Search Results

This paper describes algorithms that can recognize and track the engine crankshaft position for arbitrary cam- and crank-shaft tooth wheel patterns in both steady-state and transient operating conditions. Crankshaft position tracking resolution is adjustable to accommodate different application requirements. The instantaneous crankshaft position information provided by the position tracking module form the basis for crankshaft angle domain (CAD) engine control and measurement functions such as precise injection / ignition controls and on-line cylinder pressure CAD analyses. The algorithms described make reconfiguration of the tracking module for different and arbitrary cam- and crank-shaft tooth wheel patterns very easy, which is valuable especially for prototyping engine control systems. The effectiveness of the algorithms is shown using test engines with different cam and crank signal patterns.

In-cylinder ion sensing through sparkplug electrodes can be used to determine in-cylinder A/F ratio by using a modified production coil-on-plug ignition system having ion sensing capability. The in-cylinder ionization can be characterized by the height of the peak, location of the peak from ignition command and area under the ionization signal curve. The effects of A/F ratio on the in-cylinder ionization can be isolated from other affecting factors by conducting tests on a constant volume combustion device in which the initial pressure and temperature can be well controlled. This results in a parabolic correlation of the ionization characteristics with the mixture equivalence ratio. Additionally the ionization characteristics show strong dependence on engine load and speed. Equivalence ratio characteristics during engine cranking and warm up are investigated, and a method for on-line calibration of ionization detection is discussed.

This paper explores the possibility of on-board fuel classification for fuel property adaptive compression-ignition engine control system. The fuel classifier is designed to on-board classify the fuel that a diesel engine is running, including alternative and renewable fuels such as bio-diesel. Based on this classification, the key fuel properties are provided to the engine control system for optimal control of in-cylinder combustion and exhaust treatment system management with respect to the fuel. The fuel classifier employs engine input-output response characteristics measured from standard engine sensors to classify the fuel. For proof-of-concept purposes, engine input-output responses were measured for three different fuels at three different engine operating conditions. Two neural-network-based fuel classifiers were developed for different classification scenarios. Of the three engine operating conditions tested, two conditions were selected for the fuel classifier to be active.

Tremendous amount of useful information can be extracted from the cylinder pressure signal for engine combustion control. However, the physical cylinder pressure sensors are undesirably expensive and their health need to be monitored for fault diagnostic purpose as well. This paper presents the results of the development of a virtual cylinder pressure sensor (VCPS) with individual variable-oriented independent estimators. Two neural network-based independent cylinder pressure related variable estimators were developed and verified at steady state. The results show that these models can predict the variables correctly compared with the extracted variables from the measured physical cylinder pressure sensor signal. Good generalization capabilities of the developed models are observed in the sense that the models work well not only for the training data set but also for the new inputs that they have never been exposed to before.