Model Predictive Control-based Adaptive Traffic Signaling Scheme for Philippine Urban Traffic

Traffic congestion is a recurring issue on Philippine urban roads resulting in increased travel times. One of the contributors to this is the inefficient fixed-time signaling in local stoplights. This work explored a Model Predictive Control (MPC)-based adaptive traffic system to solve the congestion issue. Using the SUMO simulator, the optimal green times of stoplights were computed and applied in a simulation of traffic in the Katipunan Avenue - Aurora Boulevard intersection in Metro Manila. The MPC-based scheme performed consistently better than the existing fixed-time signaling, reducing average queue time and queue length by 57.43 % and 41.85 %, respectively, for the same flow rate. The adaptive traffic scheme was also robust, with negligible degradation in performance when subjected to 20 % vehicle count errors.