EP0352705A1

EP0352705A1 - Process for regulating the fuel to air ratio of an air/fuel mixture feeding an internal-combustion engine, of the closed-loop type and without the working of a physical fuel to air ratio measurement

Info

Publication number: EP0352705A1
Application number: EP89113602A
Authority: EP
Inventors: Mariano Sans
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 1988-07-27
Filing date: 1989-07-24
Publication date: 1990-01-31
Also published as: DE68901822D1; DE68901822T2; FR2634824A1; FR2634824B1; EP0352705B1; ES2032080T3

Abstract

The process involves determining an injection time ti for the control of the opening of a fuel injector, by adding to a static componen s of this time a sampled dynamic correction Δ t i d _n, the measurement of which is given by a linear combination of at least one of the values Δ t i d _n-k of the dynamic corrections made at the immediately preceding moments of sampling and of the estimated fuel to air ratio differences e_n-k in relation to a nominal fuel to air ratio Rc, which are measured at the current moment and at at least one immediately preceding moment of sampling.

The invention is used for regulating the fuel to air ratio of an air/fuel mixture in the event of the failure of an oxygen sensor placed in the exhaust gases of the engine.

Description

The present invention relates to a process for regulating the fuel to air ratio of an air/fuel mixture feeding an internal-combustion engine and, more particularly, to such a process of the "closed-loop" type, without the working of any physical fuel to air ratio measurement, designed to replace a process for the closed-loop regulation of such an engine from the signal supplied by an oxygen sensor placed in the exhaust gases of this engine, as a result of a temporary or permanent failure of this sensor.
There are known processes for regulating the fuel to air ratio of an air/fuel mixture by modulating an opening time (or duration) of a fuel injector by means of a multiplicative term, this regulation being adjusted so as to obtain a uniform and continuous oscillation of the air/fuel ratio about a nominal value. Whether the exhaust gases of the engine pass or do not pass through a catalytic converter limiting the pollution of the environment by these exhaust gases, the modulation of the air/fuel ratio is made the best possible, if appropriate in relation to the characteristics of this converter, by means of two types of correction, proportional and integral, introduced into the calculation of the opening time of the injectors equipping the engine to be regulated.
These corrections are usually calculated at the top dead centre of the engine cylinder in question. As a result of this, the frequency of the oscillations of the air/fuel ratio is not controlled correctly since this depends on the speed of the engine. Likewise, the amplitude of these oscillations cannot be controlled correctly because it depends on the amplitude of the corrections calculated and applied, and owing to the fact that the corrections are calculated systematically at the top dead centre it is not possible to take the dynamics of the system into account, thus causing a loss in the information to be processed and therefore less accuracy in the calculations used for controlling the amplitude of the oscillations of the mixture fuel. to air ratio.
French patent application No. 88 10114 filed on 27th July 1988 and entitled "Process and device for regulating the fuel to air ratio of an air/fuel mixture feeding an internal-combustion engine" describes a process for the closed-loop regulation of fuel to air ratio which works a signal supplied by an oxygen sensor placed in the exhaust gases of the engine and which makes it possible to control the frequency and amplitude of the oscillations of the fuel to air ratio completely by making these independent of the speed of the engine. The process employs a law of regulation of fuel to air ratio having a recurrent nature, that is to say depending on the prior states of the fuel to air ratio measurements and the values of the preceding corrections of the opening time of the injectors.
Whether a regulating process with proportional and integral correction or with a recurrent law of regulation is used, a failure of the device for carrying out the process is always to be feared. The failure is very often caused by a breakdown of the oxygen sensor used, because this detector is placed in an especially aggressive environment formed by the exhaust gases of the engine.
It is therefore expedient to provide means making it possible to ensure a suitable control of the fuel to air ratio of the air/fuel mixture in the event of a failure of the fuel to air ratio regulation normally used for this purpose and, more particularly, in the event of a failure of the oxygen sensor used for this regulation.
An object of the present invention is, therefore, to provide a process for regulating the fuel to air ratio of an air/fuel mixture feeding an internal-combustion engine, designed to function in the absence of fuel to air ratio information supplied by an oxygen sensor placed in the exhaust gases of the engine. This process will then be able to function independently or replace a fuel to air ratio-regulating process utilizing this fuel to air ratio information, in the event of a temporary or permanent failure of this oxygen sensor.
Another object of the present invention is to provide such a process capable of ensuring a continuous control of the fuel to air ratio of the mixture from a nominal fuel to air ratio value which can be different from that corresponding to the stoichiometry, under some driving conditions.
Yet another object of the present invention is to provide such a process which is effectively protected from disturbances and which allows the use of corrections distributed according to the type of corrections, so as to reduce the calculation time for these corrections.
These objects of the invention are achieved by means of a process for regulating the fuel to air ratio of an air/fuel mixture feeding an internal-combustion engine, of the closed-loop type, by controlling the opening time of a fuel injector, this opening time consisting of at least the sum of a static injection time and of a dynamic correction of the injection time, the said process being characterized in that, to form a signal representing this correction, an estimation of the instantaneous fuel to air ratio of the mixture is made by means of a model of the response of the engine to a signal representing this opening time, the difference between this fuel to air ratio and a nominal fuel to air ratio is sampled, the sampled difference is processed in a corrector with a sampled output in synchronism with the sampling of the fuel to air ratio difference, to form a signal Δ
representing the dynamic correction of the injection time, the sampled measurement Δ
_n of this signal at the moment of sampling n taking the form of a linear combination of at least one of the values Δ
_n-_k of the dynamic corrections made at the immediately preceding moments of sampling and of the estimated fuel to air ratio differences measured ê_n-k measured at the current moment and at at least one immediately preceding moment of sampling.
The farthest moment of sampling (n-k) taken into account in this linear function defines the order of the control thus obtained.
According to a first implementation of the process according to the invention, said to be of the first order, the measurement Δ
_n of the sampled signal representing the dynamic correction of the injection time is of the form:
where y and are constants adjusted as a function of the dynamics desired for the regulation, ê_n and ê_n-1 are the measurements of the estimated fuel to air ratio differences at the moments of sampling n and n-1.
According to a second implementation of the process according to the invention, said to be of the second order, the measurement Δ
_n of the sampled signal representing the dynamic correction of the injection time is of the form:
where α, β, go, g1, g₂ are constants adjusted as a function of the dynamics desired for the regulation, ê_n, ê_n-1, ê_n-2 are measurements of the estimated fuel to air ratio differences of the moments of sampling n, n-1 and n-2.
The static injection time
n, to which the dynamic correction Δ
_n is added, itself introduces an estimated value of the steady-state injection control with corrections attributable to certain operating conditions of the engine: correction of idling, of the recoupling of an air compressor or other auxiliary appliance, of altitude, etc.
Other corrections are applied to the nominal value of the fuel to air ratio of the mixture, to take into account other operating conditions of the engine: operation under full load, cold, during acceleration or deceleration, etc.
Corrections can also be applied to the static gain of the model of the response of the engine, to take into account some operating parameters of the engine, namely speed, intake pressure, temperature of the air, of the coolant, etc.
Other characteristics of the present invention will emerge from a reading of the following description and from an examination of the single Figure which illustrates a functional diagram of the regulating process according to the invention.
By assumption, the process according to the invention cannot utilize a signal representing the actual fuel to air ratio of the air/fuel mixture, such as that supplied by an oxygen sensor, currently called a "lambda sensor", placed in the exhaust gases of the engine.
According to the present invention, the fuel to air ratio measurement, which can be obtained from the signal supplied by an oxygen sensor, is replaced by an estimation of this fuel to air ratio and a substitution of this estimation for the fuel to air ratio measurement normally supplied by the sensor in a regulating process operating by closed loop, so as to preserve the advantages afforded by such functioning in terms of stability and dynamics. It is, in fact, a "pseudo-regulation" by closed loop, since the fuel to air ratio measurement used is not a physical measurement, but merely an estimation calculated from a modelling of the process, as will be described in more detail later.
Thus, with reference to the single Figure of the accompanying drawing, the present invention makes use of a reference model
representing the dynamic relation between the fuel to air ratio of the mixture and an opening time (or duration) of a fuel injector used to compose the air/fuel mixture burnt in an internal-combustion engine M, the fuel to air ratio of this mixture being regulated by means of the process according to the invention.
If:

Rc denotes a nominal fuel to air ratio set for the air/fuel mixture and
R denotes the instantaneous fuel to air ratio of the mixture, estimated by means of the model, it appears from the functional diagram of the Figure that the estimated fuel to air ratio difference:
ê = Rc- R̂
is sampled at E1 with a constant sampling period Te and is processed in a corrector C which supplies a signal representing the dynamic correction of the injection time Δ
, itself added to an estimated "static" injection time
, the sum ti being sampled at E2 in synchronism with the sampling E1. The signal ti represents an effective resulting injection time used with the model
for establishing the estimated fuel to air ratio R.

According to a preferred embodiment of the present invention, the corrector C used is a digital corrector.
The modelling of the engine makes it necessary to determine a dynamic model
(p), such that:
where Gs is the static gain of the model in the stabilized state, Md(p) characterizes the dynamics of the response of the fuel to air ratio R as a function of the control signal ti, and p is the La- place operator (sometimes referred to as s in English speaking countries.
In the steady state there is:
with:

ti = K.R.R_A
where K is a coefficient characteristic of the engine and of the operating conditions,
R is the desired fuel to air ratio of the mixture set at R = 1 and
R_A is the volumetric efficiency. It is shown that:
where N is the speed of the engine, or engine rating, Pr is the pressure at the intake manifold of the engine, go = ao [Pr + po],
ao and P_o are constants.

Bench measurements made on the engine make it possible to prepare a mapping of the values of the coefficient K in the pressure Pr/speed N system. The sampling of the above expression (1) of the dynamic model gives:
where Md(Z) is the Z-transform of the dynamic model (with Z-¹ = discrete delay), Bo is the "zero-order hold" function introduced by sampling.
The model Md(Z) is thus a recurrent model which makes it possible to estimate the instantaneous fuel to air ratio at the moment of sampling n as a function of the effective injection time Ti_n:
thus giving the sampled fuel to air ratio difference:
By means of the process according to the invention, it is possible to give the nominal fuel to air ratio any chosen value. It is possible, for example, to set:

Rc = 1 for pollution control functioning,
Rc > 1 in an acceleration period,
Rc < 1 for functioning with a "lean" mixture, likewise for pollution control purposes.

If a catalytic converter is associated with the engine in order to treat the exhaust gases, it is known that a proper functioning of this catalytic converter requires an oscillation of the fuel to air ratio of the mixture. For this purpose, according to the invention, a periodic variation as a function of the time of the nominal fuel to air ratio Rc is programmed in the form of a square-wave signal, for example alternating on either side of the desired mean value.
If C(Z) denotes the Z-transform of the transfer function of the corrector C used in the present invention, there is:
and the sampled effective injection time is given by:
where
_n is the estimated static injection time which introduces possible corrections, such as:

- an "idle speed" correction,
- a correction for the coupling to the engine of an air compressor forming part of an air-conditioning device supplied with mechanical energy by the engine,
- a correction of altitude.

Of course, the closed-loop dynamics will be characterized by the characteristic equation of the system:
Bench measurements of the fuel to air ratio of the mixture as a function of the injection time make it possible to identify the dynamic behaviour of the engine.
If these measurements reveal a dynamic relation of the first order in the fuel to air ratio and the mixture, transfer functions of the first order are chosen for the corrector C and the model
.
Thus, bv choosino:
where r is a time constant, there follows: Md(Z) = (1-a)/(Z-a) with a = e^·Te/τ.
It is demonstrated that such a choice provides the following laws of recurrence:
where r, η are functions of a, of Gs and of the dynamics desired for the regulation.
If the bench measurements reveal a dynamic behaviour of the second order, transfer functions of the second order are chosen for the corrector C and the model, that is to say, for Md(Z): Md(Z) = (α'Z + β')/(Z² - γ'Z + δ') thus giving the following recurrent relations for the estimated fuel to air ratio and the dynamic injection
with: ê_n = Rc - R̂_n where a, β, go, g1, g₂ are functions of a', β', γ', δ' and of Gs(N,Pr) and of the desired closed-loop dynamics.
In order to integrate the physical disturbances acting on the actual engine system and affecting the static and dynamic behaviour of the latter, and to ensure a zero static error in the steady state, an integral 1/(Z-1) must be preserved in the writing of C(Z), thus implying:
The signals t i _nand A tid _nobtained in this way and forming essential components of the total opening control time Ti of the injector must be combined with various corrections aimed at taking into account special operating conditions of the engine or even the ageing of the latter.
In particular, the static gain Gs = 1/K.g_o is subject to a mapped correction of the factor K as a function of N and of Pr, as seen above. This correction is of the "fast" type, that is to say it is capable of changing at each calculation cycle or at each moment of sampling.
A correction of the injector ageing can also be introduced by adding a term 5K to the factor K, the estimated value of which then becomes:
Of course, this ageing correction is of the "slow" type.
Likewise, the term go = ao (Pr + po) of Gs can experience a "slow" altimetric correction δp_alt and a "slow" self-adaptive correction on estimated value α̂ ₀ and p̂ of ao and of ^po. The corrected value:
is then used.
Finally, corrections of air temperature Cair and of coolant temperature Cwater can be applied to the static gain, the estimated and corrected value of which then becomes:
Furthermore, various corrections can be introduced by action on the nominal fuel to air ratio value Rc, particularly:

a "full load" correction,
a correction of fuel to air ratio during cold starting, a correction of "transient phase" (acceleration/deceleration).

A correction, for example in the event of the coupling of an air compressor to the engine, can also be introduced by acting on the fuel to air ratio, in addition to that mentioned above, introduced by acting on the opening duration of the injector.
All these corrections are of the "fast" type.
The total opening time of the injector can also include a term to, an "offset" correction representing a dead time in the injector control and a term st_bat representing a variation in the electrical supply voltage of the injector, this voltage being supplied by the battery of the vehicle.
Thus, with the process according to the invention which provides a regulation of the closed-loop type without using a fuel to air ratio signal supplied by an oxygen sensor, the total duration Ti of the opening control signal of the injector is expressed by the relation:
with:
This process, which is carried out by means of a law of recurrent control, is intended more particularly for replacing the regulating process described in the abovementioned patent application in the event of a failure in the oxygen sensor used, the latter process itself likewise being carried out by means of a law of recurrent control governing a sampled additive dynamic correction Δ
_n which is added to an estimated static injection time in order to establish an effective resulting injection time ti.
If the failure is temporary, the regulation by means of the process according to the present invention is abandoned as soon as the sensor returns to its normal operation, and the regulating process of the abovementioned patent application resumes the control of the fuel to air ratio regulation of the air/fuel mixture.
Of course, the regulating process according to the invention could be used independently or, in the event of a failure of an oxygen sensor, be associated with a regulating process other than that described in the abovementioned patent application. However, the process according to the invention is closely complementary to the latter process in that they both employ recurrent laws of control which can be put into effect by calculation means of the same type.
Thus, by means of the regulating process according to the invention, a complete failure of the fuel to air ratio regulation of the mixture. of an internal-combustion engine in the event of a breakdown of an oxygen sensor can be avoided by simulating the presence of this sensor.
The process according to the invention also has the advantage of allowing a continuous control of the fuel to air ratio from values below that corresponding to the stoichiometry up to values higher than this. The recurrent law of control used is designed to preserve the dynamics and stability of the regulation, despite the presence of disturbances. The corrections applied can be distributed according to the particular type (slow, fast, constant) and therefore made only at the appropriate time, thus achieving a saving of the calculation time for these corrections.
It will also be seen that the constant-interval sampling of period Te chosen in the process according to the invention, allows the law of control to be made insensitive to the variations in the engine speed, this not occurring when there is the conventional choice of a sampling at the moment of passage of a piston of the engine through the top dead centre.
Of course, it is possible to use mapped corrections in systems other than the intake- pressure/engine-speed system described above, for example in the intake mass air flow/engine- speed system, or in another system utilizing any other combination of physical measurements accessible on the engine, without departing from the scope of the present invention.

Claims

1. Process for regulating the fuel to air ratio of an air/fuel mixture feeding an internal-combustion engine, of the closed-loop type, by controlling the opening time of a fuel injector, this opening time consisting of at least the sum of a static injection time and of a dynamic correction of the injection time, the said process being characterized in that, to form a signal representing this correction, an estimation of the instantaneous fuel to air ratio of the mixture is made by means of a model of the response of the engine to a signal representing this opening time, the difference between this fuel to air ratio and a nominal fuel to air ratio is sampled, the sampled difference is processed in a corrector with a sampled output in synchronism with the sampling of the fuel to air ratio difference, to form a signal Δ

representing the dynamic correction of the injection time, the sampled measurement A tid _n of this signal at the moment of sampling n taking the form of a linear combination of at least one of the values Δ

_n-k of the dynamic corrections made at the immediately preceding moments of sampling and of the estimated fuel to air ratio differences ê_n-k measured at the current moment and at at least one immediately preceding moment of sampling.

2. Process according to Claim 1, characterized in that the measurement Δ

_nof the sampled signal representing the dynamic correction of the injection time is of the form:

where γ and are constants adjusted as a function of the dynamics desired for the regulation, and ê_n and ê_n-1 are the measurements of the estimated fuel to air ratio differences at the moments of sampling n and n-1.

3. Process according to Claim 1, characterized in that the measurement Δ

_n of the sampled signal representing the dynamic correction of the injection time is of the form:

where a, β, go, g₁, g₂ are constants adjusted as a function of the dynamics desired for the regulation, and ê_n, ê_n-1 and ê_n-2 are the measurements of the estimated fuel to air ratio differences at the moments of sampling n, n-1 and n-2.

4. Process according to Claim 3, characterized in that the coefficients a and β satisfy the relation: α + β = 1.

5. Process according to any one of the preceding claims, characterized in that the said signal representing the dynamic correction of the injection time Δ

is added to a static injection time
tis representing the estimated static injection control in the steady state and of at least one correction of the group: correction of idling, correction of coupling to the engine, correction of altitude, in order to form a control signal of the opening time of the injector.

6. Process according to Claim 5, characterized in that a signal δt_bat is added to the signals representing the static injection time and the dynamic correction, in order to correct variations of the supply voltage of an electrically controlled injector, so as to form a control signal for the opening time of the injector.

7. Process according to Claim 6, characterized in that the sum of the signals tis Δ

and δt_bat has added to it a constant offset correction (to) characteristic of the response time of the injector, in order to form a control signal for the opening time of the injector.

8. Process according to any one of Claims 1 to 7, characterized in that the model of the response of the engine to a controlled opening time of the injector comprises a component consisting of a static gain of the form G₃ = 1/K.g_o, K being a factor which experiences a mapped correction in the intake pressure Pr/engine speed N system, and go = αo (Pr + po) where αo and po are constants.

9. Process according to Claim 8, characterized in that the factor K is corrected by an additive correction 5K, taking into account the ageing of the injectors.

10. Process according to Claim 8, characterized in that an additive altimetric correction dP_alt is applied to the term (pr + po) of go.

11. Process according to Claim 10, characterized in that a self-adaptive correction is applied to the coefficients ao and po of go.

12. Process according to any one of Claims 8 to 11, characterized in that a correction of air temperature and a correction of water temperature are applied to the static gain Gs.

13. Process according to any one of the preceding claims, characterized in that the fuel to air ratio difference and the output signal from the corrector are sampled at a constant interval Te independent of the rotational speed of the engine.

14. Process according to any one of the preceding claims, characterized in that it is put into effect as a result of the detection of a failure of an oxygen sensor placed in the exhaust gases of the engine and forming part of a device for the closed-loop regulation of this engine.

15. Process according to any one of the preceding claims, characterized in that the nominal fuel to air ratio is corrected as a function of predetermined operating conditions of the engine.

16. Process according to Claim 15, used on an engine associated with a catalytic converter ensuring the anti-pollution control of the exhaust gases of the engine, characterized in that the value of the nominal fuel to air ratio is controlled according to a periodic law as a function of time, alternating on either side of the mean nominal value, in order to ensure the proper functioning of the catalytic converter.