Constrained multivariable extremum-seeking for online fuel-efficiency optimization of Diesel engines. (June 2019)

• Record Type: Journal Article
• Authors: van der Weijst, Robert; van Keulen, Thijs; Willems, Frank
• Is Part Of: Control engineering practice. Volume 87(2019)
• Page Start: 133
• Page End: 144
• Publication Date: 2019-06
• Subjects: Extremum-seeking -- Constraint handling -- Multivariable control -- Real-time optimization -- Diesel engine control -- Fuel-efficiency optimization -- Real-world driving emissions; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2019.03.008
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: This work presents a new method for online fuel-efficiency optimization of Diesel engines, using constrained extremum-seeking. A two-input optimization problem, which is suitable for extremum-seeking, is integrated into a tracking control system. As a result, both air-path and fuel-path actuators are used for tracking and extremum-seeking. A key element of the proposed method is a cost function based on real-time BSFC estimation. Moreover, an existing constrained extremum-seeking method is extended to multiple output constraints. Experiments on a Euro-VI heavy-duty Diesel engine demonstrate the constraint handling, robustness with respect to real-world disturbances, and the fuel saving potential of the control design.