Real-time Autonomous Cruise Control of Connected Plug-in Hybrid Electric Vehicles Under Uncertainty

Abstract

Advances in embedded digital computing and communication networks have enabled the development of automated driving systems. Autonomous cruise control (ACC) and cooperative ACC (CACC) systems are two popular types of these technologies, which can be implemented to enhance safety, traffic flow, driving comfort and energy economy. This PhD thesis develops robust and adaptive controllers for plug-in hybrid electric vehicles (PHEVs), with the Toyota Plug-in Prius as the baseline vehicle, in order to enable them to perform safe and robust car-following and platooning with improved vehicle performance.

Three controllers are designed here to achieve three main goals. The first goal of this thesis is the development of a real-time Ecological ACC (Eco-ACC) system for PHEVs, that is robust to uncertainties. A novel adaptive tube-based nonlinear model predictive control (AT-NMPC) approach to the design of Eco-ACC systems is proposed. Through utilizing two separate models to define the constrained optimal control problem, this method takes into account uncertainties, modeling errors and delayed data in the design of the controller and guaranties robust constraint handling for the assumed uncertainty bounds. {In addition, it adapts to changes in order to improve the control performance when possible.} Furthermore, a Newton/GMRES fast solver is employed to implement the designed AT-NMPC in real-time. The second goal is the development of a real-time Ecological CACC (Eco-CACC) system that can simultaneously satisfy the frequency-domain and time-domain platooning criteria. A novel distributed reference governor (RG) approach to the constraint handling of vehicle platoons equipped with CACC is presented. RG sits behind the controlled string stable system and keeps the output inside the defined constraints. Furthermore, to improve the platoon's energy economy, a controller is presented for the leader's control using NMPC method, assuming it is a PHEV. The third objective of this thesis is the control of heterogeneous platoons using an adaptive control approach. A direct model reference adaptive controller (MRAC) is designed that enforces a string stable behavior on the vehicle platoon despite different dynamical models of the platoon members and the external disturbances acting on the systems. The proposed method estimates the controller coefficients on-line to adapt to the disturbances such as wind, changing road grade and also to different vehicle dynamic behaviors. The main purpose of all three controllers is to maintain the driving safety of connected vehicles in car-following and platooning while being real-time implementable. In addition, when there is a possibility for performance enhancement without sacrificing safety, ecological improvement is also considered.

For each designed controller, Model-in-the-Loop (MIL) simulations and Hardware-in-the-Loop (HIL) experiments are performed using high-fidelity vehicle models in order to validate controllers' performance and ensure their real-time implementation capability.