关键词: 城市交通, 迭代学习控制, 快速路入口匝道控制, 输入受限, 状态受限, 收敛性

Abstract: In actual freeway traffic system, the on-ramp flow and the mainline speed are constrained. Therefore, it is necessary to take these constraints into consideration in controlling the freeway traffic flow. ILC based freeway ramp metering has become a research hotspot these years, however, there is no corresponding convergence analysis under input and state constraints. In this paper, the traffic density control problem is first formulated into an output tracking problem. Next, with rigorous analysis, the ILC based freeway ramp metering is able to guarantee the asymptotic convergence of the traffic density to the desired one, despite the presence of velocity and input constraints. The control performance under constraints is further verified by intensive simulations.

Key words: urban traffic, iterative learning control, freeway ramp metering, input constraint, state constraint, convergence