Performance enhancement of internal combustion engines using crank -angle domain control

Andrew W Osburn, Purdue University

Abstract

The crank-angle domain is examined as a framework for engine controller design. The crank-angle domain refers to replacing time with engine crankshaft position as the independent variable for modeling and signal processing. Three properties of the crank-angle domain lead to improved performance of crank-angle domain controls. First, it will be shown that engine dynamics are generally less varying in the crank-angle domain. This property can reduce design effort or widen the operating range of a controller. Second, the periodicity of signals that are correlated to the engine cycle is demonstrated to be one engine cycle. This property is useful in designing individual cylinder controls and misfire diagnostics. Third, signal processing in the crank-angle domain allows synchronization of algorithms with the inherent sampling processes in the engine and avoids the creation of a multi-rate system. It is shown that this can improve the results of signal processing tasks. The crank-angle domain translates signal processing problems into a spatial domain. Crankshaft displacement is a monotonically increasing function of time. As long as the engine is running, there is no conflict in representing functions of time as functions of rotational displacement. Therefore, any signal processing method which can be performed on a time domain function can be performed on a crank-angle domain function. Furthermore, all available tools for signal representation and transformation, system identification, controller design, and algorithm implementation are directly applicable in a crank-angle domain framework; no new theory is required. This dissertation focuses on demonstrating how the crank-angle domain can be used to design solutions for engine controller functions. Each of the major control tasks for spark-ignition engines is addressed and new perspectives on these problems are presented.

Degree

Ph.D.

Advisors

Chiu, Purdue University.

Subject Area

Automotive materials

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS