and the better the load balancing among microgrids can be achieved when the electric vehicle charges at the charging station.

[0060] S3, compare ILBMS with other three charging methods.

[0061] S3-1, introduction of the control method

[0062] The proposed algorithm ILBMS is compared with another ordered charging strategy, load balancing matching strategy (LBMS), and two unordered charging strategies, random matching strategy (RMS) and shortest distance matching strategy (SDMS). LBMS refers to selecting the charging station with the lowest load in the system when the candidate solution set is selected by the electric vehicle when sending a charging request. RMS refers to randomly selecting a charging station within the electric vehicle's reachable range. SDMS refers to selecting the nearest charging station.

[0063] S3-2, the simulation parameters are set as follows,

[0064] The initial load of the MGs in the system is scaled according to the California grid data, and the load data within a day is intercepted as the base load for the numerical simulation. The sampling interval time in the data is 5 minutes, so the total number of time samples T d = 288, and the sampling time window is 0:00-23:55. The charging stations are evenly distributed in a range of 50x50km. Electric vehicles randomly appear in this range. Other parameter settings are shown in Table 1,

[0065] Parameter Value Parameter Value p 50 kW v 30 km / h [R max ]] 300 km [CAT full ]] 240 min a 2.096 b 0.0669 c 0.0469

[0002] N

[0004] G

[00221] ​ 100

[0066] Table 1 Parameter settings

[0067] S3-3, Through the analysis of simulation results, the charging recommendation algorithm described in S2 can better optimize the charging behavior of electric vehicles and alleviate the impact of electric vehicle charging on the safe operation of power grids.

[0068] By recommending electric vehicles to suitable charging stations to balance the power load among MGs, the impact of a large number of electric vehicle charging on the safe operation of the system can be alleviated. A valley-peak ratio η(t) is introduced to measure the load balancing degree in the system at time t, which is defined as follows,

[0069]

[0070] Since there is a valley-peak gap in the load distribution in the system at the same time, we use P valley (t) and P peak (t) to represent the minimum and maximum values of the load in the system at time t, respectively. The larger the valley-peak ratio, the smaller the gap between the valley and peak values in the system at time t, and the more stable the system operation.

[0071] In this embodiment, η in different time periods in the sampling window is compared. As shown in Figure 2 , it can be seen that when N = 16000, the unordered charging strategy RMS and SDMS have no obvious effect on improving the η of the system. This shows that ILBMS can effectively balance the load in the power grid and enhance the safety and stability of the power grid operation at time.

[0072] Further, an average valley-peak ratio is introduced to measure the load balancing degree in the sampling window of the multi-microgrid system. The definition of is as follows,

[0073]

[0074] From Figure 3 , it can be seen that under RMS and SDMS hardly changes with the increase of N. This shows that these two scheduling strategies have no significant effect on improving the valley-peak ratio. The under the LBMS algorithm shows a trend of first increasing and then decreasing, which shows that the performance of the LBMS algorithm is not very stable. On the contrary, the under the ILBMS algorithm shows a monotonically increasing trend. This shows that the valley-peak gap in the system is gradually decreasing. And from Figure 3 , it can be directly seen that the of ILBMS is the largest. This shows that ILBMS can better reduce the valley-peak gap in the system.

[0075] The size of the load standard deviation within the system can fully reflect the load difference. The smaller the load standard deviation, the smaller the load difference between MGS at a certain moment, the more balanced the load distribution within the system, and the more stable the operation of the multi-microgrid. Specifically, we introduce the average load standard deviation to reflect the degree of load difference within the sampling window,

[0076]

[0077] where D(t) represents the difference in the total load size of each MG at time t.

[0078] In this embodiment, as shown in Figure 4 , the load fluctuation of MGs under four algorithms is compared. From Figure 4 , it can be seen that the load standard deviation of the unordered scheduling strategies RMS and SDMS has a large numerical difference compared with the two ordered charging strategies. This shows that RMS and SDMS cannot fully utilize electric vehicles to achieve the purpose of reducing load fluctuation. On the contrary, the ordered scheduling strategies LBMS and ILBMS can effectively schedule electric vehicles to smooth the load fluctuation within the multi-microgrid. In addition, as the number of electric vehicles N in the system increases, the load standard deviation under the two unordered strategies remains almost unchanged, while the load standard deviation under LBMS and ILBMS decreases more obviously. This shows that the ordered charging strategy can better smooth the load fluctuation within the system. Although ILBMS and LBMS can both reduce load fluctuation, there is still a significant difference in performance between the two methods. From Figure 4 , it can be found that compared with LBMS, the load standard deviation under the ILBMS method is smaller. And as N increases, the difference between the two gradually becomes larger. This shows that ILBMS can more effectively utilize the charging behavior of electric vehicles to smooth the load fluctuation within the system. In addition, as N increases, the load standard deviation under the ILBMS method is a downward trend, while the trend under the LBMS method is first downward and then slowly larger. Therefore, ILBMS can more effectively reduce the load fluctuation within the system whether N is large or small, and the method performance is more stable. In summary, ILBMS has a significantly better method performance than the other three methods in smoothing the load fluctuation within the system.

[0079] The above compares the system load valley peak difference under four different strategies, but this comparison only focuses on the size of the two MG loads at the same time at the system peak and valley, and cannot represent the load distribution of all MGs in the system. For the integrity and rigor of the simulation, the load distribution of all MGs in the system at 18:00 when N=16000 is further compared in FIG. 5. As can be seen from FIG. 5(a), when there is no electric vehicle charging in the system, the load distribution of each MG at the same time fluctuates greatly. The simulation results of FIG. 5(b) show that ILBMS not only does not increase the peak load of the system, but also can effectively fill the system valley. The remaining three scheduling methods to some extent make the charging behavior of electric vehicles coincide with the peak of MG power consumption, which undoubtedly increases the risk of safe operation of the power grid. Therefore, ILBMS can effectively utilize the charging activities of electric vehicles to balance the load distribution in the system, and further proves the advantages of ILBMS method under the above standard deviation and valley peak ratio indicators.

[0080] When scheduling electric vehicles to appropriate charging stations, not only the load distribution at a certain time in the system needs to be concerned, but also the total load P sl of each MG in the entire time period needs to be valued. sl,j The total load P sl,j of the jth MG is defined as follows,

[0081]

[0082] The smaller the total load fluctuation in the system, the more smoothly the system runs. From Figure 6 it can be seen that under the ILBMS method, the system has the smallest load peak, the largest valley, and the smallest interquartile range, which shows that ILBMS can better guide the charging behavior of electric vehicles, thereby reducing the load fluctuation between MGs and being beneficial to the stable operation of the power grid.

[0083] Through this specific embodiment, it can be proved that the method of recommending electric vehicle charging stations based on multi-microgrid load balancing proposed in the present application. Compared with random matching, shortest distance matching and load balancing matching, the ILBMS method proposed by us can optimize the charging behavior of electric vehicles, effectively balance the load distribution in the system to reduce the impact of a large number of electric vehicle charging on the safe operation of the system.

[0084] The embodiments of the present application are described in detail above in combination with the drawings, but the present application is not limited to the described embodiments. For those skilled in the art, various changes, modifications, replacements and variations of the embodiments including components are made without departing from the principles and spirits of the present application, and still fall within the protection scope of the present application.

Claims

1. A method for recommending an electric vehicle charging station based on load balancing of multiple microgrids, characterized in that, The method comprises the following steps: S1, determining a candidate charging station solution set of the electric vehicle according to the spatial positions of the electric vehicle and the charging stations and the initial electric quantity of the electric vehicle; S2, recommending an optimal charging station for the electric vehicle in the candidate charging station solution set of the electric vehicle; S2-1, determining the charging start time of the electric vehicle after the electric vehicle arrives at the charging station according to the time of the electric vehicle on the road; S2-2, establishing a double exponential function model of the electric vehicle battery charging, and then determining the charging period of the electric vehicle; The method for determining the charging period of the electric vehicle according to the double exponential function model is as follows: Setting the electric car from 0% to Required charging time This can be solved as follows, Time ΔT required for an electric vehicle to be fully charged i,j is expressed as follows, where t full is the time required to charge the battery of an electric vehicle from 0% to 100%, by default a constant; S2-3, selecting the charging station with the lowest average load in the charging period of the candidate charging station solution set as the recommended charging station of the electric vehicle; The method for recommending the charging station is as follows: By introducing the average load P i,j This parameter quantitatively describes the load of each MG during the charging period of the i-th electric vehicle after it arrives at the j-th charging station, where P i,j is defined as follows: where P j (t) is the real-time load of the jth MG at time t, P j (t) is defined as follows, P j (t) = P 0,j (t) + σ t,j p where P 0,j (t) is the initial load of the jthMG at time t, i.e., the base load when no electric vehicle is charging in the MG; σ t,j is the number of electric vehicles that are being charged at the jthcharging station at time t; p is the constant maximum power of electric vehicle charging; S3, performance verification. 2.The method of claim 1, wherein, The step S1 comprises the following sub-steps: S1-1, obtaining the spatial distance matrix of the electric vehicle and the charging stations according to the position information of the time when the electric vehicle issues a charging request and the spatial positions of the charging stations; S1-2, according to the initial electric quantity B of the electric vehicle ini,i and the distance from the charging station, the farthest driving distance R of the i-th electric vehicle is calculated i , and the charging station candidate solution set Sol of the i-th electric vehicle is obtained i。 3.The method of claim 2, wherein, The calculation method in the step S2-1 is as follows: during the time when the electric vehicle travels to the charging station, the electric vehicle is set to travel to the charging station at a constant speed v, and the power consumption on the road is also linear consumption, which is specifically characterized as follows: the i-th electric vehicle goes to the j-th charging station on the road and the time consumption is tij= d ij v, (1) where d ij is the distance between the i-th electric vehicle and the j-th charging station. power consumed on the road defined as where R max is the driving range of an electric vehicle at full charge, R i,j represents the distance between the ith electric vehicle and the jth charging station. 4.The method of claim 1, wherein, The battery charging process satisfies a double exponential function model, and the expression is as follows: SOC(t) = 1.0 + ae -bt - (1 + a)e -ct Wherein a, b and c are proportional constants in the battery charging process.

5. The method of claim 1-4, wherein, In the step S3, the method for performance verification is to compare the algorithm performance of the load balancing matching strategy, the shortest distance matching strategy and the random matching strategy in balancing the system load distribution through simulation. 6.The method of recommending an electric vehicle charging station based on multi-micro-grid load balancing according to claim 5, wherein, The performance comparison method is as follows: the system valley-peak ratio in the sampling window, the system average valley-peak ratio, the difference degree of the load distribution in the system at the same time, the load distribution in the system at a certain time and the difference degree of the total load of the micro-grid in the sampling window, and the valley-peak ratio η(t) is defined as follows, Since there is a difference between the minimum and maximum values of the load in the system at the same time, the minimum and maximum values of the load in the system at time t are respectively denoted as P valley (t) and P peak (t) represent that the greater the valley-peak ratio, the smaller the difference between the minimum and maximum values of the load in the system at time t, and the more stable the system runs.

Images