Research on Big Data Mining Technology of Electric Vehicle Charging Behaviour

Abstract

Thousands of electric vehicles (EV), which are large in number and flexible in their use of electricity, will be connected to the power system in the near future, which will bring more uncertainty to the power system. Therefore, it is necessary to study the general characteristics of EV charging behaviours. In the charging process, big data regarding charging behaviour of EVs are generated. This paper proposes a big data mining technique based on Random Forest and Principle Component Analysis for EV charging behaviour to identify and analyse clusters with different charging characteristics from the big data. This paper uses Dundee’s January 2018 EV charging data to conduct experiments, and obtains the charging behaviour clusters of the workdays, weekends, and holidays of January. The superiority of the random forest algorithm in the EV clustering problem is reflected when compared to the Euclidean distance method. The clusters obtained by the random forest algorithm have clearer characteristics, including the user’s charging method and travel behaviour. The results show that the charging behaviour of EVs has certain regularity, and the charging load has obvious peak-to-valley difference that is necessary to be regulated.

A Literature Review on the Charging Behaviour of Private Electric Vehicles

“…Different approaches have been used to classify private charge behaviors, applying statistical analysis techniques [87,88], clustering [89,90], or data mining [91], substantially confirming the prevalence of slow night charging on weekdays. Using aggregate analysis of charging demand, one can gain insights into the charging behaviors of different types of users [92].…”

“…The third type concerns a wider range of factors, such as charging power, dwell time, and the cost of home charging [53]. Another study by Y. Liu et al [91] classified EV charging behavior during weekdays, weekends, and holidays for a month in 2018. The findings showed that the largest group of users recharged their EVs primarily during weekdays after dawn, with slow recharges and small amounts of energy.…”

Planning of distributed renewable energy systems under uncertainty based on statistical machine learning

“…In [86], regarding the movements of electric taxis as random walks, the Markov process is used to simulate the distribution of charging demand in static space, while in [87], the random forest (RF) method based on a regression tree is used to predict the driving characteristics of each EV, so as to obtain the travel mode data set of the EV. From the measured charging information and big data mining technology, the EV charging behavior model is presented based on RF and principal component analysis (PCA), which captures the EV with different charging characteristics based on a data-driven model [88]. The gradient boosting model (GBM) is used to simulate the SOC state at the end of a plug-in EV (PEV) daily schedule [89], and in [90], a generalized regression neural network (GRNN) is used to realize power prediction.…”

Integrated profile of EV user behavior and value based on charging data mining

“…The performance of unsupervised user behavior classification is improved by incorporating an adaptive generation strategy of initial clustering centers. The literature [10] divided the big data of charging behavior into three parts according to weekdays, weekends and holidays, and then used the random forest algorithm to perform multidimensional clustering respectively, and the results showed that the charging behavior of electric vehicles has certain regularity. The literature [11] clusters cities into five clusters based on electric bus data from 14 cities using the k-means algorithm for characteristic parameters such as charging power and charging time.…”