A learning approach for freeway traffic control

In this work, several learning control algorithms are developed to regulate freeway density and flow under a macroscopic level freeway environment. A detailed analysis on the traffic model adopted in this work is first conducted. Next, to regulate the traffic density and flow, learning control method is used based on the repeatability of daily traffic patterns. The regulation is achieved either through ramp metering or speed control. Finally simulations are conducted to verify the efficacy of the proposed control algorithms.