Celebrates Recent PhD Graduates
We are thrilled to announce that three excellent members of our lab have successfully defended their PhD dissertations and graduated on 10th September! This significant milestone reflects their hard work, dedication, and contributions to the field of intelligent traffic and transportation systems.
Graduates:
Advisors: Prof. Xiqun Chen, Dr. Simon Hu
Dissertation Title: “Road Traffic Flow Prediction via Knowledge Transfer and Deep Learning”
Summary: Timely and accurately inferring road traffic flow dynamics is vital in Intelligent Transportation Systems (ITS). Extensive research on traffic state prediction has been carried out in the past, and massive empirical studies have demonstrated the best performance of deep learning techniques. However, the model assumptions and data requirements of deep learning methods are very stringent, and the computational cost is extensively high. Unfortunately, computational resources are normally limited, and data scarcity problems are also inevitable in large urban road networks. Those issues result in a significant prediction performance drop and limited model application scope. To fulfill these research and engineering gaps, this dissertation proposes a series of transfer learning and physics-guided learning approaches, which identifies transferable traffic flow patterns from multiple data-intensive traffic regions and adapts obtained knowledge to target data-sparse traffic regions. Such a novel learning philosophy leads to more accurate, robust and responsive traffic prediction models in the presence of insufficient and imperfect data.
Advisors: Dr. Simon Hu, Prof. Xiqun Chen
Dissertation Title: “Deep Reinforcement Learning-Based Regional Traffic Signal Optimization Under Multi-Source Data Environment”
Summary: This thesis systematically explores the optimization of traffic signal control using multi-source data provided by Cooperative Vehicle Infrastructure Systems. By deeply analyzing macro and micro traffic flow characteristics, the thesis first optimizes the signal timing at the single intersection level. Based on the dual requirements of communication and generalization, this thesis further extends the optimization strategy to the regional level, developing communication collaborative control and scalable regional control methods. These methods not only improve traffic signal control strategies but also adapt to the dynamic traffic environment. They provide new theories and methodologies for intelligent traffic systems, which are of significant practical importance for enhancing urban traffic management efficiency, alleviating congestion, and increasing traffic throughput.
Advisors: Dr. Simon Hu, Prof. Xiqun Chen
Dissertation Title: “On-Demand Meal Delivery: System Modeling and Service Design”
Summary: With the rise of mobile payment technologies worldwide, On-Demand Meal Delivery (ODMD) services have become increasingly popular. Due to the network effect among stakeholders, platforms are expected to take more social responsibilities. Therefore, optimizing ODMD system management has become a new focus in academic research. To this end, important scientific questions include understanding the endogenous interaction among stakeholders, quantifying the impact of ODMD service design on service levels, and exploring the application performances of platform management strategies. To this end, this dissertation focuses on the monopoly market dominated by a single ODMD, aiming to measure the impact of ODMD system management strategies, optimize the strategies, and provide managerial insights for the platform and other system decision-makers.
We congratulate Dr. Junyi Li, Dr. Bin Zhou, and Dr. Anke Ye on their remarkable achievements! Their research has not only advanced our understanding of intelligent transportation systems but has also paved the way for future innovations in sustainable mobility and transportation.
As they embark on the next chapter of their careers, we look forward to seeing their continued contributions to academia and industry. Please join us in celebrating their accomplishments!