นิภาภัทร เรืองแข. Simulation-based method for locating the charging stations for electric vehicles on the National Highway of Thailand. Master's Degree(Logistics and Supply Chain Management). มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าธนบุรี. สำนักหอสมุด. : King Mongkut's University of Technology Thonburi, 2017.
Simulation-based method for locating the charging stations for electric vehicles on the National Highway of Thailand
Abstract:
Charging Stations (CSs) are required for electric vehicles (EVs) in order to recharge the battery. EVs cannot travel in the long journey if there is no CS available on the
highways, In this study, we present a conceptual model for identifying the location of CSs on the National Highway No.1 (NH-1) also known as Phahonyothin Road. This
model was simulated with ARENA 14.0. There are 53 survey points (SPs) on NH-1. This study considers two different cases of road track: one road track and two parallel road track. As for on road track, our study considers, only from the survey point No.8 (SP-8) to the survey point No.10 (SP-10) on NH-1. there are 2 intervals (from SP-8 to SP-9, and from SP-9 to SP-10). As for two parallel road track, our study considers only the survey pont No.2 (SP-2) to the survey point No.3 (SP-3) on NH-1. There is only an interval. We forecasted and estimated the number of EVs travelling through SP-2, SP-3, SP-8, SP-9 and SP-10 for the year 2036 by using the historical data. We also consider the inflow and outflow of EVs that enter or exit NH-1 via all intersections and imaginary points.
From SP-8 to SP-10, there are 159 actual intersections. We estimated the inflow and outflow from these intersections using the historical data of population density of all
districts that are passed by NH-1 during these 2 intervals. To simplify our simulation model, we ereated 25 imaginary points that represents 159 intersections. We ran the
experiments on our simulation model with your charging behaviors of EV drivers; charging, when the remaining capacity battery (RCB( is less than equal 24, 48, 72, and
96 km under two different scenarios: (1) if one charging circuit per one CS is set up, and (2) if two charging ciruits per one CS are set up.
One the orther hand, from SP-2 to SP-3, there are 33 actual intersections but this interval is more complex than the intervals between SP-8 to SP-10, Since it has a main road and a frontage road that is parallel in this interval thus this conceptual model does not have the imaging point like one road track case. The model of two road tack has four charging behaviors similar to the one road track. However, the experiment of two road 96 km under two different secenarios: (1) if one charging circuit per one CS is set up, and (2) if two charging circuits per one CS are set up.
On the orther hand, from SP-2 to SP-3 there are 33 actual intersections but this interval is more complex than the intervals between SP-8 to SP-10. Since it has a main road and a frontage road that is parallel in this interval thus this conceptual model does not have the imaging point like one road track case. The model of two road tack has four charging behaviors similar to the one road track. However, the experiment of two road track was run under five differnt scenarios. Thery are 5, 10, 15, 20, and 25 charging circuit per on CS, respectively.
The result of experiments indicated that the charging behaviors of EV driver will have an impact on the appropriate locations of CSs. More impotantly, the average waiting time was considered to cetermine the appropriate locations of CS and the appropriate number of charging circuits per one CS. As for one road track case, if there was one charging circuit per one CS, the average waiting time would be the least when CS was located at SP-9. Furthemore, if there were charging circuits per one CS, the average waiting time would be the least when CS was also located at SP-9. As for two parallel road track case, the average waiting time was the least when CS was located at the Exit-6 to Frontage Rd. In other words, our study found that the increasing number of charging circuits per one CS resulted in the decreasing average waiting time. Moreover, if the number of charging circuits were increased, the position of CS with the least average waiting time would be closer to the position of CS with the least simulation model can be used as a fundamental step to identify the appropriate location for constructing a CS for a specific interval of any highway.