Siting public electric vehicle charging stations in Beijing using big-data informed travel patterns of the taxi fleet

Abstract

Charging infrastructure is critical to the development of electric vehicle (EV) system. While many countries have implemented great policy efforts to promote EVs, how to build charging infrastructure to maximize overall travel electrification given how people travel has not been well studied. Mismatch of demand and infrastructure can lead to under-utilized charging stations, wasting public resources. Estimating charging demand has been challenging due to lack of realistic vehicle travel data. Public charging is different from refueling from two aspects: required time and home-charging possibility. As a result, traditional approaches for refueling demand estimation (e.g. traffic flow and vehicle ownership density) do not necessarily represent public charging demand. This research uses large-scale trajectory data of 11,880 taxis in Beijing as a case study to evaluate how travel patterns mined from big-data can inform public charging infrastructure development. Although this study assumes charging stations to be dedicated to a fleet of PHEV taxis which may not fully represent the real-world situation, the methodological framework can be used to analyze private vehicle trajectory data as well to improve our understanding of charging demand for electrified private fleet. Our results show that 1) collective vehicle parking "hotspots" are good indicators for charging demand; 2) charging stations sited using travel patterns can improve electrification rate and reduce gasoline consumption; 3) with current grid mix, emissions of CO 2 , PM, SO 2 , and NO x will increase with taxi electrification; and 4) power demand for public taxi charging has peak load around noon, overlapping with Beijing's summer peak power.

Two-Stage Stochastic Choice Modeling Approach for Electric Vehicle Charging Station Network Design in Urban Communities

“…The authors of [27] formulated a charging station location problem with a focus on human factors. To support recent developments in the electrification of public transportation, the authors of [7], [28], and [56] developed models to optimally determine the locations of charging stations for electric taxis and buses. The impact of different types of EVs ( [14], [52]) , locations and sizes of charging infrastructures ( [31], [43], [45], [63]) on power networks has also been investigated by various studies.…”

“…In the problem, constraints (7), (12), (13), and (15) are referred to as the linking constraints because of the presence of the first-stage variables x and z in the second-stage, which links the two stages. Let A be the coefficient matrix for variables z n,j in the linking constraints (7), where a i,q is the entry of matrix A at indices i and q, and a i,q ∈ R |N ||J||T |×|N ||J| . In addition, let F and G be the coefficient matrices for variables x n,j in the linking constraints (12) and ((13), (15)), respectively, where f i,q and g i,q are the entries of the matrices F and G at indices i and q, and f i,q ∈ R |N ||J||T ||B|×|N ||J| , g i,q ∈ R 2|T ||N ||N ||B||J||S|×|N ||J| .…”

Conceptualizing Big Data: Analysis of Case Studies

“…Several studies discussed enhancing the decision-making processes, enabling data-driven decision-making, or providing actionable insights to managers (Bärenfänger et al, 2014;Krumeich, Jacobi, Werth, & Loos, 2014;Dutta & Bose, 2015). Several studies (Cai et al, 2014;Tao, Corcoran, Mateo-Babiano, & Rohde, 2014;Kalakou, Psaraki-Kalouptsidi, & Moura, 2015) also investigated transportation or passenger patterns, providing insights into planning and decision-making. Embedding analytics and insights into processes and decision-making routines is important (Bekmamedova & Shanks, 2014).…”

A realistic framework to a greener supply chain for electric vehicles

“…Pevec et al proposed a data‐driven approach using predictive analytics to decide optimal charging station locations. Former studies have focused on determining the optimal station locations based on vehicles' movement and driving patterns . Khalaf and Wang have studied ways to introduce tariffs that may sway EVs charging patterns towards low grid impact times.…”

“…Former studies have focused on determining the optimal station locations based on vehicles' movement and driving patterns. [10][11][12] Khalaf and Wang 13 have studied ways to introduce tariffs that may sway EVs charging patterns towards low grid impact times. The tariffs' economic and environmental impacts were also analyzed.…”