Development and evaluation of feedback-based freeway ramp metering strategies

One of the most widely used control measures for freeway control is ramp metering. In an effort to relieve recurrent and nonrecurrent congestion on freeways, various ramp metering strategies have been employed to regulate the inputs to freeways from on-ramps.; In this dissertation, an isolated "feedback-based" ramp metering strategy MIXCROS (Kachroo and Ozbay 2003) that explicitly considers ramp delays is implemented and tested. In addition to the regulation of ramp input from the freeway, the proposed strategy calls for regulation of ramp queues by explicitly incorporating them into the system dynamics used to derive the feedback control law. MIXCROS is tested using PARAMICS, a stochastic microscopic traffic simulation package, on a calibrated network located in Hayward, California. In addition to MIXCROS, 2 other ramp control strategies, namely ALINEA (Papageorgiou et al. 1991) and New Control (Kachroo and Ozbay 2003), are implemented using PARAMICS.; Implementations of the local controls have elicited a need for managing traffic on a freeway system more effectively, improving equity in overall network, and improving the traffic performance of a corridor during nonrecurring congestion (e.g., due to incidents, and environmental conditions). These improvements are addressed by "feedback-based" coordinated ramp metering strategies C-MIXCROS and D-MIXCROS (Kachroo and Ozbay 2003), which ascertain both efficiency and equity via explicit consideration of on-ramps in their objective function by assigning weights to both freeway and ramp. C-MIXCROS and D-MIXCROS are implemented and evaluated using both macroscopic (MATLAB) and microscopic (PARAMICS) simulation models (on an 11-mile-long corridor of I-295 in South Jersey) under different demand conditions. In addition to the newly proposed coordinated ramp metering strategies, a well-known coordinated strategy (METALINE (Papageorgiou et al., 1990)) and 3 other local strategies (ALINEA, New Control, and MIXCROS) are also implemented using the same network and results are compared. C-MIXCROS and D-MIXCROS both perform better than all other control strategies tested for all the demand scenarios both for the ramp system and at the network level.; The critical density of a freeway link is subject to changes over time owing to such circumstances as environmental conditions and traffic incidents. Because of the critical density's impacts on the performance of some ramp metering strategies that make use of it as a threshold value for control action, a strategy that automatically adapts to the real-time change of critical occupancy, based on the freeway traffic conditions at the bottleneck locations, can prove efficient. 4 methods for the online estimation of critical density using the extended Kalman filter and Kalman filter are proposed. The effects of the proposed methods on the performance of MIXCROS, C-MIXCROS and D-MIXCROS are evaluated using macroscopic and microscopic simulation models. The proposed methods show promise improving the overall system-wide performance of all ramp metering strategies on multiple ramps. In addition, all ramp metering strategies have superior results compared with the ones collected using constant critical density.