Connected Vehicle Data-Driven Fixed-Time Traffic Signal Control Considering Cyclic Time-Dependent Vehicle Arrivals Based on Cumulative Flow Diagram
Chaopeng Tan,Yumin Cao,Xuegang Ban,Keshuang Tang
DOI: https://doi.org/10.1109/tits.2024.3360090
IF: 8.5
2024-01-01
IEEE Transactions on Intelligent Transportation Systems
Abstract:Fixed-time control is a widely adopted and cost-effective method for signalized intersections. However, existing studies utilizing connected vehicle (CV) data have not effectively addressed fixed-time control due to their reliance on specific vehicle arrival assumptions. To overcome this limitation, this study presents a novel traffic control approach for fixed-time signalized intersections based on a cumulative flow diagram (CFD) framework. The proposed method comprises a CFD model and a multi-objective optimization model. The CFD model establishes analytical relationships between traffic flow operations and varying signal timing parameters, with intersection demand estimated using a novel weighted maximum likelihood estimation method. A multi-objective optimization model based on CFD is formulated to minimize exceeded queue dissipation time as the primary objective and average delay as the secondary objective, which is applicable under both undersaturated and oversaturated traffic conditions. Leveraging the data-driven nature of the CFD model, a specially designed bi-level particle swarm optimization-based algorithm is employed to determine optimal cycle length (and offset if applicable) and green ratios separately. Evaluation results demonstrate that the proposed method outperforms Synchro, a conventional approach, in terms of average delay and queue under various traffic conditions. Moreover, the proposed method exhibits the capability to handle specialized scenarios involving spillbacks.
engineering, electrical & electronic,transportation science & technology, civil