The iterative learning approach for vehicle queuing length balanced-control of the signalized isolated intersection

In this work, the iterative learning control approach is proposed to address the signalized isolated intersection control problem. The signal timing should be adjusted according to the real-time daily-repeated traffic demand to ensure the full utilization of green time. However, it is difficult to establish the exact mathematic model of isolated intersection with the time-varying demand. The iterative learning control approach requires much less system modeling knowledge and also rejects the exogenous traffic perturbations. Rigorous analyses show that the proposed learning control scheme can guarantee the convergence of the vehicle queuing length difference to the desired one. The effectiveness of the proposed approach is also verified through simulations.