The objective of this PhD study is to develop an eco-driving system for a mixed traffic consisting of CAVs and human-driven vehicles (HVs) on urban roads considering multiple objectives, including efficiency, safety, fuel consumption and emissions.
Congestion, fuel consumption, and emissions are major costs of modern transportation systems, which impede sustainable development due to their enormous negative impact on mobility, the economy, public health, and the environment.
The recent advances in mobile cloud computing and Internet of Things (IoT) have greatly enabled the modern intelligent transport system with high safety and efficiency. In particular, the emergence of connected and automated vehicles (CAVs) holds great potential for improving traffic efficiency and energy savings.
On urban roads, traffic is interrupted by signal control or giving way to vehicles from other approaches at the non-signalised intersection. The stop-and-go traffic oscillations lead to hard acceleration and braking as well as idling, which are associated with higher energy consumption and emission.
The proposed eco-driving system’s performance will be rigorously assessed by using both simulations and real traffic data.
Even in partial CAV environments, a scenario which is expected to form at the early stage of CAV development, information related to neighbouring vehicles, road condition, traffic management, etc. will be available through a vehicular communication network, which can be used to guide CAVs using customised driving trajectory for various purposes.
Nevertheless, saving fuel consumption and reducing emissions are often considered as secondary, compared with other objectives such as improving traffic safety, reducing traffic congestion, etc.
For many studies that focus on fuel consumption and vehicle emissions, traffic efficiency is often ignored or inadequately considered, that is, the developed eco-driving strategies can instead result in harming traffic efficiency.
In fact, in the literature there is a lack of a comprehensive modelling framework that jointly considers energy saving, emission reduction, traffic efficiency, and safety.
This is largely due to four factors:
- Road traffic has been traditionally operated and controlled with the goal of moving traffic safely and efficiently.
- The management and control strategies are based on traffic flow models (TFM), which are derived purely from vehicular movements (trajectories) without considering the impact of energy consumption and emissions, and, consequently, lack a jointly optimal design for traffic efficiency and energy saving.
- Comprehensive data including trajectory, fuel and emission rate are very difficult to collect, which makes it challenging to develop models that aim to simultaneously optimise traffic efficiency, and minimise fuel consumption and vehicle emissions, especially for CAVs; and
- It still lacks a credible and authentic simulation platform that integrates traffic dynamics with energy consumption and vehicle emission, and provides extensive assessments for traffic efficiency, energy efficiency, and vehicle emissions at both the individual level and the large-scale network level.
In transitioning to the era of CAVs, it is urgent to develop traffic operation and control strategies that not only maximise travel efficiency but also simultaneously minimise vehicles’ energy consumption and emission. It is this that has motivated this project.
This project aims to develop eco-driving strategies in a mixed traffic consisting of CAVs and human-driven vehicles (HVs) to improve both mobility and sustainability, which will be simulated and evaluated for various real-world scenarios in an authentic simulation platform.
In addition, lane-changing (LC) in the vicinity of an intersection is also a disruptive behaviour, which can be classified into mandatory lane-changing (MLC) and discretionary lane-changing (DLC). The former usually happens when vehicles have to change to the correct lane in order to reach the planned destination while the latter occurs when vehicles are pursuing a better driving experience, e.g., for a higher speed.
Inappropriate decision or planning of LC can create safety hazards, cause delay, unnecessary oscillations and extra energy use and emissions in at least two lanes (the original and the target lane). Accordingly, eco-strategies proposed by this project will mainly focus on:
- Longitudinal speed guidance; and
- MLC and DLC decision and execution in a mixed traffic scenario consisting of CAVs and HVs in the vicinity of signalised intersection, with the goal of minimising fuel consumption and guaranteeing safety without sacrificing traffic efficiency.
Develop an information-based control framework for CAVs and develop novel control strategies in a mixed traffic network including both CAVs and HVs, using cutting-edge optimisation control algorithms to reduce congestion while minimising environmental impairment. Strategies for longitudinal and lane-changing motion are designed and integrated into the system.
Evaluate the performance of these control strategies based on both simulated traffic and real-world data using a comprehensive simulation platform that is able to integrate a traditional traffic simulation module, a vehicle energy consumption and emission module, a vehicular communication module, and an autonomous driving simulator.
More from iMOVE