A self-adaptive toll rate algorithm for high occupancy toll (HOT) lane operations.

DTRT07-G-0010Dramatically increasing travel demands and insufficient traffic facility supplies have resulted in severetraffic congestion. High Occupancy Toll (HOT) lane operations have been proposed as one of the mostapplicable and cost-effective countermeasures against freeway congestion. By allowing Single OccupancyVehicles (SOVs) to use High Occupancy Vehicle (HOV) lanes with a toll, excess capacities of HOV lanes canbe utilized. Through balancing pricing and vehicle occupancy constraints, an HOT lane system can optimizetraffic allocations between HOT and General Purpose (GP) lanes and hence enhance the overall infrastructureefficiency if a proper tolling strategy is employed. Although there exist several tolling strategies, two majorproblems with these tolling strategies may significantly degrade the HOT lane system performance. First, theunder-sensitive tolling algorithm is incapable of handling the hysteresis properties of traffic systems and maycause severe response delays. Secondly, unfavorable flow fluctuation on both HOT and GP lanes may resultfrom over-sensitive tolling strategies and generate agitating traffic operations.To address these problems, a new self-adaptive dynamic tolling algorithm is developed in this study tooptimize HOT lane operations. To reduce the computational complexity, a second-order control scheme isexploited in this algorithm. Based on traffic speed conditions and toll changing patterns, the optimum flowratio for HOT lane utilization is calculated using feedback control theory. Then the appropriate toll rate isestimated backward using the discrete route choice model. To examine the effectiveness of the proposedtolling algorithm, simulation experiments were conducted. A microscopic traffic simulation software tool,VISSIM, is utilized. The proposed algorithm is implemented and integrated with the VISSIM package throughan external module specifically developed for this study. Data from the Washington State Route (SR) 167HOT lane system is used to build and calibrate the simulation model. The simulation experiment results showthat the proposed tolling algorithm is capable of responding to traffic changes promptly and effectively. Itperformed reasonably well in optimizing overall traffic operations of the HOT lane system under varioustraffic conditions. This algorithm is logically straightforward and not difficult to implement