A Simulation Study on Dynamic Speed Harmonization (DSH) Found That Implementing DSH With a 10 Market Penetration Rate of Connected and Automated Vehicles (CAVs) Led to a 9.1 to 9.6 Percentage Point Reduction in Collision Probability.
Northern California Simulation Study Explored the Impacts of Deploying CAVs with DSH.
California
Dynamic Coordinated Speed Control and Synergistic Performance Evaluation in Connected and Automated Vehicle Environment
Summary Information
Dynamic Speed Harmonization (DSH), also known as variable speed limit, is a common traffic control strategy to reduce speed oscillation at freeway merge sections and before work zones or incidents. The objective of this study was to investigate the effects of DSH in mixed traffic flow involving Human-Driven Vehicles (HDVs) and Connected and Automated Vehicles (CAVs) on freeways. A safety-oriented DSH strategy based on Deep Reinforcement Learning (DRL) was developed to understand how CAVs could improve operational performance under different CAV Market Penetration Rates (MPRs) in multiple simulated scenarios of recurrent and non-recurrent bottleneck conditions. Safety and mobility benefits of the proposed approach were measured.
METHODOLOGY
The study used the simulation model of the 1.4-mile-long freeway segment of I-80 in District 4 of California as a test bed and used Caltrans Performance Measurement System (PeMS) data. The selected freeway segment included a 2,100- foot-long weaving section. The simulation analyses included two cases: (1) recurrent bottleneck conditions and (2) non-recurrent bottleneck conditions. Both compared the effectiveness of DSH to a no-DSH scenario.
- For the recurrent case, the simulation model was run with five different MPR levels using various measures of effectiveness (MOEs), such as accumulated emergency deceleration, mean travel time, and fuel consumption.
- For the non-recurrent case, the simulation runs included an incident at the downstream weaving area, and the model was run with three different MPR levels with and without a special event that inflated the on-ramp traffic demand by 50 percent. MOEs such as Collision probability, Deceleration Rate to Avoid the Crash (DRAC), average travel time, and average time loss were assessed.
FINDINGS
- For the non-recurrent bottleneck scenario, all MOE indicators decreased with an increase in the MPR of CAVs. With 10 percent MPR and no special event, the DHS approach reduced collision probability by 9.1 percentage points (30.9 percent to 21.8 percent) compared to the no-DHS approach. Similar benefits were observed at 50 percent and 90 percent MPR, but to a lesser extent.
- With a special event and 10 percent MPR, collision probability was reduced by 9.6 percentage points, from 31.8 percent to 22.2 percent. Similar benefits were observed at 50 percent and 90 percent MPR, but to a lesser extent.
