Keywords

Cooperative Adaptive Cruise Control (CACC), Information Flow Topology, Long Platoon, Multi-Agent Systems (MASs), Multi-hop Broadcast, Piggybacking

Abstract

The use of Vehicle-to-Everything (V2X) communication in cooperative driving has the potential to greatly enhance transportation safety and efficiency. However, the success of Cooperative Vehicle Safety (CVS) applications is reliant on the reliability of the data system, which may experience data loss due to issues such as sensor failure or poor performance of V2X technologies in heavily trafficked areas. To compensate for non-ideal communication, cooperative vehicles must predict each other's behavior. The Model-Based Communication (MBC) concept was introduced to improve perception in cooperative vehicles under non-ideal communication by introducing a flexible content structure for broadcasting joint vehicle dynamics and driver behavior models.

Cooperative Adaptive Cruise Control (CACC) has been extensively studied in recent years as a way to alleviate traffic congestion and improve traffic flow, throughput, and highway capacity. This research aims to advance the coordinated operation of a large network of Connected and Automated Vehicles (CAVs) in an environment that includes both manned and automated vehicles, using the concept of perceptive stochastic coordination. The goal is to model vehicle movements and apply Stochastic Model Predictive Control (SMPC) to solve the mass platoon control problem. The performance of different information flow topologies, where vehicles receive information from multiple predecessors, was investigated in ideal and non-ideal communication setups. Additionally, the proposed CACC controller utilized a discrete hybrid stochastic MPC design that leveraged MBC. Simulation studies were conducted to evaluate the controller performance, and the results confirmed the effectiveness of the proposed approach.

Significant progress has been made in the development of Automated Vehicles (AVs), but their widespread adoption will not be possible until solutions are found that allow AVs to coexist with Human-driven Vehicles (HVs). Despite advances in self-driving technology, many drivers are expected to continue to prefer to drive themselves. Consequently, HVs and AVs will co-exist in mixed traffic for an extended period. For AVs to maneuver safely and efficiently in this mixed traffic, they must understand how humans perceive and react to risks while driving. However, the driving environment is continually changing and it is difficult to predict every possible scenario and hard-code controllers for each. To address these challenges, this research incorporates a human decision-making model into Reinforcement Learning (RL) to enable the safe and efficient operation of AVs.

Our research investigates V2X communication, emphasizing vehicle platoons as a transformative approach in transportation systems to enhance fuel efficiency and traffic flow. This method relies on maintaining closely monitored distances between vehicles, especially at high speeds, facilitated by rapid and reliable wireless data exchange. However, the inherent instability of wireless channels poses challenges such as data loss and delays, which can impair platoon functionality. Current models struggle to form long platoons due to the limited range of vehicle-to-vehicle (V2V) communication. Quick traffic information sharing is critical for vehicle response times, necessitating both reliability and ultra-low latency. To address these challenges, we propose a distance-based, network-aware relaying policy specifically designed for long platoons of connected vehicles. Our simulation results show that this relaying approach significantly reduces communication breakdowns and narrows the error gap between vehicles, achieved with minimal computational demand.

Completion Date

2024

Semester

Summer

Committee Chair

Fallah, Yaser

Degree

Doctor of Philosophy (Ph.D.)

College

College of Engineering and Computer Science

Department

ECE

Degree Program

Computer Engineering

Format

application/pdf

Identifier

DP0028876

URL

https://stars.library.ucf.edu/cgi/viewcontent.cgi?article=1398&context=etd2023

Language

English

Rights

In copyright

Release Date

2-15-2030

Length of Campus-only Access

5 years

Access Status

Doctoral Dissertation (Campus-only Access)

Campus Location

Orlando (Main) Campus

Accessibility Status

Meets minimum standards for ETDs/HUTs

Download

Restricted to the UCF community until 2-15-2030; it will then be open access.

Share

COinS