Studying Autonomous Vehicles Policies with Urban Planning of Toa Payoh in Singapore

(WP4 – Electric Autonomous Vehicle Charging Infrastructure Planning with Power Consideration)

Professor Meng Qiang, Professor Ong Ghim Ping, Raymond, Professor Lee Der-Horng

This summary describes the work we (NUS team) have done throughout the project “Studying Autonomous Vehicles Policies with Urban Planning of Toa Payoh in Singapore”. This is a 3.5-years project (10/02/2017 – 09/08/2020), funded by the Ministry of National Development under the Land and Liveability National Innovation Challenge Research Programme. The project involves three research institutions (SMART-FM, SEC-FCL, and NUS) and four agencies (MOT, LTA, URA and HBD). The project opened an initiative to explore the feasibility of deploying electric autonomous vehicles (EAVs) as a complimentary option of public transportation in Singapore, and investigate their land and transport implications. NUS team is responsible for work package 4, “Electric autonomous vehicle charging infrastructure planning with power consideration”, where two research objectives/focuses are included:

• Charging infrastructure planning: Identifying optimal charging station location and the charging infrastructure configuration (the no. of fast and/or normal chargers at each charging station) for EAVs in Singapore.
• Energy consumption estimation: Estimating energy consumption of EAVs at optimum charging infrastructure deployments under future urban planning scenarios.

For the first research focus, we develop a modelling framework based on microscopic traffic data of EAV trajectories generated from traffic simulation platforms for the charging facility deployment problem. It takes into account several practical factors, including multiple optimization objectives, battery degradation, charger mix and multi-type EAVs. The developed model comprises three functional modules: (i) candidate charging location determination, (ii) charging facility deployment (CFD) and (iii) refinement of charging infrastructure planning solution. Under the CFD module, charging demand prediction considering battery degradation and vehicle heterogeneity, charging demand allocation and charger configuration optimization are executed sequentially. The CFD solution is tuned by a backward elimination method to find a more economic planning solution where the minimal number of chargers at each station can be specified.For the second research focus, we propose an easy-to-implement method to estimate network-wide EAV energy consumption based on their trajectories and the unit energy consumption functions. The unit energy consumption functions in battery degradation rate are derived by linear regression method using EV operational data in the US. For energy consumption, we assume EAVs’ battery degradation rates follow truncated normal distribution and calculate daily travel mileage of each EAV based on its trajectories, and thus obtained expected total energy consumption in the network. We also propose a prorated assignment approach to determine the expected energy consumed at each EAV charging point based on the estimated temporal-spatial charging demand distribution from EAVs' trajectories.

The developed models were validated using both real taxi data and island-wide traffic simulation outputs under full and partial automation planning scenarios. Our primary findings are summarized as follows:

• 15% less chargers are needed and 7.5% energy are saved under partial automation planning scenario where all private cars are gasoline powered.
• Compared to the case where we have ICE-based private cars, a partial automation planning scenario of 25% private cars converted to EAVs will result in a 26.8% increase in energy consumption.
• More than 10% of investment could be saved by using our refinement mechanism.
• Overlooking battery degradation would lead to over 10% estimation error for energy consumption.

1. Wang, H., Zhao, D., Meng, Q., Ong, G.P., Lee, D.H. (2019). A four-step method for electric-vehicle charging facility deployment in a dense city: an empirical study in Singapore. Transportation Research Part A: Policy and Practice, 119: 224-237.
2. Cai, Y.T., Wang, H., Meng, Q., Ong, G.P., Lee, D.H. (2019). Investigating user perception on autonomous vehicle (AV) based mobility-on-demand (MOD) services in Singapore using the logit kernel approach. Transportation, 46, 2063-2080.
3. Wang, H., Zhao, D., Cai, Y.T., Meng, Q., Ong, G.P. (2019). A trajectory-based energy consumption estimation method considering battery degradation for an urban electric vehicle network. Transportation Research Part D: Transport and Environment, 74: 142-153.
4. Xu, M., Meng, Q. (2020). Optimal deployment of charging stations considering path deviation and nonlinear elastic demand. Transportation Research Part B: Methodological, 135, 120-142.
5. Wang, H., Zhao, D., Meng, Q., Ong, G.P., Lee, D.H. (2020). Network-level energy consumption estimation for electric vehicles considering vehicle and user heterogeneity. Transportation Research Part A: Policy and Practice, 132: 30-46.
6. Wang, H., Zhao, D., Cai, Y.T., Meng, Q., Ong, G.P. Taxi Trajectory Data Based Fast-charging Facility Planning for Urban Electric Taxi Systems. Submitted to Transportation Research Part D for review.