Highly Automated Electric Vehicle (HAEV)-based mobility-on-demand system modeling and optimization framework in restricted geographical areas

yy Recent rapidly developing automated vehicle (AV) technology has been widely recognized as one focus of the industry, government administrations, and academia. To leverage and accelerate the real-life applications of AV technology, a potential also feasible way is to only operate lower level AVs in Operation Design Domain (ODD)s. This way could significantly improve the mobility and safety of AV-based transportation systems. However, works to date are only focusing on one or two scenarios where AV technology has been fully developed and tested. Also, the sophisticated new features introduced by lower-level AV technologies are not considered comprehensively in existing AV-driven transportation systems. Thus, to leverage and accelerate AV-driven transportation systems, one lower-level automation technology, highly automated vehicle (HAV), operating in a prescribed ODD, is comprehensively quantified. And its integration with both electric vehicle (EV) technology and commercially running mobility-on-demand (MoD) system is fully modeled and evaluated in this work. In detail, to describe the application of the HAEV-based MoD system, a proposed integrated methodology thoroughly investigates multiple system uncertainties. The proposed method includes a quantitative traffic ODD modeling architecture, a driving complexity quantification method of ODDs, system continuity variation, and a unified joint optimization model of vehicle-trip and station-vehicle assignments. Numerical experiments are conducted to validate the proposed methodology. The experimental results demonstrate the effectiveness of the proposed method from multiple perspectives. 1) Via the proposed ODD modeling and driving complexity quantification method, the representation and driving complexity of eight typical ODDs are explored, and the study area is classified into four levels of ODDs. 2) The ODD-based MoD s system can offer much more efficient vehicle service and charging performance while maintaining a high vehicle and station utilization. 3) With the AV technology, there is a significant decrease in the cost time for users on-board and vehicles driving to charge activities. With the proposed assignment optimization scheme, both the trip and charging wait time decrease sharply, and the unusual trip and charging wait events happen much less. 4) The overall saved energy consumption and reduced CO2 emissions by HAEVs are massive, which are 1,984.51 GJ and 2,323.51 kg, respectively. The total ridership length grows a lot at a complexity Level 6 ODD compared to at a complexity Level 1 ODD though the cost per user ride per km is doubled. (C) 2020 Elsevier Ltd. All rights reserved.