Rapid urbanization and population growth will cause severe urban traffic congestion that will impose alarming economic and social costs. In the modern world of smart technologies, it is past time to develop more sustainable, energy-efficient transportation systems to alleviate congestion and reduce harmful emissions. Novel connected and automated vehicle (CAV) technologies provide significant opportunities for reducing traffic crashes, enhancing mobility, and minimizing energy cost by using vehicle-to-vehicle and vehicle-to-infrastructure communications. However, key challenges of CAV-based traffic network control remain pertaining to how to coordinate multiple intersections and how to solve the resulting large-scale problems efficiently. In our lab, we are developing computational optimal control and machine learning techniques to reach satisfactory solutions for traffic signal and CAV control within limited computational budget. Both centralized and distributed methods are pursued for control of isolated intersections, multi-intersection traffic corridors, and traffic networks.