With the development of communication technology, artificial intelligence and new energy resources, transportation will undergo revolutionary changes.
Alternative-fuel vehicle (e.g., electric vehicles (EVs)) can greatly reduce urban traffic pollution and emissions of greenhouse gas, while the further promotion of EVs desires transportation support. THU-LEAD is dedicated to the study of driving behaviors of EV drivers, so as to set up the planning and operation methods of urban and intercity charging facilities (including public charging stations, wireless charging lanes, etc.). In order to promote the electrification of transportation system, THU-LEAD also studies the methods to realize interconnection of traffic network, energy network and information network. At the same time, THU-LEAD is also exploring new charging service modes (such as mobile charging service) and expanding them to multi-mode transportation systems covering taxis, ride-sourcing cars and buses through reasonable mechanism design, and thus the economic costs of various participants in electric transportation can be further reduced.
Traffic big data provides large space/a broad view for scholars and engineers to understand the traffic mechanism. Through deep cooperation with the government and enterprises, THU-LEAD has obtained high-quality multi-source traffic data (e.g., monitoring control system (MCS) data, mobile phone cellular data, smart cards data and so on). Based on data mining, machine learning and other frontier technologies, THU-LEAD has constructed a set of data-driven research methods, including traffic pattern recognition, congestion analysis and state prediction methods. Real-world networks in Beijing, Tianjin and other cities have been taken as examples to calibrate and test the methods. THU-LEAD has also developed a real-time regional traffic control method based on surrogate model and reinforcement learning, combining active guidance with traffic big data. The model effect has been validated and evaluated through computer simulation.
The development of automated driving technology has promoted the transformation of traffic industry. The key research content of THU-LEAD is how to make better use of automated driving technology to improve the level of transportation service. Automated vehicles (AVs) are highly controllable and can accurately execute a variety of motion instructions, which makes it possible to design new organization and control methods at the intersection, arterials and even road network level, and thus greatly improve the capacity of road systems. Roadside intelligent infrastructure can further enhance the advantages of AVs. On the other hand, automated driving technology frees the driver's hands and enables people to plan their travel arrangements more flexibly. For this reason, THU-LEAD proposes a new human-vehicle interaction model to adapt to new changes, improve travel efficiency and safety, and reduce transportation and operation costs.