08-Cooperative Driving Automation (CDA) for Multi-intersection Traffic Control
Date and Time: Tuesday, July 30, 2024: 5:00 PM - 6:30 PM
Location: Indigo BC
Xiao Yun Lu
Senior Scientist, Lawrence Berkeley National Lab; PI Researcher, PATH, ITS, University of California Berkeley, Lawrence Berkeley National Lab; University of California ,Berkeley
Presentation Description
1. Relevance to Automated Road Transportation
Cooperative Driving Automation (CDA) provides a tangible means to achieve traffic level energy saving, emission reduction and safety and mobility improvement. It enables traffic cooperation via connectivity that includes vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), infrastructure to (I2V), Vehicle-to-Pedestrian (V2P), Pedestrian-to- Vehicle (P2V) and infrastructure-to-infrastructure (I2I). The connectivity will make situation-awareness feasible and reliable for all road users, thus creating significantly more benefits than just using vehicle on-board sensors. Besides, CDA can provide benefits before full connectivity is achieved. Some applications could be beneficial with low CDA market penetration levels of Connected Automated Vehicles (CAV) and/or V2X.
This proposal intends to present a typical CDA application project that integrates CAV trajectory control with multi-intersection arterial corridor signal optimization at low CDA market penetration levels. To evaluate the performance of the application, the project team developed a Hardware-in-the-loop test platform consisting of three Cooperative Adaptive Cruise Control (CACC) capable passenger cars with different powertrains, a two-lane physical test corridor with six signalized intersections equipped with the latest V2X Roadside Units (RSUs), and a real-time microscopic simulation that generates mixed traffic (with CAVs) streams on the test corridor. The system will be extensively tested at American Center for Mobility (ACM) and the data will be quantitatively analyzed to evaluate the performance.
2. Objective
The objective of this project is to design, develop and implement an arterial CDA application that integrates CAVs trajectory planning and intersection signal control and conduct quantitative hardware-in-the-loop tests to evaluate performances for mobility and energy consumption. The project also involves defining V2X messages required by the proposed CDA-based Active Traffic Management (ATM) of arterial multi-intersection traffic signals. The project is expected to demonstrate that: (a) CAVs can be used as mobile sensors to provide other connected vehicles (CVs), CAVs and the infrastructure devices more information for refined traffic state parameter estimation; and (b) CAVs can be used as the controller to regulate the traffic toward desired traffic flows through the CAVs movements.
3. Approach and Methodology
The on-going project intends to (1) define V2X message lists for the proposed CDA application and implement the messages with the most recently commercial OBUs and RSUs; (2) Implement CACC on three passenger cars with different powertrains; (3) Develop optimal signal coordination, signal control, trajectory planning for CAVs to achieve traffic level energy saving, emission reduction, and mobility improvement along a multi-intersection arterial corridor.
4. Status - Project Accomplishments
(1) We have initially developed message lists for V2V, V2I, U2V and I2I and implemented on Cohda V2X OBUs and RSUs;
(2) We have implemented CACC/platooning capability on three cars with different powertrains including: IC engine, hybrid-electric, and fully electric;
(3) We have developed Model Predictive Control (MPC) algorithms for multi-signals coordination and CAVs trajectory planning for energy savings, emission reduction, and mobility improvement in an arterial corridor [4, 5];
(4) We have developed real-time simulation algorithms in Aimsun to implement the control and coordination algorithms with the three physical CAVs integrated/imbedded;
(5) We have integrated the real-time simulation, the traffic control systems, and the trajectory planning on Berkeley side with the traffic control system on the test track in ACM in Michigan remotely through a well-designed communication protocol and user interface.
5. Key findings
We have initially identified message categories for V2V [1], V2I, I2V, and I2I and recognized the differences between freeways and arterial intersections applications. In the implementation of CACC/platooning on vehicles of different makes/models/powertrains [2, 3], it is very critical to recognize the differences in lower-level interface and control actuations. Those differences also cause internal delays and responses: EV has the fastest response, hybrid-electric is the second, and IC engine vehicle is the slowest. Recognizing those factors is important in vehicle positioning, control system design, implementation and tuning for the highest platooning performance in the real-word traffic.
Different powertrains of vehicles lead to differences in lower-level interface and control actuation, and differences in internal delays and responses. As vehicle manufacturer adding more and more Advanced Driver Assistance System (ADAS) such as Adaptive Cruise Control (ACC), more limits will exist. Examples are (a) limits on acceleration capability for a given speed; (b) acceptable command signals through the CAN (Control Area Network) Bus; (c) internal control logics for the system faulting out (disabled); and (d) accessibility of remote sensor (such as radar) raw or processed data. Those are important factors to take into consideration in vehicle automation.
We have tested the performance of the proposed corridor-level control policy in a microscopic simulation environment against a non-CDA baseline case. This is an essential step before executing the field test. The simulation results suggest that the corridor-level signal coordination can provide about 10% mobility and energy improvement when the control algorithms are enhanced with the CAV information. The CAV trajectory planning can generate an additional 4% benefit. These findings will be further validated and updated by the upcoming hardware-in-the-loop tests that involve the three physical CAVs.
6. References
[1]. X. Y. Lu, Shladover, A. Kailas, and O. Altan, Messages for Cooperative Adaptive Cruise Control Using V2V Communication in Real Traffic, Fei Wu Eds, Taylor & Francis Group, CRC Press, 2018
[2]. C. Flores, J. Spring, D. Nelson, S. Iliev, and X. Y. Lu, Enabling Cooperative Adaptive Cruise Control on Strings of Vehicles with Heterogeneous Dynamics and Powertrains, Vehicle System Dynamics 61(1), March 2022, DOI:10.1080/00423114.2022.2042568
[3]. C. Flores, J. Muñoz, C. A. Monje, V. Milanés, X. Y. Lu, Iso-damping Control for Robust Automated Car-Following Approaches, J. Adv Res. 2020 Sep; 25: 181–189. Published online 2020 Jun 17. doi: 10.1016/j.jare.2020.05.013
[4]. H. Liu, A. Kurzhanskiy, W. Hong, and X. Y. Lu, Integrating Vehicle Trajectory Planning and Arterial Traffic Management to Facilitate Eco-Approach and Departure Deployment, TRB 2024 Annual Meeting, Washington D. C. Jan 7-11 2024
[5]. H. Liu, Q. Wang, W. Hong, Y. Shao, and X. Y. Lu, Connected Automated Vehicles as Actuators of Active Traffic Management: State-of-the-Art Methodologies, TRB Annual Meeting Jan 8-12, 2023
Speaker Biography
Dr. Xiao-Yun Lu is a Principal Investigator (PI) Researcher at California PATH, U. C. Berkeley, and affiliated Senior Scientist at Lawrence Berkeley National Laboratory. He got BSc. in Mathematics from Sichuan University, China (1982), MSc. in Applied Mathematics from the Institute of Systems Science, Chinese Academy of Sciences (1985), and Ph. D. in Systems and Control from University of Manchester, UK (1994). He has 36 total years of experience in systems and control theory and design; traffic systems detection, modeling, micro/macro simulation, and real-time field implementation; connected automated vehicle (CAV) dynamics modeling, control design and implementation; and vehicle active safety. He has been a technical leader and project PI of several US DOE and DOT and Caltrans projects including: (1) algorithm and field test of coordinated ramp metering and variable speed advisory; (2) passenger car and truck platooning, and integration of CAVs with active traffic management for energy saving, emission reduction and mobility improvement; (3) coordinated merging algorithm development and implementation on 3 passenger cars; (4) practical string stability analysis for vehicle following; (5) integrated ACC (Adaptive Cruise Control) and CACC (Cooperative ACC) for heavy-duty truck platooning; (6) multi-vehicle longitudinal collision avoidance and impact mitigation with V2V; and (7) active control to prevent truck rollover. He is a member of TRB Committee on Vehicle and Highway Automation (ACP30), former member of TRB Committee on Highway Traffic Monitoring ABJ35, and ATM (Active Traffic Management) Subcommittee; reviewer of Mathematical Review.
Presentation File
Cooperative Driving Automation (CDA) for Multi-intersection Traffic Control
Category
Energy and Environmental Implications of Transportation Automation