Use this URL to cite or link to this record in EThOS:
Title: A model-based method for optimising emissions of diesel engines through non-linear model predictive control
Author: Wang, Xiaoming
ISNI:       0000 0004 7969 6759
Awarding Body: University of Gloucestershire
Current Institution: University of Gloucestershire
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Thesis embargoed until 03 Jun 2023
Access from Institution:
For diesel engines, based on the legislative emission limits set by Euro I to Euro VI, the reductions required in particulate matter and nitrogen oxide are 80% and 90%, respectively. Meanwhile, fuel-consumption efficiency is still an important consideration for customers due to ever-increasing fuel prices. Modern diesel engines employ advanced fuel-injection systems which can efficiently reduce emissions and fuel consumption, as they have good fuel distribution in the combustion chamber and produce a close-to-homogeneous charger-compression ignition. However, ideal combustion conditions can be achieved only in combination with optimal control of the air-path system of the engine. Therefore, the aim of this study is to research, design and develop a new algorithm for the nonlinear, model-predictive control of air-path systems of diesel engines. In this study, which is conducted on the basis of measurements taken from a virtual test-bench under near-real load conditions, a linear parameter-varying model is created and parameterised by dynamical system identification. The results of simulation show that a linear parameter-varying modelling approach can be used to represent this air-path system more precisely than is possible with other, more conventional methods. The data-based modelling approach through an engine-simulation platform and the model-based optimisation framework developed in this study are used to design an innovative, non-linear, model-predictive controller for a diesel-engine air-path. The idea behind the proposed non-linear model-predictive control strategy is to represent the plant model as a linear parameter-varying model, and the control-objective function in searching for an optimal solution to the quadratic programming problem is extended to the parameter-varying cost function by utilising the given linear parameter-varying model. This concept is aimed at optimising the efficiency of engine-air-path systems with respect to intake-manifold pressure and air-mass flow tracking in transient operations. The problems of a prediction-model mismatch and the cross-coupling effects of two actuators are overcome by the application of a multiple-input, multiple-output linear parameter-varying model. The results reveal that, compared to existing approaches, the proposed non-linear model-predictive control method significantly improves the accuracy and computational efficiency of engine-air-path system control-even in large, transient operations. Finally, significant potential exists to improve the performance of the control. Thus, emissions and fuel consumption in the certification driving cycle of the vehicle can be optimised on the basis of the model.
Supervisor: Zhang, Shujun ; Bechkoum, Kamal Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: TK Electrical engineering. Electronics Nuclear engineering ; TL1.484 Motor vehicles. Cycles