A study on the charging load model of electric vehicles in medium-sized cities

Abstract

With the rapid growth of the number of electric vehicles, the impact of the electric vehicle charging load on the grid has gradually increased, which has become an urgent issue to be solved. To reduce the impact of large-scale electric vehicles connecting to the grid, accurate load forecasting is particularly critical. In this paper, the influencing factors of the EV charging load were studied, the charging load characteristics of different types of EVs were analyzed, calculation models of different EV charging loads in a certain area were constructed, a program simulation was performed, a prediction curve of the future charging load was drawn, and the regional EV charging load was established. The future charging load levels of EVs in a certain area were calculated and analyzed to establish the basis for future optimization control schemes.

Keywords

electric vehicle charging load model Monte Carlo simulation charging load calculation

Introduction

With the development of the world economy, the improvement of people’s quality of life, and the sharp increase in population, the worldwide demand for energy is increasing. Although China’s energy reserves are relatively large, the per capita amount for a population of 1.4 billion is relatively low. Environmental deterioration and the energy crisis have provided opportunities for the development of electric vehicles, which have attracted the attention of governments around the world due to their cleanliness and environmental friendliness.¹ Electric vehicles have the following advantages: (1) no or low noise, which is the most intuitive feature; and (2) high efficiency, since the motor will not idle under the condition of uninterrupted starts and stops, which can improve the use of energy. (3) Low environmental pollution. It uses power batteries instead of common fuel oil, generates no exhaust gas during driving, and thus does not pollute the atmosphere; (4) it is less affected by environmental factors and thus has a wide range of use. (5) Simple structure, easy use and maintenance, and durable.

Rapid economic development has vigorously promoted the development of the automobile industry. At present, the number of motor vehicles in China has reached 415 million, of which the number of cars has reached 318 million, which marks the further establishment of China’s status as a major automobile nation. Although the prosperity of the automobile industry can promote rapid economic development, it has also brought many problems, such as environmental pollution and a shortage of gasoline energy, which has caused the development of the automobile industry to be criticized. Therefore, electric vehicles are currently recognized as the best choice for new energy vehicles. Electric vehicles use electric energy as energy source, have almost zero greenhouse gas emissions, and have obvious advantages in operating costs compared with traditional fuel vehicles.²

In the study of charging load model, Monte Carlo simulation method is the most commonly used one. It constructs the probability process of car travel according to the data of car travel by statistical simulation method, and then obtains the simulated and predicted load model from the probability sampling.³ At present, the data model based on the results of the National Household Vehicle Survey released in 2018 (NHTS2017) is widely used, and Monte Carlo method is adopted for load forecasting.^4,5 In order to solve the problem that Monte Carlo simulation requires a large number of samples to build a probability model, some studies combined Markov chain and Monte Carlo method to obtain samples consistent with stable probability distribution, so that the charging load prediction of electric vehicles on the day is only related to the charging load of the previous day.⁶ Some studies also use the spatial-temporal distribution method of travel chain for load forecasting. Compared with Monte Carlo method, this method introduces the user’s travel purpose as an influencing factor and combines the two aspects of time and space distribution for load forecasting of electric vehicles.⁷ However, the above research failed to conduct a detailed analysis of the characteristics of different types of electric vehicles, and was only classified by general data.

In this paper, by discussing the influencing factors of the EV charging load and the charging characteristics of different types of EVs in a certain area, the Monte Carlo simulation method was used to extract the initial charging time and initial SOC, and the charging load of different types of EVs was investigated. Modeling was performed to analyze the characteristics of the charging load. Through the discussion and analysis in this paper, an accurate charging load model of EVs in a certain area can be established, and a future charging load forecast curve can be drawn from this model, which lays the foundation for subsequent optimization control.

Factors affecting the charging load of EVs

Construction of charging facilities

The size and distribution of the charging load are affected by various factors, including the number and distribution of charging facilities, the capacity and type of charging stations (whether it is fast charging or slow charging), and the utilization rate of charging piles. The number and distribution of charging facilities determine the spatial attributes of large-scale EVs connected to the grid. The capacity and type of charging station determine the load characteristics and scale of the charging station. The utilization rate of charging piles determines the impact of charging stations on the overall charging load in different time periods.

Electricity demand

The electricity demand is affected by various factors, including the type of EV, the driving needs of car owners, the type of charging location, and the charging time period. The electricity demand of different types of electric vehicles varies greatly. For example, the charging demands of pure electric vehicles and hybrid electric vehicles are completely different. The impact of the charging time period on the grid is more strongly reflected in the issue of large load fluctuations. Electric vehicle charging is related to the travel pattern of user groups, and the similarity in the travel pattern of a large number of users leads to simultaneity in charging EVs. Therefore, during the charging period, there are clear peaks and troughs. The driving needs of car owners and charging locations are important factors in the size and distribution of electricity demand.

Electricity price and policies

Different electricity prices and policies affect the charging behavior and decision-making of EV users, thus affecting the size and distribution of the charging load. The electricity price and the degree of policy support directly affect people’s use of EVs. On the one hand, low electricity prices can attract more users to purchase and charge EVs, thereby increasing the charging load. On the other hand, high electricity prices may cause users to reduce the use of EVs to charge, resulting in a reduction in charging load.^7,8

Climatic factors

Meteorological factors affect the EV charging load by affecting the ability of EV owners to adjust their travel plan or adjust their charging behavior due to the travel plan, so meteorological factors have the most significant impact on the EV charging load.⁹ In addition, the temperature and road conditions caused by the climate will cause changes in the electricity consumption of electric vehicles per 100 km.

Time factor

The charging load is also affected by the time factor. For example, the distribution of the charging load is different during different time periods, such as holidays and workdays, daytime and night, and weekends and workdays. During different time periods, the charging demands of trams will differ. To analyze the load of EVs, it is necessary to analyze the charging pattern of EVs by time period.

Characteristics of the charging equipment

Different types of charging equipment have different charging efficiencies and powers, and the selection of charging equipment will also affect the charging load. The characteristics of charging equipment include the comprehensive expression of the type, power, charging speed, charging method, and charging efficiency of the charging pile. For example, AC charging piles and DC fast charging piles are quite different in terms of charging power, charging methods, and charging efficiency. AC charging piles are often used for low-power charging and have a slow charging speed. However, intelligent charging methods such as scheduled charging can be used. It is generally used in scenarios such as home charging and small commercial charging. A DC fast charging pile has a higher charging power and can achieve a faster charging speed, which is suitable for scenarios such as expressways.

Modeling method of the EV charging load model

Types of electric vehicles

The load distribution of EVs is application dependent, and the charging behavior and time of different types of vehicles are different. The charging behavior of private cars is uncertain, and the charging time is flexible; the charging regularity of buses is often carried out in the parking lot at the starting station, the charging location of taxis is not fixed, and the charging time is mostly when the owners are resting; the charging time of official vehicles is mostly after performing official duties. Accurately predicting the load distribution of EVs requires the study of load modeling for each type of vehicle. In this study, we divided EVs into four categories, that is, private cars, buses, taxis, and official cars, and performed load modeling.

Charging methods for EVs

There are three charging methods for electric vehicles: home charging, public charging and fast charging.

Home charging: Home charging refers to charging at home using a home charging pile or a common power supply. This charging method is the most common and convenient, but the charging speed is relatively slow, and it generally takes several hours or tens of hours to reach full charging.

Public charging: Public charging refers to charging at the charging facilities provided in public places, shopping malls, hotels and other places. Public charging facilities are generally equipped with various charging modes, such as fast charging and slow charging, and the charging speed is faster than home charging.

Fast charging: Fast charging refers to the use of specific fast charging equipment to quickly charge electric vehicles, and the charging speed is very fast. However, the high charging power of fast charging may cause some damage to the batteries of EVs.

Daily mileage of EVs

The driving mileage of electric vehicles has a major impact on the number of charging times and charging time. The study data show that the average daily mileage of different types of EVs is different. Since the rise of electric vehicles is to replace traditional fuel vehicles, the mileage data of electric vehicles have not yet reached a sufficient level. In the present study, it is assumed that when EVs are used as travel tools, the travel mode of car owners is not affected.

Based on the analysis of bus operation data in a certain area,^7,8 electric buses drive approximately 200 km per day, but the range of battery cruising does not exceed 150 km, which cannot meet daily driving needs. To ensure the normal operation of electric buses, they must be charged at least once during the rest period of the day and be recharged when they are out of service at night to meet their daily driving needs. Research data show that the average daily mileage of electric taxis is approximately 400 kilometers, and their maximum driving mileage is approximately 300 kilometers. To ensure the normal operation of an electric taxi, it must be charged at least twice a day to maintain its normal operation. The driving range of electric business vehicles is relatively short, and they are parked for a long time after finishing official business. Therefore, this paper uses the charging mode of once a day to ensure that the vehicles are always in a good charging state while driving.

In view of the great randomness of the travel pattern of electric private cars, the current research on travel mileage data generally uses American Household Travel Survey data. According to NHTS 2009 data, the average daily driving mileage of electric private cars reaches 65 kilometers, and the fitted probability density distribution function follows a lognormal distribution.⁶ As shown in the following equation:

f (L, μ, σ) = \frac{1}{L σ \sqrt{2 π}} \exp [- \frac{{(\ln x - μ)}^{2}}{2 σ^{2}}]

(1)

According to the literature,¹⁰ the fitting results for μ and σ are 3.20 and 0.88, respectively. The function distribution diagram is shown in the Figure 1.

Figure 1.

Probability density distribution map of daily mileage driven by private cars.

Start time of EV charging

The charging time of EVs will have a profound impact on the distribution pattern of the charging load. Considering the interests of both the grid and the user, we discuss the characteristics that need to be considered when charging EVs and how to reasonably select the charging time. The charging time of electric vehicles varies according to their use; therefore, personalized charging strategies are needed according to different types of vehicles.

Electric buses need to be charged twice during the operation period. They are not charged during peak hours. They are fast charged during the day and conventionally charged at night. According to the operating hours of the electric buses, one fast charge is applied during the day from 10:00 to 16:00, and the other charge is applied at 22:00 at night. Routine charging starts at 24:00.

To ensure the normal operation of taxis, taxis need to perform fast charging during rest hours to meet the needs of long-distance travel. According to the data analysis, the charging time of taxis was between 2:00 and 4:00 in the morning and between 12:00 and 14:00 at noon.

For official vehicles, the charging start time is when the vehicle returns to the parking lot after performing its official duties. Because there is a long parking time for electric official vehicles, this chapter assumes that the starting time period of charging for electric official vehicles is 18:00 to 7:00, and the charging method is conventional charging.

For a private car, it is assumed that its charging start time is when we return home from the last trip. According to the NHTS 2009 statistical data, the end time of the last trip of an electric private car satisfies the normal distribution; that is, the charging start time satisfies the normal distribution,¹¹ as shown in the formula (2):

f (t_{s}, μ_{s}, σ_{s}) = {\begin{cases} \frac{1}{σ_{s} \sqrt{2 π}} \exp [- \frac{{(t_{s} - μ_{s})}^{2}}{2 {σ_{s}}^{2}}] & < t_{s} < 24 \\ \frac{1}{σ_{s} \sqrt{2 π}} \exp [- \frac{{(t_{s} - μ_{s} + 24)}^{2}}{2 {σ_{s}}^{2}}] & 0 < t_{s} < (μ_{s} - 12) \end{cases}

(2)

Right μ_s, σ_s the values are 17.6 and 3.4, respectively, and the function diagram is shown in the Figure 2.

Figure 2.

Probability density distribution of private cars at the initial charging moment.

Calculation of the EV charging load based on the Monte Carlo method

Monte Carlo simulation is a prediction method based on user travel patterns. The main inference is that when a large number of continuously and repeatedly occurring random events exhibit regularity, this regularity is unavoidable. This phenomenon is called “randomness.” “Deterministic” means that people know some basic statistical characteristics and probability parameters in advance so that people can use certain mathematical tools to estimate and infer the relationships between these variables. In recent years, the Monte Carlo simulation method has become widely used in China. The core of the Monte Carlo simulation method is to simulate the daily travel habits of users by performing a large number of repeated random samplings on the travel data of users and using a probabilistic model to make predictions. Monte Carlo estimator F_N The expression for¹² is.

F_{N} = \frac{1}{N} \sum \frac{f (X_{i})}{p (X_{i})}

(3)

where f(X_i) is a random variable X_i is a probability distribution function; p(X_i) is a random variable X_i is a probability density function; and N is the number of simulations.

To calculate the distribution of the charging load versus time for each EV, the continuous charging time of the vehicle and the starting charging time and initial SOC of each EV are also needed. In this paper, it is assumed that the electric car owner will continue to charge the car until the battery is full. However, for an electric car, which rarely stops, such as a taxi, if the permitted charging time is exceeded, regardless of whether the electric car is fully charged, the charging needs to be terminated. Therefore, the following equation is used to calculate the charging duration of an EV:

T = \min (T_{p a r k}, \frac{(1 - S O C) * C}{p * k})

(4)

where T_park represents the parking time of the EV, C represents the battery capacity, P represents the charging power, and k represents the charging efficiency.

Modeling and forecast curve of the EV charging load in a certain area

The specific methods of the regional EV charging load model are as shown in Figure 3.

(1) According to the charging data of charging stations in relevant studies, the total number of electric vehicles m and the probability density function of relevant data are determined.

(2) The mileage, charging start time and residence time of each electric vehicle are extracted according to the probability density function.

(3) Calculate the required charging time according to the charging power, and add up the charging load. The charging load curve of electric vehicle is obtained.

Figure 3.

Calculation process of the EV charging load based on the Monte Carlo method.

The distribution curves of different models of the EV charging load versus time can be obtained based on the Monte Carlo method. Finally, the distribution curve of the total amount of EV charging load in an area is obtained by superimposing the loads of these four types of EVs.

Modeling and forecast curve of the electric bus charging load in a certain area

According to the analysis in the previous section, electric buses need to be charged twice a day to meet their operational needs. Fast charging is used during the daytime between 10:00 and 16:00, depending on the operational power consumption and operation pattern of buses. The battery starts to rotate and perform fast charging at approximately 14:00 noon. The initial state-of-charge (SOC) fit a normal distribution. At night, the buses were stopped from 22:00 to 5:00 the next day. The conventional charging method was used. The charging start time was evenly distributed between 22:00 and 24:00, and the initial SOC value also obeyed a normal distribution. Based on these conclusions, a charging load model for electric buses can be established¹⁰ as Table 1.

Table 1.

Description of the electric bus charging load model in a certain area.

Number of charges	Charging time period	Charging start time	Starting SOC	Chosen charging method	Charge probability
1	10:00–16:00	N (13,0.5²)	N (0.5,0.1²)	Quick charging	1
2	22:00–5:00	U (22,24)	N (0.5,0.1²)	Conventional charging	1

The assumption is that the capacity and charging power of the electric bus battery remain unchanged. The following Table 2 lists the simulation parameter settings for the electric bus.¹³

Table 2.

Parameter settings for electric buses in a certain area.

Reference type (taking a certain region as an example)	Setting value
Number of electric buses in 2030 (units)	20,000
Battery capacity (kWh)	180
Conventional charging power (kW)	22
Quick charging power (kW)	120
Charging efficiency	0.9

With the help of the above flow chart, software simulation is performed to obtain the distribution curve of the electric bus charging load in an area in 2030.

Modeling and forecast curves for the electric taxi charging load in a certain area

According to a survey of taxis in a certain area, electric taxis often charge during the meal or rest time of drivers. Therefore, this chapter considers the uniqueness of electric taxis, assuming that the initial charging time obeys a uniform distribution and that the initial battery state-of-charge (SOC) obeys a normal distribution, to develop a fast charging load model for electric taxis. In summary, a traditional electric taxi charging load model is established⁹ as Table 3.

Table 3.

Description of the charging load model of electric taxis in a certain area.

Number of charges	Charging time period	Charging start time	Starting SOC	Charging method	Charge probability
1	0:00–4:00	U (0,4)	N (0.3,0.1²)	Quick charging	1
2	12:00–14:00	U (12,14)	N (0.3,0.1²)	Quick charging	1

In this paper, the assumptions are that the battery capacity and charging power of the electric taxis are constant, the batteries need to be charged twice a day, and the fast charging method is used for all of them. The simulation parameters of the electric taxi are set as shown in the Table 4.¹¹

Table 4.

Parameter settings for electric taxis in a certain area.

Reference type (taking a certain region as an example)	Setting value
Number of electric taxis in 2030 (units)	50,000
Battery capacity (kWh)	50
Quick charging power (kW)	80
Charging efficiency	0.9

With the help of the above flow chart, software simulation is performed to obtain the distribution curve of the charging load of electric taxis in a certain area in 2030.

Modeling and forecast curve for the official electric vehicle charging load in a certain area

Electric official vehicles are charged only after the completion of official tasks, the daily driving mileage is relatively short, and only one charge is needed to meet travel needs. According to the survey data of official vehicles in a certain area, the initial charging times of official electric vehicles are uniformly distributed, and the distribution of the initial SOC is in line with the normal distribution. In summary, there is a conventional charging load model for official electric vehicles⁹ as Table 5.

Table 5.

Description of the charging load model for official electric vehicles in a certain area.

Number of charges	Charging time period	Charging start time	Starting SOC	Charging method	Charge probability
1	19:00–7:00	U (19,23)	N (0.4, 0.1²)	Conventional charging	1

In this paper, the assumptions are that the battery capacity and charging power of the electric business vehicles are constant, the batteries are charged once a day after performing official duties, and the conventional charging method is used. The simulation parameters for the electric official vehicle are set as shown in the Table 6.¹¹

Table 6.

Parameter settings for official electric vehicles in a certain area.

Reference type (taking a certain region as an example)	Setting value
Number of electric office cars in 2030 (units)	200,000
Battery capacity (Kw∙h)	45
Conventional charging power (kW)	7
Charging efficiency	0.9

With the help of the above flow chart, software simulation is performed to obtain the distribution curve of the charging load of official electric vehicles in a certain area in 2030.

Modeling and forecast curves for the charging load of electric private cars in a certain area

The charging time of electric private cars is random, and there is no fixed time period. Based on the above description, the initial charging time of private cars satisfies the normal distribution, and the SOC of the battery during charging can be calculated depending on the daily driving mileage. The calculation formula is shown in the following equation:

S O C = (1 - \frac{L}{L_{m}}) * 100 %

(5)

It is assumed that an electric private car is charged as soon as it comes home, and slow charging and fast charging coexist. In view of the fast charging behavior, the present study assumes that the possibility that private electric cars will use fast charging to supply electric energy in residential areas is 0.2. In summary, the charging load model of electric private cars is established¹⁴ as Table 7.

Table 7.

Description of the disordered charging load model for private cars with fast charging behavior in a certain area.

Number of charges	Charging time period	Charging start time	Daily mileage	Charging method	Charge probability
1	All day	Piecewise normal distribution	Logn (3.2,0.88²)	Slow charging	0.8
1	All day	Piecewise normal distribution	Logn (3.2,0.88²)	Quick charging	0.2

Based on the above analysis, this paper assumes that the battery capacity and charging power of the electric private car are constant, the charging is once a day, the charging method is fast charge charging behavior, and the simulation parameters for the electric private car are set as Table 8.

Table 8.

Parameter settings for electric private cars in a certain area.

Reference type (taking a certain region as an example)	Setting value
Number of electric private cars in 2030 (units)	800,000
Battery capacity (kWh)	40
Slow charging power (kW)	3.3
Quick charging power (kW)	28
Electricity consumption per 100 kilometers (kWh)	20
Charging efficiency	0.9

With the help of the above flowchart, software simulation is performed to obtain the charging load distribution curve of the disordered fast charging behavior of electric private cars in a certain area in 2030.

Forecast curve of the total EV charging load in a certain area

The above four different charging load datasets are superposed to obtain the total curve of the EV charging load. The results are as follows.

Results analysis

In 2030, it is estimated that there will be 20,000 electric buses, 50,000 electric taxis, 200,000 office cars and 800,000 private cars, satisfying the demand of a medium-sized city.

Electric bus

The prediction curve in Figure 4 shows that the high charging load of electric buses is caused by the large battery capacity. During the day, in the fast charge mode, the load reaches the maximum value. The maximum load point is approximately 2 pm, and the peak value reaches 1074691 kW. This shows that the charging load of electric buses is quite large during the day. When conventional charging is used at night, the peak load is more than twice that of fast charging.

Figure 4.

Charging load curve of electric buses in a certain area in 2030.

Therefore, a fine arrangement of the operating hours and charging times of electric buses and the avoidance of centralized fast charging during the day are effective means to reduce the burden on the grid and improve the stability of the grid.

Electric taxis

According to the prediction curve shown in Figure 5, due to the fast charging method for electric taxis, their peak charging load is relatively high. During a day, the charging load of electric taxis will have two peaks, which correspond to the rest of the time of the electric taxi drivers. During these two charging time periods, the peak loads are approximately 1:00 am and 12:30 noon, and the peak loads are 260371 kW and 259201 kW, respectively.

Figure 5.

Distribution curve of the charging load of electric taxis in a certain area in 2030.

In the face of such load characteristics, the characteristics of the taxi charging load and the local load conditions should be fully considered when selecting the site and capacity of the electric taxi charging stations, and the load curves of the two parties should be made complementary to each other to promote the stable operation of the grid.

Electric official vehicle

The charging load forecast curve for official electric vehicles is shown in Figure 6. The main load is distributed between 5 pm and 4 am. At approximately 22:40, the peak load of the electric service vehicles was 860819 kW. In addition, the charging time is concentrated at night, which meets the requirements of orderly charging. Smart charging methods such as ordered charging can be combined to implement peak shaving of the grid load and prevent load peaks caused by centralized charging at night.

Figure 6.

Distribution curve of the charging load of official electric vehicles in a certain area in 2030.

Electricity private car

According to the curve shown in Figure 7, when EVs are quickly charged out of sequence, the peak charging load will occur between 20:00 and 24:00 at night, and the peak value can reach 115116 kW. According to the prediction curve, the peak period of EV charging is the peak period of residential electricity consumption, which leads to the “peak upon peak” phenomenon of the load in residential areas. The charging load should be guided by ordered charging means such as the time-of-use electricity price, or the impact of the charging load on the grid can be alleviated by the introduction of a new energy microgrid.

Figure 7.

Charging load distribution curve of the disordered and fast charging of electric private cars in a certain area in 2030.

Total EVs

It can be seen from the forecast curve of the total charging load of EVs in Figure 8 that the charging load of EVs fluctuates significantly due to the fast charging of taxis and buses between 22:00 and 1:00 the next day or at 12:00 between 14:00 and 14:00; additionally, two load peaks appear at 1226833 kW and 1326806 kW, respectively. These fluctuations do not necessarily have a negative impact on the grid. To change the impact of the EV charging load on the grid, the local load curve must be analyzed, and the user charging load must be guided to perform peak shaving and filling in combination with cutting-edge orderly charging, charging station siting and capacity fixing, V2G and new energy grid technologies. Valley, to tap the potential of mobile energy storage devices for electric vehicles.

Figure 8.

Distribution curve of the total EV charging load in a certain area in 2030.

Conclusions

This paper starts with the factors affecting the EV charging load, analyzes the load modeling methods of different types of EVs, and uses the Monte Carlo method to predict the EV load in a certain area. The load characteristics and impact of different types of EVs on the grid are analyzed, and suggestions are proposed on how to alleviate the impact of different types of EV charging loads on the grid. It is concluded that large-scale electric vehicle access will have a considerable impact on grid load fluctuations. In the future, technologies such as siting and capacity fixing of charging stations, orderly charging, V2G, and new energy grids need to be used to guide and dispatch electric vehicle charging loads to achieve smart charging for cars and peak shaving for grids. Although the popularity of electric vehicles brings challenges to the power grid, its energy storage characteristics, if properly deployed, will have a positive impact on the economy and stability of the power grid. Orderly charging of electric vehicles aims to optimize the load curve of the grid by guiding and controlling the charging behavior, reduce the load variance, and reduce the demand for installed power generation capacity. Through the interaction strategy between electric vehicles and the power grid, real-time power balance strategy, and orderly charging strategy, the coordination and interaction between the power grid and electric vehicles are realized, and the charging behavior is optimized to reduce the charging cost of users and maximize the consumption of renewable energy.

In the further study of future charging load modeling, considering the expansion of the scale of electric vehicles and the further improvement of charging equipment technology, the impact of future charging load on the power grid will be further increased. This means that more accurate charging load prediction is needed to advance the dispatch of the supply grid. Therefore, the future charging load modeling needs to consider more factors to improve the accuracy, such as the impact of the surrounding environment and functional buildings on the electric vehicle load, the impact of different weather on the charging load, the impact of changes in holiday travel rules on the charging load, and so on. The ideal situation is definitely to go to various places to collect the latest charging data for analysis and modeling, but the workload is too large. The historical charging data of a region can be used for modeling, and the predicted charging trend of the region can be used for simulation and simulation, and the modeling deviation can be constantly corrected to finally achieve a load model close to the real one.

Statements and declarations

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Husain

Ozpineci

Islam

, et al. Electric drive technology trends, challenges, and opportunities for future electric vehicles. Proc IEEE 2021; 109(6): 1039–1059. DOI: 10.1109/JPROC.2020.3046112.

Aghajan-Eshkevari

Azad

Nazari-Heris

, et al. Charging and discharging of electric vehicles in power systems: an updated and detailed review of methods, control structures, objectives, and optimization methodologies. Sustainability 2022; 14(4): 2137. DOI: 10.3390/su14042137.

Zhao

Wang

Shen

ZJM

, et al. Data-driven framework for large-scale prediction of charging energy in electric vehicles. Appl Energy 2021; 282: 116175. DOI: 10.1016/j.apenergy.2020.116175.

Yin

. Orderly charging strategy of electric vehicle based on improved PSO algorithm. Energy 2023; 271: 127088. DOI: 10.1016/j.energy.2023.127088.

Zhang

Yin

Zhang

, et al. A wind power curtailment reduction strategy using electric vehicles based on individual differential evolution quantum particle swarm optimization algorithm. Energy Rep 2022; 8: 14578–14594. DOI: 10.1016/j.egyr.2022.10.442.

Ghasemi-Marzbali

. Fast-charging station for electric vehicles, challenges and issues: a comprehensive review. J Energy Storage 2022; 49: 104136. DOI: 10.1016/j.est.2022.104136.

Tao

Huang

Chen

, et al. Orderly charging strategy of battery electric vehicle driven by real-world driving data. Energy 2020; 193: 116806. DOI: 10.1016/j.energy.2019.116806.

Shi

, et al. Data-driven spatial-temporal prediction of electric vehicle load profile considering charging behavior. Electr Power Syst Res 2020; 187: 106469. DOI: 10.1016/j.epsr.2020.106469.

Feng

Chang

Fan

, et al. Electric vehicle charging load prediction model considering traffic conditions and temperature. Processes 2023; 11(8): 2256. DOI: 10.3390/pr11082256.

10.

Elahe

Kabir

Mahmud

, et al. Factors impacting short-term load forecasting of charging station to electric vehicle. Electronics 2022; 12(1): 55. DOI: 10.3390/electronics12010055.

11.

Zheng

Shao

Zhang

, et al. A systematic methodology for mid-and-long term electric vehicle charging load forecasting: the case study of Shenzhen, China. Sustain Cities Soc 2020; 56: 102084. DOI: 10.1016/j.scs.2020.102084.

12.

Xing

Sun

, et al. Multi-type electric vehicle load prediction based on Monte Carlo simulation. Energy Rep 2022; 8(S9): 966–972. DOI: 10.1016/j.egyr.2022.05.264.

13.

Amara-Ouali

Goude

Massart

, et al. A review of electric vehicle load open data and models. Energies 2021; 14(8): 2233. DOI: 10.3390/en14082233.

14.

Zhang

Yan

Liu

, et al. Daily electric vehicle charging load profiles considering demographics of vehicle users. Appl Energy 2020; 274: 115063. DOI: 10.1016/j.apenergy.2020.115063.