A Simulation Tool and a Driving Simulator Experiment were Implemented to Capture the Traffic Performance when Optimizing Vehicle Speed.
Blacksburg, Virginia, United States
Integrated Optimization of Vehicle Trajectories and Traffic Signal Timings: System Development and Testing
Summary Information
Connected automated vehicles (CAVs) have opened up new possibilities for traffic signal control and vehicle management at signalized intersections, offering potential improvements in mobility, fuel economy, and traffic safety. Specifically, vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, as well as vehicle automation, provide opportunities to optimize traffic signal control and vehicle control. The objective of this study was to simulate a two-layer optimization approach for providing optimal control for both vehicles and traffic signal controllers. The first layer conducted an optimization strategy on traffic signal timings in real-time for approaching vehicles from upstream traffic, minimizing energy consumption. The secondary layer calculated the optimal number of vehicle brakes and throttle levels using a vehicle speed controller, with the goal of achieving the most efficient and safe vehicle trajectories in real-time using a dynamic programming method in simulation software. To test the efficacy of the proposed speed guidance system, a driving simulator experiment was conducted with 70 participants with a user survey.
METHODOLOGY
This proposed controller was first tested at an isolated signalized intersection to validate its performance and then on an arterial road in downtown Blacksburg with multiple traffic signals using a microscopic traffic simulation software. Five scenarios were considered in the simulation tests. The origin-destination (O-D) demand matrices were generated based on traffic counts collected during the afternoon peak period (4 ~ 6 pm) at 15 minutes intervals for the year 2012. and three O-D demand levels were considered: 25 percent, 50 percent and 100 percent of the calibrated demand levels. These were:
- Scenario 1 (S1): Base Case without signal optimization and vehicle speed control.
- Scenario 2 (S2): Signal Optimization using Webster’s method.
- Scenario 3 (S3): Signal Optimization to minimize traffic delay.
- Scenario 4 (S4): Signal Optimization to minimize fuel consumption.
- Scenario 5 (S5): Integrated Controller that combines signal optimization for minimizing fuel consumption and vehicle speed optimization.
In addition, a driving simulator experiment was conducted by recruiting 70 participants from Morgan State University and the Baltimore metro area. The participants were instructed to drive in the virtual study area based on a medium-size road network in the Baltimore metropolitan area consisting of three signalized intersections, under 17 different scenarios (uphill and downhill). A speed guidance system in the simulator was used called the Eco- CACC-I algorithm. The system computed the recommended speed to assist drivers, avoid stop-and-go behavior, and lower greenhouse gas emissions.
FINDINGS
- The test results under 25 percent demand level indicated that the integrated controller (S5) could significantly enhance traffic mobility with a 31.44 percent reduction of total delay and a 24.16 percent reduction of vehicle stops, at the same time improving the energy efficiency with a 13.98 percent reduction in fuel consumption. Similar trends were observed under 50 percent and 100 percent demand levels.
- Overall, the test results on the arterial network indicated that the proposed controller can greatly improve energy efficiency with 17.7 percent fuel savings and enhance traffic mobility with up to a 47.18 percent reduction in total delay and 24.84 percent reduction in vehicle stops.
- The simulator testing the impacts of speed guidance systems resulted in a reduction of 20 percent in emissions in the uphill scenarios and seven percent for downhill scenarios.
- In the post survey results, 60 percent of the participants preferred using the Speed-Change guidance.
- The percentage of usage of the recommended speed voice scenario for the uphill scenario was 53, 61, and 57 percent for the first, second, and third intersections respectively.
- The percentage of usage of the recommended speed voice scenario for the downhill scenario was 61, 63, and 66 percent for the first, second, and third intersections respectively.
