Comprehensive Multi-Layered and Time-Dependent Congestion Control Strategy for Mixed Traffic Flow

This paper proposed a comprehensive multilayered and time-dependent congestion control strategy for the coexistence of connected autonomous vehicles (CAVs) and human-driven vehicles in future mixed traffic systems. A linear programming model was developed with the objective of minimizing the average travel time. A genetic algorithm was used to optimize the time taken by CAVs to enter the main road via the entrance ramp, resulting in an optimal control strategy that satisfies the objective function and constraints. Typical scenarios with three different origin-destination (OD) distribution types were designed and the effectiveness of the control strategy was demonstrated through primary simulation analysis. In an approximately uniform distribution of OD scenario, 23 CAVs (representing 2.71% of all vehicles) were controlled but were able to reduce the average travel time of the entire mixed traffic system by almost 55%. Significantly improving the efficiency of the traffic system and alleviating the prevalent spatiotemporal congestion problem. Furthermore, in two scenarios with a ununiform OD distribution, by controlling CAVs that account for no more than 5% of all vehicles, it was possible to exceed the average travel time by 70% and 25% respectively. Overall, this study confirmed the effectiveness of the proposed congestion control strategy, which achieved significant control results at a relatively low cost. It showed great promise for further testing and application in traffic management.