Distributed Model-Free Adaptive Predictive Control for Urban Traffic Networks

Data-driven control without using mathematical models is a promising research direction for urban traffic due to the massive amounts of data generated every day. This article proposes novel distributed model-free adaptive predictive (D-MFAPC) approach multiregion networks. More specifically, dynamics network regions are first transformed into MFAPC models, and then, derived instead serve as prediction in design. The formulated problem finally solved with an alternating method multipliers (ADMM)-based approach. simulation results Linfen, Shanxi, China, show feasibility effectiveness proposed method.