Search Results

Modern engine management systems increasingly rely on on-line identification schemes. These are used either for self-tuning regulators or the rapid parametrization of controllers. In this paper the on-line parameter identification of the wall-wetting dynamics is studied in detail. The identification is performed by exciting the fuel path dynamics of the engine at a constant operating point. The amount of fuel injected serves as input and the air-to-fuel ratio, which is measured with a linear oxygen sensor, as output. In order to gain precise information about the amount of fuel in the cylinder, a new measurement concept is used. For one, the placement of the lambda sensor close to the exhaust valve minimizes the effects of gas mixing on the measurements. Additionally, by an appropriate collection of the data, the sensor dynamics are bypassed. This is also illustrated by a measurement with a very fast NOx sensor.

In modern engine control applications, there is a distinct trend towards model-based control schemes. There are various reasons for this trend: Physical models allow deeper insights compared to heuristic functions, controllers can be designed faster and more accurately, and the possibility of obtaining an automated application scheme for the final engine to be controlled is a significant advantage. Another reason is that if physical effects can be separated, higher order models can be applied for different subsystems. This is in contrast to heuristic functions where the determination of the various maps poses large problems and is thus only feasible for low order models. One of the most important parts of an engine management system is the air-to-fuel control. The catalytic converter requires the mean air-to-fuel ratio to be very accurate in order to reach its optimal conversion rate. Disturbances from the active carbon filter and other additional devices have to be compensated.

This paper introduces a model-based adaptive controller designed to compensate mixture ratio dynamics in an SI engine. In the basic model the combined dynamics of wall-wetting and oxygen sensor have to be considered because the only information about process dynamics originates from measuring exhaust λ. The controller design is based on the principles of indirect Model Reference Adaptive Control (MRAC). The indirect approach connotes that explicit identification of the system parameters is required for the determination of the controller parameters. Due to nonlinearities and delays inherent in the process dynamics, an adaptive extended Kalman filter is used for identification purposes. The Kalman filter method has already been described in detail within an earlier paper [1]. It proves to be ideally suited to deal with nonlinear identification problems. The estimated parameters are further used to tune an adaptive observer for wall-wetting dynamics.

The two characteristics of wide-range air/fuel ratio sensors when located in front of and behind a three-way catalytic converter are investigated. Input as well as output gas concentration measurements and sensor readouts are presented. Behind a new converter almost no oxygen can be measured for rich air/fuel ratios. The wide-range sensor's signal is sensitive to changes in the gas composition when keeping the air/fuel ratio constant at a rich value. Since the gas compositions up- and down-stream of the converter differ, the sensor signals are not identical for the same rich air/fuel ratio before and after the converter. The various diffusion coefficients of the exhaust gas species flowing through the porous coating of the sensor combinded with the different up- and downstream gas compositions are responsible for the different sensor characteristics.

The dynamics of the fuel-path subsystem of an SI engine, between fuel injection command signal and measured air-to-fuel ratio, is modeled approximately by a series connection of a first-order low-pass filter and a time delay element. The three parameters involved in this approximation, i.e., the time constant and the gain factor of the low-pass filter as well as the time delay, depend on the operating point of the engine. In order to design a gain-scheduled controller for the entire operating range of the engine, the parameters are identified for a number of operating points. For the automation of the parameter identification of all operating points desired, an on-line identification based on the recursive least-squares method is used. The algorithm for the decision of whether to increase or decrease the integer part of the current estimated time delay, which is a multiple of the sampling period, is based on an estimation of the fractional part of the time delay at each point.