Modelling of large-scale electric vehicles charging demand: A New Zealand case study

Abstract

Due to increasing electric vehicles (EVs) uptakes, power system distribution network will have to accommodate the increased load level for charging EVs. Thus, the importance of a robust power system especially in the distribution network level is indisputable. During the planning or reinforcement stage of distribution networks, it is paramount to have some estimations and analyses done on system-wide EV charging loads that will be placed in the network. Thus, this paper systematically investigates the EV fleet composition, market shares, and charging patterns within New Zealand (NZ) area. A multivariate probabilistic modelling of dependent random variables and cumulative distribution functions is adopted for the accurate estimation of aggregated EV charging demands. Several vehicle travel survey data sets are utilised to quantitatively determine charging behaviours and driving patterns of EVs. The developed methodology based on Monte-Carlo simulation (MCS) is utilised to generate results close to the real use-cases daily power demand, which can be further utilised in the analysis of EV charging strategies. In addition, non-smart

Modeling the Charging Behaviors for Electric Vehicles Based on Ternary Symmetric Kernel Density Estimation

“…To make it simple, a simulation sample was selected. To select the simulated data (whose structure is more similar to that of the original data), the axis ranges of the scatter density maps were set to AT ∈ [6,30] Figure 8. The overall 3D structures of the original data (as shown in Figure 2) and simulated data in case 1 and case 2 are approaching.…”

“…As for a power grid (PG), these challenges originate from the uncertain distribution of the EVs charging load [3][4][5]. In terms of EVs, these challenges stem from the randomness of EVs driving and charging behaviors [6]. As a result, problems such as uneven EVs distribution in the CN, congestion in some charging stations (CSs), and long waiting time for charging, and waste of charging resources are caused [7].…”

“…When EVs charging behaviors are modelled with PE, the variable to be estimated conforming to a certain known probability distribution should be presupposed before estimating parameters of the assumed probability distribution with EVs charging behaviors data for testing. Common PE methods are composed of Gaussian distribution [6], mixed Gaussian distribution [4], mixed Beta distribution [19], and Weibull distribution [22] etc. The premise of using the PE approach is to presuppose that the probability distribution is approaching the actual situation.…”

A Probabilistic Methodology to Quantify the Impacts of Cold Weather on Electric Vehicle Demand: A Case Study in the U.K.

“…On the other hand, it is assumed that all vehicles are identical and impacts of ambient temperatures were not considered. In [21], Monte Carlo simulation is developed to analyze EV charging profiles in New Zealand. Similar to previously discussed study only one type EVs are considered (Nissan Leaf) for light duty vehicles and the simulations were carried out for 253 thousand to 1.69 million EVs.…”

Integration of Electric Vehicles in the Distribution Network: A Review of PV Based Electric Vehicle Modelling

“…Spinning reserve is With the increasing level of EV penetration, the associated technical issues, e.g., system imbalance, decreased stability, and power quality, as well as increased system cost, are becoming more prominent, due to additional energy and power demand. The unidirectional approach, i.e., G2V mode, has been extensively studied in the literature in the form of topics like smart charging [19], safety [20], and control features [21]. The focus of these studies is on minimizing the charging cost [22] or minimizing the impact on the distribution system [23,24].…”