Date of Award

2015

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy in Mechanical Engineering-Engineering Mechanics (PhD)

Administrative Home Department

Department of Mechanical Engineering-Engineering Mechanics

Advisor 1

Mahdi Shahbakhti

Committee Member 1

Jeffrey D. Naber

Committee Member 2

Rush D. Robinett III

Committee Member 3

John E. Beard

Committee Member 4

Jeffrey B. Burl

Abstract

Low Temperature Combustion (LTC) holds promise for high thermal efficiency and low Nitrogen Oxides (NOx) and Particulate Matter (PM) exhaust emissions. Fast and robust control of different engine variables is a major challenge for real-time model-based control of LTC. This thesis concentrates on control of powertrain systems that are integrated with a specific type of LTC engines called Homogenous Charge Compression Ignition (HCCI). In this thesis, accurate mean value and dynamic cycleto- cycle Control Oriented Models (COMs) are developed to capture the dynamics of HCCI engine operation. The COMs are experimentally validated for a wide range of HCCI steady-state and transient operating conditions. The developed COMs can predict engine variables including combustion phasing, engine load and exhaust gas temperature with low computational requirements for multi-input multi-output realtime HCCI controller design. Different types of model-based controllers are then developed and implemented on a detailed experimentally validated physical HCCI engine model. Control of engine output and tailpipe emissions are conducted using two methodologies: i) an optimal algorithm based on a novel engine performance index to minimize engine-out emissions and exhaust aftertreatment efficiency, and ii) grey-box modeling technique in combination with optimization methods to minimize engine emissions. In addition, grey-box models are experimentally validated and their prediction accuracy is compared with that from black-box only or clear-box only models. A detailed powertrain model is developed for a parallel Hybrid Electric Vehicle (HEV) integrated with an HCCI engine. The HEV model includes sub-models for different HEV components including Electric-machine (E-machine), battery, transmission system, and Longitudinal Vehicle Dynamics (LVD). The HCCI map model is obtained based on extensive experimental engine dynamometer testing. The LTC-HEV model is used to investigate the potential fuel consumption benefits archived by combining two technologies including LTC and electrification. An optimal control strategy including Model Predictive Control (MPC) is used for energy management control in the studied parallel LTC-HEV. The developed HEV model is then modified by replacing a detailed dynamic engine model and a dynamic clutch model to investigate effects of powertrain dynamics on the HEV energy consumption. The dynamics include engine fuel flow dynamics, engine air flow dynamics, engine rotational dynamics, and clutch dynamics. An enhanced MPC strategy for HEV torque split control is developed by incorporating the effects of the studied engine dynamics to save more energy compared to the commonly used map-based control strategies where the effects of powertrain dynamics are ignored.

LTC is promising for reduction in fuel consumption and emission production however sophisticated multi variable engine controllers are required to realize application of LTC engines. This thesis centers on development of model-based controllers for powertrain systems with LTC engines.

Share

COinS