Constrained Full Form Dynamic Linearization based Model Free Adaptive Perimeter Control for Urban Traffic Networks

Most common perimeter control strategies based on the macroscopic fundamental diagram (MFD) generally depend on accurate mathematical models. However, due to the uncertainty and complexity of traffic flow models, these methods often struggle to achieve ideal results in practical applications. For a two-region traffic system, this paper proposes a data-driven strategy called constrained full form dynamic linearization based model free adaptive control (cFFDL-MFAC) to control the perimeter controllers. The advantage of this method is that it can dynamically adjust the control strategy through input and output (I/O) data, adapting to the time-varying features of the traffic system. And it comprehensively considers dynamic of the system, enabling a more accurate description of traffic flow variations and improving control performance by full form dynamic linearization technique. The performance of the cFFDL-MFAC method is demonstrated through numerical simulations, confirming its effectiveness and advantages compared to other frequently used perimeter control approaches.