Charging scheduling method and device for multi-gun charging pile

Abstract

The invention relates to a charging scheduling method and device for multi-gun charging piles, and belongs to the technical field of intelligent charging of new energy electric vehicles. The method comprises the following steps: S1, a charging pile intelligent sensor carries out data acquisition; s2, processing the acquired data; s3, principal component analysis: analyzing the processed data, and converting the acquired data into a small amount of linearly independent data by adopting a principal component analysis method and orthogonal transformation; and S4, charging scheduling and execution: the charging pile mainboard executes a charging task according to a scheduling algorithm in combination with a prediction result. According to the method, specific requirements and challenges in different charging scenes are comprehensively considered, the use efficiency and the resource utilization rate of the charging piles are maximized by optimizing the power distribution and scheduling strategy, and therefore the overall performance and the service capacity of charging facilities are improved, and the ever-increasing new energy vehicle charging requirements are met.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent charging of new energy electric vehicles, and relates to a charging scheduling method and device for a multi-gun charging pile. Background Art

[0002] As the advantages of pure electric-driven new energy vehicles in terms of environmental protection, low noise, driving experience and operating costs become increasingly prominent, more and more car owners tend to choose new energy electric vehicles as their means of travel. This trend is not only reflected in the individual consumer market, but has also been widely used in public transportation, logistics and distribution. Charging, as one of the core needs of new energy electric vehicles, its efficiency and experience are directly related to the owner's satisfaction with the use and the popularity of new energy vehicles. In order to improve the charging efficiency of charging piles, the industry has been constantly exploring new technologies and solutions. On the one hand, increasing the charging power is the most direct and effective method, but this also puts higher requirements on the hardware design of the charging pile and the battery protection mechanism. Although high-power charging can greatly shorten the charging time, too fast charging speed may damage the battery and affect the battery life. Therefore, during the charging process, it is necessary to accurately control the charging power to ensure that the charging efficiency is maximized while protecting the battery.

[0003] However, the current number of charging stations still cannot meet the charging needs of a large number of electric vehicles. Especially in some urban central areas or densely populated areas, the contradiction between supply and demand of charging piles is particularly prominent. At the same time, the high-power charging time of new energy vehicles is limited. In order to protect the battery life and ensure charging safety, electric vehicles usually enter a floating charge state after reaching a certain charging percentage, thereby reducing the charging power. Although this charging strategy is conducive to protecting the battery, it also further aggravates the tension of charging piles. In addition, differences in user charging habits also bring challenges to the management of charging piles. Sometimes after charging is completed, the user fails to unplug the gun or leave the parking space in time, resulting in the charging pile being occupied for a long time, affecting the charging needs of other car owners. In order to solve this problem, some charging pile suppliers have begun to develop charging piles with automatic power-off and reminder functions, but this also requires a higher level of intelligence to support.

[0004] Therefore, more and more charging pile suppliers have begun to develop "one pile with multiple guns" charging piles, gradually transforming from the traditional one pile with one gun to one pile with two guns, four guns and other designs. This design can not only improve the utilization rate of charging piles, but also alleviate the problem of tight supply and demand of charging piles to a certain extent. However, there are still many shortcomings in the current charging pile design. First, the lack of intelligence is a prominent problem. Most of the existing designs are limited to simple multi-charging module switch control, lacking the application of intelligent scheduling and optimization algorithms. This results in the inability of charging piles to flexibly adjust charging power and resource allocation according to actual conditions, reducing charging efficiency and user experience. Secondly, poor practicality is also a drawback of the current charging pile design. In terms of hardware design, many systems rely solely on the on-off operation of relays to achieve the switching of charging guns. This method cannot achieve frequent and accurate switching, which is prone to equipment damage and failure, shortening the service life of the equipment, and increasing maintenance costs and economic burdens.

[0005] Therefore, in order to meet the growing demand for charging and improve the charging experience of car owners, it is necessary to further optimize the intelligent control level of charging piles and improve the reliability and economy of their hardware. By introducing advanced intelligent scheduling algorithms and optimizing hardware design, we can achieve efficient, stable and reliable operation of charging piles, providing strong support for the popularization and development of new energy vehicles. Summary of the invention

[0006] In view of this, the purpose of the present invention is to provide a charging scheduling method and device for multi-gun charging piles, which focuses on intelligent scheduling and control of the power output of a single charging module. The method comprehensively considers the specific needs and challenges in different charging scenarios, and aims to maximize the efficiency and resource utilization of charging piles by optimizing power allocation and scheduling strategies, thereby improving the overall performance and service capabilities of charging facilities and meeting the growing charging needs of new energy vehicles.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A charging scheduling method for a multi-gun charging pile, the method specifically comprising the following steps:

[0009] S1: Charging pile intelligent sensor collects data;

[0010] S2: Process the collected data;

[0011] S3: Principal component analysis: Analyze the processed data using the principal component analysis method, and use orthogonal transformation to convert the above collected data into a small number of linearly independent data;

[0012] S4: Charging scheduling and execution: The charging pile mainboard executes the charging task according to the scheduling algorithm combined with the prediction results.

[0013] Further, in step S1, it specifically includes: the charging pile mainboard system collects m characteristic data x=(x 1 ,x 2 ,x 3 ,…,x m ) T Perform n collections for preliminary data preparation for intelligent charging scheduling. The data is represented by x 1 ,x 2 ,x 3 ,…,x n , where x j =(x 1j ,x 2j ,x 3j ,…,x mj ) T represents the jth feature data, x ij represents the i-th variable in the data collected for the j-th time, j = 1, 2, ..., n; the data collected by the charging pile is represented by the data matrix X, which is recorded as:

[0014]

[0015] Further, in step S2, it specifically includes:

[0016] After obtaining the data matrix X, we can obtain the data mean and data covariance; the data mean vector Expressed as

[0017]

[0018] The data covariance matrix S is

[0019] S=[s ij ] m×n

[0020]

[0021] in is the sample mean of the i-th feature data, is the sample mean of the jth feature data; the data correlation matrix R is:

[0022]

[0023] in

[0024]

[0025] Further, in step S3, the charging pile mainboard cleans the data, and uses the principal component analysis method to convert the above collected data into a small number of linearly independent data using orthogonal transformation. This method is used to determine the relationship between the eigenvalues ​​in the data collected by the charging pile, and is used to screen out the most important characteristic data; the singular value decomposition method through the data matrix is ​​used, specifically including:

[0026] 1) The data collected by the charging pile is normalized to obtain a normalized new data matrix, which is still represented by X. Specifically, the original feature data is transformed as follows:

[0027]

[0028] in

[0029]

[0030] In this method, the normalized variables Still recorded as x ij , the normalized data matrix is ​​still represented by X;

[0031] 2) According to the normalized data matrix X, calculate the data feature correlation matrix R (the calculation method refers to the above formula)

[0032]

[0033] 3) Calculate k eigenvalues ​​and corresponding k unit eigenvectors of the feature data correlation matrix R;

[0034] Solve the characteristic equation of R |R-λI|=0, where I is the identity matrix, and get the m eigenvalues ​​of R

[0035] λ 1 ≥λ 2 ≥…≥λ m

[0036] Calculate the variance contribution rate The number of principal components k that reaches the predetermined value; the i-th principal component y i The variance contribution rate of y is defined as i The ratio of the variance of to the sum of all variances, denoted as η i

[0037]

[0038] Find the unit eigenvectors corresponding to the first k eigenvalues, a i =(a 1i ,a 2i ,…,a mi )T ,i=1,2,…,k;

[0039] 4) Find the principal components of k data, perform linear transformation with k unit eigenvectors as coefficients, and find the principal components of k data:

[0040]

[0041] 5) Calculate k principal components y j With the original variable x i The correlation coefficient ρ(x i ,y j ), and k principal components for the original variable x i Contribution rate v i ;

[0042]

[0043] where α ii is the variable x i The variance of the covariance matrix Σ is the diagonal element of the covariance matrix; its mean vector is μ

[0044] μ=E(x)=(μ 1 ,μ 2 ,…,μ) T

[0045] Covariance matrix Σ = E[(x-μ)(x-μ) T ]

[0046]

[0047] According to the contribution rate v i Sort from large to small and select the first n data that meet the requirements;

[0048] 6) Calculate the k principal component values ​​of n data, and substitute the normalized feature data into the formula in 4) to obtain the principal component values ​​of n feature data; the jth data x j =(x 1j ,x 2j ,x 3j ,…,x mj ) T The i-th principal component value of

[0049]

[0050] 7) Input the k principal component values ​​of n data into the model M to obtain the prediction result p = M(y'), that is, p = M(y 11 ,y 12 ,…,y ij); Model M is not specifically defined here. Model M can download a general pre-trained model PM from the Internet for fine-tuning according to actual needs, or it can be trained by itself after cleaning the charging pile data obtained in the production environment; the data used to fine-tune (train) the model will be labeled in advance, and the label is the priority; the sorted data is input into the model PM for fine-tuning (training) to obtain a dedicated model M, so that model M can fit the label to give a priority p that meets the current specific usage scenario.

[0051] Further, in step S4, the charging pile mainboard performs the charging task according to the scheduling algorithm combined with the prediction result p. If the number of currently connected vehicles c t When it is 1, the charging time is unlimited and the waiting time is 0; if the number of vehicles is greater than 2, the charging scheduling algorithm starts to be executed, including:

[0052] 1) First, set the waiting time T w and charging time T c All are initialized to T minutes. The charging pile determines which of the peak, flat, and low peak periods (set time periods according to local electricity prices) the charging pile is in based on the current time. According to different peak periods, the waiting time and charging time are set to T / 2 minutes, T minutes, and T*2 minutes respectively.

[0053] 2) If a new vehicle is connected, charging will start immediately (waiting time T w =0), the charging time is T n minutes, after charging, the priority p is reduced by 1 level, and other vehicles wait for T w Response extension T n minute;

[0054] 3) For the vehicle being charged, calculate the output power Wo. If the output power Wo drops to the rated power W c When the charging time of the vehicle is reduced to 1 / 2 of c / 2 minutes, and lower the priority p, otherwise charge normally;

[0055] 4) When the vehicle finishes charging or is fully charged, it exits the queue;

[0056] The above steps are performed continuously until all vehicles are charged and the queue is empty.

[0057] The present invention also provides a charging scheduling device for a multi-gun charging pile.

[0058] The beneficial effects of the present invention are:

[0059] The charging scheduling method based on the priority algorithm of the present invention can significantly improve the resource utilization and charging efficiency of the charging pile. Through the multi-dimensional data acquisition of intelligent sensors, the system can monitor and analyze the key parameters of the charging process in real time, thereby providing comprehensive data support for charging scheduling. Compared with the traditional charging pile design, the present invention can filter out the most relevant feature data by introducing the principal component analysis (PCA) technology, effectively reduce redundant information, and improve data processing efficiency.

[0060] In addition, the priority scheduling algorithm of the present invention can intelligently schedule the charging power according to the real-time status of the charging pile and the needs of the connected vehicles, optimize the power distribution, and ensure that the charging process of multiple charging guns can proceed smoothly. Through this flexible control design, the charging pile can dynamically adjust the charging strategy according to the different electricity prices and load conditions in the peak, flat and low peak periods, and realize intelligent and efficient charging management.

[0061] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0063] Figure 1 It is a schematic diagram of the process of the present invention;

[0064] Figure 2 It is a circuit design diagram of the present invention. DETAILED DESCRIPTION

[0065] The technical solution of the present invention is described in detail below with reference to the accompanying drawings.

[0066] Figure 1 It is a schematic diagram of the process of the present invention, such as Figure 1 As shown, the charging scheduling method of the charging pile with multiple charging guns includes the following steps:

[0067] Before the algorithm is executed, the charging pile mainboard system collects multiple characteristic data related to the charging process through intelligent sensors. Including but not limited to the current power of the electric vehicle (percentage), charging gun connection time, charging gun temperature, battery temperature, electric vehicle BMS required voltage and current, electric pile output power, phase power consumption, phase fundamental power consumption, phase power factor, etc., a total of m characteristic data x = (x 1 ,x 2 ,x3 ,…,x m ) T The system collects data n times, and the collected data is used for the preliminary preparation of charging scheduling. The data is represented by x 1 ,x 2 ,x 3 ,…,x n . Where x j =(x 1j ,x 2j ,x 3j ,…,x mj ) T represents the jth feature data, x ij represents the i-th variable in the data collected for the jth time, j = 1, 2, ..., n. The data collected by the charging pile is represented by the data matrix X, which is recorded as:

[0068]

[0069] After obtaining the data matrix X, we can obtain the data mean and data covariance. Data mean vector Expressed as

[0070]

[0071] The data covariance matrix S is

[0072] S=[s ij ] m×n

[0073]

[0074] in is the sample mean of the i-th feature data, is the sample mean of the jth feature data. The data correlation matrix R is:

[0075]

[0076] in

[0077]

[0078] Secondly, the charging pile mainboard cleans the data, and uses the principal component analysis method to convert the collected data into a small number of linearly independent data using orthogonal transformation. This method is used to determine the relationship between the eigenvalues ​​in the data collected by the charging pile, and to filter out the most important feature data. In the present invention, the singular value decomposition method of the data matrix is ​​used. The specific steps are as follows:

[0079] (1) The data collected by the charging pile is normalized to obtain a normalized new data matrix, which is still represented by X. Specifically, the original feature data is transformed as follows:

[0080]

[0081] in

[0082]

[0083] In the present invention, the normalized variables Still recorded as x ij , the normalized data matrix is ​​still represented by X.

[0084] (2) Based on the normalized data matrix X, calculate the data feature correlation matrix R (the calculation method refers to the above formula)

[0085]

[0086] (3) Calculate the k eigenvalues ​​and corresponding k unit eigenvectors of the feature data correlation matrix R.

[0087] Solve the characteristic equation of R |R-λI|=0, where I is the identity matrix, and get the m eigenvalues ​​of R

[0088] λ 1 ≥λ 2 ≥…≥λ m

[0089] Calculate the variance contribution rate The number of principal components k that reaches the predetermined value. The i-th principal component y i The variance contribution rate of y is defined as i The ratio of the variance of to the sum of all variances, denoted as η i

[0090]

[0091] Find the unit eigenvectors corresponding to the first k eigenvalues

[0092] a i =(a 1i ,a 2i ,…,a mi ) T ,i=1,2,…,k

[0093] (4) Find the principal components of k data

[0094] Use k unit eigenvectors as coefficients to perform linear transformation and find k principal components of the data

[0095]

[0096] (5) Calculate k principal components y j With the original variable x i The correlation coefficient ρ(x i ,y j ), and k principal components for the original variable x i Contribution rate v i .

[0097]

[0098] where α ii is the variable x i The variance of is the diagonal element of the covariance matrix Σ. Its mean vector is μ

[0099] μ=E(x)=(μ 1 ,μ 2 ,…,μ m ) T

[0100] Covariance matrix Σ = E[(x-μ)(x-μ) T ]

[0101]

[0102] According to the contribution rate v i Sort from large to small and select the first n data that meet the requirements.

[0103] (6) Calculate the k principal component values ​​of n data

[0104] Substituting the normalized feature data into the formula in (4), we can obtain the principal component values ​​of the n feature data. j =(x 1j ,x 2j ,x 3j ,…,x mj ) T The i-th principal component value of

[0105]

[0106] (7) Input the k principal component values ​​of n data into the model M and obtain the prediction result p = M(y'), that is, p = M(y 11 ,y 12 ,…,y ij). Model M is not specifically defined here. Model M can download the general pre-trained model PM from the Internet for fine-tuning according to actual needs, or it can be trained by itself after cleaning the charging pile data obtained in the production environment. The data used for fine-tuning (training) the model will be labeled in advance, and the label is the priority. The sorted data is input into the model PM for fine-tuning (training) to obtain a dedicated model M, so that the model M can fit the label to give a priority p that meets the current specific usage scenario.

[0107] Finally, the charging pile mainboard executes the charging task according to the scheduling algorithm combined with the prediction result p. If the current number of connected vehicles ct is 1, the charging time is unlimited and the waiting time is 0. If the number of vehicles is greater than 2, the charging scheduling algorithm is started.

[0108] (1) In the present invention, the waiting time Tw and the charging time Tc are both initialized to T minutes. The charging pile determines which of the peak period, flat period, and low peak period (the time period is set according to the local electricity price) the charging pile is in according to the current time, and sets the waiting time and charging time to T / 2 minutes, T minutes, and T*2 minutes respectively according to different peak periods.

[0109] (2) If a new vehicle is connected, charging starts immediately (waiting time Tw = 0), and the charging time is T n minutes. After charging, the priority p is reduced by 1 level, and the waiting time Tw for other vehicles to respond is extended by Tn minutes.

[0110] (3) For the vehicle being charged, the output power Wo is calculated. If the output power Wo drops to 1 / 2 of the rated power Wc, the charging time of the vehicle is reduced to Tc / 2 minutes and the priority p is lowered. Otherwise, charging is performed normally.

[0111] (4) When the vehicle finishes charging or is fully charged, it exits the queue.

[0112] The above steps are repeated until all vehicles are fully charged and the queue is empty.

[0113] By executing the above charging scheduling algorithm, the system can complete the intelligent scheduling of the output power of one pile with multiple guns. Because the execution of this intelligent algorithm requires a high frequency operation of the on-off switch, the present invention designs a hardware device for multi-gun jump based on IGBT (insulated gate bipolar transistor). The circuit design diagram of a charging pile with four guns is shown in the figure below. Figure 2 As shown, the specific design logic is as follows:

[0114] (1) First, the switch control signal from the control mainboard is isolated by photoelectric isolation, so that the host system and the high-voltage control part are isolated. Photoelectric isolation devices can achieve an isolation effect of more than 3000V, achieving almost absolute isolation of high and low voltage systems.

[0115] (2) The power supply for the high-voltage IGBT comes from the isolated DC / DC power conversion (U6, U8, U10, U24), which generates a maximum IGBT drive voltage of 15V.

[0116] (3) The high-voltage positive electrode outputs current to the charging gun through the opening and closing of the IGBT. The positive electrode control adopts a common drain method, and the three IGBTs (Q10, Q11, Q12) use a common power supply drive (MOS_A+). According to the Thevenin theorem, even if the three IGBTs share the drain, the independent current of each IGBT will not be affected.

[0117] (4) The three IGBTs (Q7, Q8, Q9) of the high voltage negative electrode are connected together with their sources and driven by three independent controls (MOS7, MOS8, MOS9).

[0118] (5) By adopting the direct drive method after end element conversion, the influence of IGBT parasitic capacitance can be ignored. At the same time, R21, R22, and R33 in this circuit can also release parasitic capacitance to achieve the effect of fast shutdown.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A charging scheduling method for multi-gun charging piles, characterized by: The method specifically comprises the following steps: S1: Charging pile intelligent sensor collects data; S2: Process the collected data; S3: Principal component analysis: Analyze the processed data using the principal component analysis method, and use orthogonal transformation to convert the above collected data into a small number of linearly independent data; S4: Charging scheduling and execution: The charging pile mainboard executes the charging task according to the scheduling algorithm combined with the prediction results.

2. A charging scheduling method for a multi-gun charging pile according to claim 1, characterized in that: In step S1, the charging pile mainboard system collects m characteristic data x=(x1,x2,x3,…,x4) through intelligent sensors, including the current power of the electric vehicle, the charging gun access time, the charging gun temperature, the battery temperature, the electric vehicle BMS required voltage and current, the electric pile output power, the phase power consumption, the phase fundamental power consumption, the phase power factor, etc. m ) T Perform n collections for the preliminary data preparation of intelligent charging scheduling. The data is represented by x1, x2, x3, …, x n , where x j =(x 1j ,x 2j ,x 3j ,…,x mj ) T represents the jth feature data, x ij represents the i-th variable in the data collected for the j-th time, j = 1, 2, ..., n; the data collected by the charging pile is represented by the data matrix X, which is recorded as:

3. A charging scheduling method for a multi-gun charging pile according to claim 2, characterized in that: In step S2, it specifically includes: After obtaining the data matrix X, we can obtain the data mean and data covariance; the data mean vector Expressed as The data covariance matrix S is S = [s ij ] m×n in is the sample mean of the i-th feature data, is the sample mean of the jth feature data; the data correlation matrix R is: in 4. A charging scheduling method for a multi-gun charging pile according to claim 3, characterized in that: In step S3, the charging pile mainboard cleans the data, and uses the principal component analysis method to convert the collected data into a small number of linearly independent data using orthogonal transformation. This method is used to determine the relationship between the eigenvalues ​​in the data collected by the charging pile, and to screen out the most important characteristic data; the singular value decomposition method of the data matrix is ​​used, which specifically includes: 1) The data collected by the charging pile is normalized to obtain a normalized new data matrix, which is still represented by X. Specifically, the original feature data is transformed as follows: in In this method, the normalized variables Still recorded as x ij , the normalized data matrix is ​​still represented by X; 2) According to the normalized data matrix X, calculate the data feature correlation matrix R, 3) Calculate k eigenvalues ​​and corresponding k unit eigenvectors of the feature data correlation matrix R; Solve the characteristic equation of R |R-λI|=0, where I is the identity matrix, and get the m eigenvalues ​​of R λ1≥λ2≥…≥λ m Calculate the variance contribution rate The number of principal components k that reaches the predetermined value; the i-th principal component y i The variance contribution rate of y is defined as i The ratio of the variance of to the sum of all variances, denoted as η i Find the unit eigenvectors corresponding to the first k eigenvalues, a i =(a 1i ,a 2i ,…,a mi ) T ,i=1,2,…,k; 4) Find the principal components of k data, perform linear transformation with k unit eigenvectors as coefficients, and find the principal components of k data: 5) Calculate k principal components y j With the original variable x i The correlation coefficient ρ(x i ,y j ), and k principal components for the original variable x i Contribution rate v i ; where α ii is the variable x i The variance of the covariance matrix Σ is the diagonal element of the covariance matrix; its mean vector is μ μ=E(x)=(μ1,μ2,…,μ m ) T Covariance matrix Σ = E[(x-μ)(x-μ) T ] According to the contribution rate v i Sort from large to small and select the first n data that meet the requirements; 6) Calculate the k principal component values ​​of n data, and substitute the normalized feature data into the formula in 4) to obtain the principal component values ​​of n feature data; the jth data x j =(x 1j ,x 2j ,x 3j ,…,x mj ) T The i-th principal component value of 7) Input the k principal component values ​​of n data into the model M to obtain the prediction result p = M(y'), that is, p = M(y 11 ,y 12 ,…,y ij ); Model M is not specifically defined here. Model M can download a general pre-trained model PM from the Internet for fine-tuning according to actual needs, or it can train the model M by itself after cleaning the charging pile data obtained in the production environment; the data used to fine-tune the model will be labeled in advance, and the label is the priority; the sorted data is input into the model PM for fine-tuning to obtain a dedicated model M, so that model M can fit the label to give a priority p that meets the current specific usage scenario.

5. A charging scheduling method for a multi-gun charging pile according to claim 4, characterized in that: In step S4, the charging pile mainboard executes the charging task according to the scheduling algorithm combined with the prediction result p. t When it is 1, the charging time is unlimited and the waiting time is 0; If the number of vehicles is greater than 2, the charging scheduling algorithm starts to execute, including: 1) First, set the waiting time T w and charging time T c The waiting time and charging time are initialized to T minutes. The charging pile determines whether it is in the peak period, flat period or low peak period according to the current time. According to different peak periods, the waiting time and charging time are set to T / 2 minutes, T minutes and T*2 minutes respectively. 2) If a new vehicle is connected, charging will start immediately, and the charging time is T n minutes, after charging, the priority p is reduced by 1 level, and other vehicles wait for T w Response extension T n minute; 3) For the vehicle being charged, calculate the output power Wo. If the output power Wo drops to the rated power W c When the charging time of the vehicle is reduced to 1 / 2 of c / 2 minutes, and lower the priority p, otherwise charge normally; 4) When the vehicle finishes charging or is fully charged, it exits the queue; The above steps are performed continuously until all vehicles are charged and the queue is empty.

6. A charging scheduling device for multi-gun charging piles, characterized by: The device adopts the method as described in any one of claims 1 to 5.

Images