Observer-based internal model air–fuel ratio control of lean-burn SI engines. (30th September 2019)

• Record Type: Journal Article
• Authors: Wu, Hsiu-Ming; Tafreshi, Reza
• Is Part Of: IFAC journal of systems and control. Volume 9(2019)
• Publication Date: 2019-09-30
• Subjects: Air–fuel ratio (AFR) -- Lean-burn engines -- Observer-based internal model control; Automatic control -- Periodicals; Relay control systems -- Periodicals; Embedded computer systems -- Periodicals; Feedback control systems -- Periodicals; Artificial intelligence -- Periodicals; Artificial intelligence; Automatic control; Embedded computer systems; Feedback control systems; Relay control systems; Electronic journals; Periodicals; 629.89
• DOI: 10.1016/j.ifacsc.2019.100065
• Languages: English
• ISSNs: 2468-6018

Abstract

Abstract: Precise control of air–fuel ratio (AFR) in lean-burn spark ignition (SI) engines can reduce carbon dioxide emissions and other harmful pollutants and improve fuel economy. However, the large time-varying delay caused by inherent engine cycle and gas transport is a major challenge in the AFR performance control. Furthermore, AFR measurement via a universal exhaust gas oxygen (UEGO) sensor often has a biased error and measurement noise that significantly affects states needed for the feedback control. Herein, an observer-based internal model controller (IMC) is proposed to accurately track the desired AFR values in the presence of system parameter uncertainty and large time-varying delay. This is accomplished by implementing a control law via approximating the system delay using Pade first-order technique, including a state observer plus an integrator for estimating the system state and even improving the steady-state error in state feedback. The stability of the overall closed-loop system is validated using robust Nyquist theorem. The results show that the proposed observer-based IMC design based on Pade first-order approximation has better performance than based on Pade second-order approximation and is easily implemented in real-time. Moreover, the control scheme can accurately control the AFR system under different operating conditions to demonstrate acceptable performance in terms of robustness and fast response. Besides, the comparisons to a PI combined with a Smith predictor is conducted to show its superiority.