Source-linked AI summary
A Review on Cooperative Adaptive Cruise Control (CACC) Systems: Architectures, Controls, and Applications
Ziran Wang, Guoyuan Wu, Matthew Barth
TL;DR
Transportation systems face safety, mobility, and sustainability challenges, motivating cooperative vehicle technologies such as CACC. This paper reviews CACC architectures, control methodologies, applications, and related issues, reporting benefits including reduced intersection delay and energy consumption while identifying testing and implementation-cost challenges.
Problem
CACC is studied as a promising cooperative technology for addressing safety, mobility, and sustainability issues in transportation systems.
Method
The paper reviews CACC system architectures, control methodologies, applications, and their related issues.
Results
Average intersection delay was reduced by 90% and energy consumption by 45% in an optimal-control-based intersection CACC simulation.
Takeaways & Limitations
CACC applications can provide safety, mobility, energy, and environmental benefits through cooperative vehicle operation and control.
Takeaways & Limitations
CACC applications rely on assumptions and may face substantial costs for infrastructure, testing, policies, and vehicle adoption.
Abstract
from arXiv · showhide
Connected and automated vehicles (CAVs) have the potential to address the safety, mobility and sustainability issues of our current transportation systems. Cooperative adaptive cruise control (CACC), for example, is one promising technology to allow CAVs to be driven in a cooperative manner and introduces system-wide benefits. In this paper, we review the progress achieved by researchers worldwide regarding different aspects of CACC systems. Literature of CACC system architectures are reviewed, which explain how this system works from a higher level. Different control methodologies and their related issues are reviewed to introduce CACC systems from a lower level. Applications of CACC technology are demonstrated with detailed literature, which draw an overall landscape of CACC, point out current opportunities and challenges, and anticipate its development in the near future.
I. INTRODUCTION
CACC integrates connectivity and automation so vehicles can cooperate through V2V communication, forming platoons with shorter headways. The reviewed architecture spans vehicle subsystems and multiple information-flow topologies, while applications target safety, capacity, energy use, and emissions.
- CAVs combine onboard sensing and autonomous driving with vehicle-to-vehicle communication, enabling cooperative vehicle operation.
- CACC extends adaptive cruise control with cooperative maneuvers performed by connected and automated vehicles.
- CACC vehicles share acceleration, speed, position, and related parameters through distributed V2V communication to coordinate platoon driving.
- Shorter time or distance headways can increase roadway capacity, while reduced velocity changes and aerodynamic drag can lower energy consumption and pollutant emissions.
- A CACC-enabled CAV architecture includes perception, planning, and actuation phases, with sensor and V2V data carried through the vehicle’s CAN bus.
- Information-flow topologies include predecessor-following, predecessor-leader following, two-predecessor variants, and bidirectional communication.
III. CONTROLS
Longitudinal control is central to CACC because vehicles must coordinate speed and maintain longitudinal spacing or headway. Controllers have been studied for objectives including platoon formation and fuel-consumption optimization.
- CACC longitudinal controllers coordinate vehicle speed while maintaining a constant inter-vehicle distance or headway relative to the preceding vehicle.
A. Model Predictive Control
MPC uses an explicit process model to predict plant responses and optimize control over prediction and control horizons. In CACC, MPC has been adapted for fuel savings and distributed coordination, partly addressing the limitations of centralized computation and global-state requirements.
- A. Model Predictive Control: Model predictive control uses an explicit process model to predict a plant’s future response.
- A. Model Predictive Control: Traditional MPC represents a single-agent system with a linear discrete-time state-space model and computes control inputs from system states.
- A. Model Predictive Control: A receding-horizon MPC formulation uses a prediction horizon of length p and a control horizon of length m.
- A. Model Predictive Control: A stochastic linear MPC-based CACC strategy produced 11~15% fuel savings in simulation.
- A. Model Predictive Control: Centralized MPC can be impractical when gathering all agents’ information and solving large-scale optimization problems is difficult.
- A. Model Predictive Control: Distributed MPC lets local controllers solve MPC problems using local information while sharing information to improve overall performance.
B. Consensus Control
Consensus control makes networked agents move toward agreement using local interactions and can be extended from single-integrator to double-integrator models for CAV dynamics. Its linear design has limitations with nonlinearities, constraints, time delays, and explicit string-stability requirements.
- B. Consensus Control: Distributed consensus uses local interactions rather than centralized global knowledge to make network agents reach agreement.
- B. Consensus Control: A single-integrator consensus algorithm drives each agent’s information state toward the information states of its communication neighbors.
- B. Consensus Control: Double-integrator consensus models CAV position, velocity, and acceleration, reaching consensus when both position and velocity differences converge.
- B. Consensus Control: Consensus algorithms can incorporate time-variant or time-invariant communication delays into the state-update model.
- B. Consensus Control: Linear consensus methods cannot directly handle CACC nonlinearities and constraints, and string stability requires additional topology or spacing-policy conditions.
C. Optimal Control
Optimal control formulations for CACC commonly optimize energy consumption or travel time while incorporating vehicle nonlinearities and constraints. Applications include Eco-CACC, intersection coordination, truck coordination, and platoon-wide energy and emissions optimization.
- Optimal CACC control can be formulated as a structured convex optimization problem minimizing energy consumption or travel time.
- Total energy consumed while traversing a designated area is often used as the objective function.
- 90% lower average intersection delay and 45% lower energy consumption were reported for an optimal-control intersection CACC system in simulation.
- Other studies optimize truck coordination fuel savings, platoon-wide energy consumption, pollutant emissions, or intra-platoon vehicle sequence.
D. String Stability
String stability is required for safe CACC platoons because disturbances should not amplify upstream. It is distinct from internal Lyapunov stability and depends on information flow, spacing policy, and actuator delay.
- String stability requires distance error, velocity, or acceleration disturbances to attenuate along the upstream direction of a platoon.
- The following vehicle’s scalar output and preceding vehicle’s scalar output are used to formulate string-stability conditions.
- Internal Lyapunov stability does not necessarily ensure string stability, and unattenuated errors can eventually cause collisions between consecutive vehicles.
- Actuator delay significantly limits the minimum inter-vehicle distance required by string-stability requirements.
- With constant-distance spacing, predecessor-following alone cannot guarantee string stability, whereas broadcasting leader information can extend information flow and ensure it.
IV. APPLICATIONS
CACC applications target safety, mobility, and sustainability across different traffic networks. Researchers have also conducted field implementations to test their effectiveness.
- CACC applications have been proposed and developed across different traffic networks to address safety, mobility, and sustainability.
- Field implementations have been conducted to test the effectiveness of CACC applications.
- The paper presents detailed application examples alongside the broader development of CACC technology.
A. Vehicle Platooning
Vehicle platooning uses CACC to tightly coordinate vehicles through distributed cooperation. The literature addresses capacity, energy, control nonlinearities, communication imperfections, and limited real-world validation.
- A. Vehicle Platooning: CACC platooning uses shorter inter-vehicle distances to increase roadway capacity and reduce energy consumption through lower aerodynamic drag and fewer unnecessary speed changes.
- A. Vehicle Platooning: Distributed consensus algorithms have been widely used to control inter-vehicle distance and achieve weighted or constrained platoon consensus.
- A. Vehicle Platooning: MPC provides a substitute for consensus when platooning controllers must handle nonlinearities through multiple local convex problems over a predictive horizon.
- A. Vehicle Platooning: Cooperative distributed platooning strategies aim to coordinate vehicles while addressing communication delay, limited range, packet loss, and sampling intervals.
- A. Vehicle Platooning: Relatively few proposed platooning control strategies have been experimentally tested in real-world traffic situations.
B. Eco-Driving on Signalized Corridors
CACC applications at signalized intersections and arterials use trajectory optimization, game-theoretic coordination, and cluster-wise organization to reduce energy use and coordinate vehicle movement.
- Eco-CACC at isolated intersections: Eco-CACC estimates intersection queues and optimizes vehicle trajectories so vehicles arrive as the last queued vehicle departs.The approach targets energy consumption at isolated signalized intersections.
- Eco-CACC at isolated intersections: 40% energy savings were reported for Eco-CACC at a 100% CAV market penetration rate in microscopic traffic simulation.
- Intersection coordination: A heuristic game theory-based iCACC system was compared with a four-way stop baseline and later extended with controller-to-CAV advice on optimal actions.The extension targeted intersection delay and energy consumption.
- Intersection coordination: 75% reduction in energy use was reported in a simulation study of integrated CACC and intelligent traffic signal control.The study analyzed mobility, energy, and environmental impacts.
- Cluster-wise coordination: A cluster-wise system groups approaching CAVs into deterministic-sequence clusters, each containing lane-level CACC platoons coordinated through their platoon leaders.Cluster followers can follow the leader’s eco-driving maneuver with respect to traffic signal information.
C. Cooperative Merging at Highway On-Ramps
CACC cooperative merging methods address stop-and-go effects on highway on-ramps through virtual vehicles, arrival-based sequencing, fuzzy control, and distributed vehicle assignment.
- Merging motivation: CACC is used to let CAVs merge cooperatively on highway on-ramps, where ramp metering can otherwise create stop-and-go traffic, extra energy consumption, and time waste.
- Virtual-vehicle methods: A virtual or “ghost” vehicle is mapped onto the main road before merging to support safer and smoother merging maneuvers.
- Virtual-vehicle methods: Vehicles receive sequence IDs based on arrival time at the merging point and cooperate with neighboring real or virtual vehicles through V2V communications.
- Control methodologies: A fuzzy-logic controller regulates longitudinal motion while an on-ramp vehicle cooperatively merges with CACC vehicles on the main road.
- Control methodologies: A distributed control protocol assigns vehicles to CACC systems in the merging scenario.
V. DISCUSSIONS
The review identifies unresolved CACC challenges involving architectures for dynamic environments, validation of market-ready controllers, and implementation costs under realistic assumptions.
- Open questions: The review covers CACC system architectures, control methodologies, and applications, while identifying open questions for future work.
- Architecture: Reliable CACC architectures must address changing information-flow topologies, varying workload distribution, and V2V packet loss in dynamic traffic networks.
- Control methodology: Market-ready CACC controllers require testing across diverse conditions and environments over relatively long mileage.Testing is difficult because CACC systems often involve several CAVs.
- Implementation: High implementation costs may impede CACC applications because proposed systems rely on assumptions and may require new policies, roadside infrastructure, and testing.