Internal-combustion engines – Combustion chamber means having fuel injection only – Using multiple injectors or injections
Reexamination Certificate
2003-04-17
2004-11-30
Kwon, John T. (Department: 3747)
Internal-combustion engines
Combustion chamber means having fuel injection only
Using multiple injectors or injections
C123S381000, C123S446000, C123S480000
Reexamination Certificate
active
06823834
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to fuel injection systems for internal combustion engines, and more specifically to techniques for estimating pilot and/or post-injected fuel quantities and minimizing variations between such fuel quantities.
BACKGROUND OF THE INVENTION
In recent years, advances in fuel systems for internal combustion engines, and particularly for diesel engines, have increased dramatically. However, in order to achieve optimal engine performance at all operating conditions with respect to fuel economy, exhaust emissions, noise, transient response, and the like, further advances are necessary. As one example, operational accuracy with electronically controlled fuel systems can be improved by reducing variations in injected fuel quantities.
A number of techniques are known for reducing injected fuel quantity variations such as, for example, robust system design, precision manufacturing, precise component matching, and electronic control strategies. However, conventional manufacturing approaches for improving performance, such as tightening tolerances and the like, are typically cost prohibitive, and conventional control approaches such as open-loop look-up tables have become increasingly complex and difficult to implement as the number of degrees of freedom to control have increased, particularly with multiple-input, multiple-output (MIMO) control systems. In fact, both of these approaches improve accuracy only during engine operation immediately after calibration in a controlled environment, and neither compensate for deterioration or environmental noise changes, which affect subsequent performance. Closed-loop control systems for controlling injected fuel quantity variations are accordingly preferable, but typically require additional sensors to measure appropriate control parameters.
One known technique for implementing such a closed-loop control system without implementing additional sensors is to leverage existing information to estimate injected fuel quantity; i.e., implementation of a so-called “virtual sensor.” One example of a known control system
10
including such a virtual sensor is illustrated in FIG.
3
. Referring to
FIG. 3
, system
10
includes a two-dimensional look-up table
14
receiving an engine speed/position signal via signal line
12
and a desired fuel injection quantity value from process block
16
via signal path
18
. Table
14
is responsive to the engine speed/position signal and the desired fuel injection quantity value to produce an initial fueling command as is known in the art. The virtual injected fuel quantity sensor in system
10
typically comprises a two-dimensional look-up table
20
receiving the engine speed/position signal via signal path
12
and a fuel pressure signal from signal path
22
. Table
20
is responsive to the fuel pressure and engine speed/position signals to produce an injected fuel quantity estimate that is applied to summing node
24
. Node
24
produces an error value as a difference between the desired fuel injection quantity and the injected fuel quantity estimate and applies this error value to a controller
26
. Controller
26
is responsive to the error value to determine a fuel command adjustment value, wherein the Initial fueling command and the fuel command adjustment value are applied to a second summing node
28
. The output of summing node
28
is the output
30
of system
10
and represents a final fueling command that is the initial fueling command produced by table
14
adjusted by the fuel command adjustment value produced by controller
26
.
While system
10
of
FIG. 3
provides for a closed-loop fuel control system utilizing a virtual sensor to achieve at least some control over variations in injected fuel quantities, it has a number of drawbacks associated therewith. For example, a primary drawback is that prior art systems of the type illustrated in
FIG. 3
are operable to compensate for variations in only a single operating parameter. Control over variations in additional parameters would require prohibitively large and difficult to manage multi-dimensional look-up tables, wherein such tables would be limited to only operating parameters capable of compensation via look-up table techniques. For operating parameters that deteriorate or change with time, for example, compensation via look-up tables simply does not work without some type of scheme for updating such tables to reflect changes in those operating parameters.
As another drawback of prior art systems of the type illustrated in
FIG. 3
, such systems are not closed-loop with respect to injector-to-injector fueling variations. For example, referring to
FIG. 16
, a plot
35
of measured fuel injection quantity vs. injector actuator commanded on-time (i.e., desired fueling command) for each injector (cylinder) of a six-cylinder engine, is shown wherein the between-cylinder fueling variations are the result of various mismatches in the fueling system hardware. As is apparent from plot
35
, the between-cylinder fuel injection quantity variations are quite pronounced and generally unacceptable in terms of accurate fueling control. While known cylinder balancing techniques could reduce such cylinder-to-cylinder fueling variations, the fuel control system of
FIG. 3
would be ineffective in reducing such variations. Moreover, the fuel control system of
FIG. 3
would further be ineffective in reducing engine-to-engine fueling variations. Referring to
FIG. 17
, for example, plots of average injected fuel vs. injector on-time for three engine fueling extremes are illustrated. Nominal engine fueling requirements are illustrated by curve
36
, minimum engine fueling conditions are illustrated by curve
38
and maximum engine fueling conditions are illustrated by curve
40
. While engines of the same type may be designed for identical fueling requirements, their actual fueling requirements may fall anywhere between curves
38
and
40
. Unfortunately, the prior art fuel control system of
FIG. 3
cannot compensate for such engine-to engine fueling variations. In general, if such control parameter variations are not attributable to the operating parameter for which the system is designed to compensate for, but are instead attributable to other error sources for which the control system of
FIG. 3
is not designed to compensate for, the system performance may actually be worse than would otherwise be the case with conventional fuel control techniques.
By the nature of their uses in a wide variety of applications, engines are typically required over their operating lifetimes to work in environments wherein many internal and external parameters that affect engine performance may vary, cannot be controlled and/or cannot be, or typically are not, measured. Heretofore, known control systems have attempted to improve injected fueling accuracy using a parameter that is both measurable and controllable. Such systems typically operate by making control changes, based on an estimated sensitivity in the fueling quantity, to this measurable and controllable parameter using assumed values for other internal and/or external parameters rather than taking into account performance effects and interactions of these other parameters. By contrast, if the injected fueling quantity can be estimated utilizing a sensor or virtual sensor that is independent of many of the internal and external parameters that affect the engine's injected fueling quantity, a robust closed-loop fueling quantity control can be performed directly on the estimated fuel quantity rather than on only one of the control parameters that affect the fueling quantity. What is therefore needed is an improved strategy for adaptively estimating injected fuel quantities based on real-time performance of certain fuel system operating conditions throughout an injection event to thereby allow for robust and accurate operation as well as straightforward integration into complex fuel control systems. Ideally, such a strategy should be capable of minimizing
Benson Donald J.
Carey David M.
Carroll, III John T.
Tarr Yul J.
Tuken Taner
Barnes & Thornburg LLP
Cummins Inc.
Kwon John T.
LandOfFree
System for estimating auxiliary-injected fueling quantities does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System for estimating auxiliary-injected fueling quantities, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System for estimating auxiliary-injected fueling quantities will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3325987