DE102008029346B4 - Method for operating an internal combustion engine at lambda control, method for determining the storage capacity of an oxygen storage in an exhaust system and motor vehicle - Google Patents

Abstract

A method of operating an internal combustion engine (12) in a system (10) in which a controller (16) determines the amount of air and fuel supplied to the internal combustion engine (12) in which at least one broadband lambda probe (20) supplies exhaust gas from the internal combustion engine wherein the broadband lambda probe (20) provides an output signal which is corrected in dependence on the air-fuel ratio in the exhaust gas, wherein a difference between a setpoint defined by the setpoint function and the corrected output signal is formed and fed to a regulator input of a regulator which is associated with a control circuit (14) for generating control signals for the control unit (16) which determine the air-fuel ratio, characterized in that the provided output signal is differentiated, the derivative thus formed being provided multiplied by a factor Output signal is added and this sum corrected Output signal represents.

Description

The invention relates to a method for operating an internal combustion engine according to the preamble of patent claim 1, as it is known from US 5 179 929 A It also relates to a method according to the preamble of claim 4 for determining the storage capacity of an oxygen storage in an exhaust line. Furthermore, the invention relates to a motor vehicle in which a method according to the invention can be carried out.

The operation of an internal combustion engine should in the present case take place within the framework of a so-called lambda control. There is a controller that determines the amount of air supplied to the engine and fuel, that is, the air-fuel ratio. This controller operates in response to control signals provided in a control loop. The exhaust gas from the internal combustion engine is fed to a broadband lambda probe. As is known, a broadband lambda probe is one which provides an output signal (eg a voltage) and thus a measured value as a function of the air-fuel ratio in the exhaust gas supplied to it. This output signal is now used in the control loop.

Broadband lambda probes may have functional limitations due to aging or poisoning. In the typically used Nernst probes, aging leads to the effect that the signals provided by the non-fully functional broadband lambda probe have been low-pass filtered compared to signals otherwise provided by a fully functional broadband lambda probe. For low-pass filtering, a dead time may occur.

In a controlled system, these limitations in the functionality of the lambda probes can have extremely strong and harmful effects, so that during operation of the internal combustion engine low efficiency is achieved, too much fuel is consumed, and consequently also the associated emission regulations can not be met.

Before broadband lambda probes were used in an exhaust system, so-called jump probes were used. A jump probe detects a transition from a rich to a lean exhaust, but the exact air-fuel ratio could not be determined. Accordingly, a regulation was possible only to a limited extent. Switching times played a role, which were measured with the help of the jump probes. In the DE 44 36 121 C2 It is described that abnormalities in the determined switching times can be determined and corrected.

With broadband lambda probes, the situation is more complex.

The problem of limited functionality of broadband lambda probes also arises in the method of determining the storage capacity of an oxygen storage in an exhaust line. Such oxygen reservoirs are usually integrated in the catalyst provided in the exhaust system or coupled thereto. The storage capacity must be determined to check if the oxygen storage is fully functional or not. In this determination, a oxygen storage upstream broadband lambda probe is used as well as a oxygen storage downstream lambda probe. The exhaust gas line is acted upon by a defined first air-fuel ratio, followed directly by a change to the application of the defined second air-fuel ratio. Typically, rich exhaust gas is supplied first, then lean exhaust gas. Due to signals of the upstream broadband lambda probe, a first detection time is determined, and a second detection time is determined on the basis of (voltage) signals of the downstream lambda probe. The detection times are those times at which the change in the admission of the exhaust gas train reflected in the signals of the lambda probe. In the case of the broadband lambda probe, the detection time is assigned to the exceeding (possibly undershooting) of a threshold value by the measured value. In a downstream jump probe, the time of the jump is determined as the second detection time. The second detection time is that time at which the defined second air-fuel ratio is no longer obscured by the effect of the oxygen storage of the catalyst. If the second air-fuel ratio is a lean exhaust gas, the oxygen storage gradually stores oxygen. As a result, the lean exhaust gas supplied to the oxygen storage becomes less lean as it leaves the oxygen storage. Only when the oxygen storage is filled, the lean exhaust gas is passed virtually unchanged to the downstream lambda probe. Only then does the jump occur in the signal of the lambda probe. If an integral of output signals of the upstream broadband lambda probe or the associated measured values measured between the first and the second detection time is calculated, this integral is a measure of the oxygen storage capacity. However, it is assumed that the broadband lambda probe (as well as the downstream lambda probe) works correctly. Due to aging of the broadband lambda probe, the corresponding integral can be strongly falsified and thus make an incorrect statement about the storage capacity of an oxygen storage.

In the US 5 179 929 A is described that the value of a signal output of a sensor (a lambda probe) is corrected. The correction is made according to a previously recognized extent of aging.

The US 2005/0 072 139 A1 discloses a control loop for exhaust gas control, wherein the control loop comprises a PD controller, which, however, does not connect directly to a lambda probe. Namely, an adder is initially connected to an oxygen sensor, and the output signals of the oxygen sensor are supplied to its inverting input, and a comparison variable from a unit is supplied to the noninverting input, in which case the output of the adder is fed to the PD regulator.

From the DE 10 2006 010 769 A1 For example, a catalyst condition monitoring is known which also includes determining the storage capacity of an oxygen storage in such a catalyst.

The DE 101 619 01 A1 describes making an offset correction for signals from lambda probes, such as those needed for temperature drift.

It is an object of the invention, both in a method for operating an internal combustion engine according to the preamble of patent claims 1 and 2, ie in a lambda control, as well as in a method for determining the storage capacity according to the preamble of claim 4 to ensure that the effects a functional impairment of broadband lambda probe be reduced or even completely eliminated.

The object is achieved by methods having the features according to claim 1, 2 and 4. Suitable for the method according to claims 1 and 2, a specific motor vehicle is provided, namely the motor vehicle with the features according to claim 3. claim 5 represents an advantageous development of the method according to claim 4.

All aspects of the present invention have in common that at least one variable is corrected to compensate for possible malfunctions of the broadband lambda probe.

The method according to claim 1 alters the design of the controller system: In the lambda control, the output signal provided by the broadband lambda probe, ie the measured values, together with a desired value function, is used to determine a controller input of the controller. While usually a difference between a setpoint defined by the desired value function and the provided measured value is formed and supplied to the regulator input, in the present case the provided measured value is first corrected before the difference is formed.

This method is based on the recognition that there are laws in the output signals emitted by a degraded broadband lambda probe, so that an at least approximately reconstruction of the signal, as it would appear, would be the broadband lambda probe fully functional.

In this regard, it has been recognized by the inventor of the present application that by differentiating the provided measurement value, multiplying the provided measurement value by a factor and adding the derivative formed by the differentiation and the product provided by the multiplying, a corrected measurement value can be provided which is at least approximately it looks like the broadband lambda probe would be fully functional. The correction works the better the more the effects of degraded broadband lambda sensor performance affect it as if it were a first order filter. The described measure provides a first-order differentiator. The filter of the first order corresponds to an integration, the differentiator to a differentiation, and both effects are exactly the same.

In the motor vehicle according to the invention, according to claim 3 for performing this method, means for differentiating to generate a derivative, means for multiplying to produce a product and means for adding the derivative and the product are provided. Derivative, product and the sum generated by the means for adding can be processed as signals, but also abstractly treated as data values.

In the previously mentioned aspect of the invention it was assumed that the broadband lambda probe is used in the context of a rapid control, for which typically the Vorkatsonde is used. However, the inventive method can also be used when the broadband lambda probe is a Nachkatsonde, that causes a slow control. If the broadband lambda probe is a postcart probe, the voltage provided by it is used along with a setpoint (which is typically constant) to determine a setpoint function. This setpoint function is then the above-mentioned setpoint function, which is usually compared with the measured value provided by the Vorkatsonde. Also, if the aftercart probe is not fully functional, the regulation may be affected. Therefore, it is also preferred to correct the output signal provided by the postcutter and form a difference between the reference and the corrected output, and this difference is then used to determine the setpoint function for the output signal provided by the pre-sample probe.

A broadband lambda probe used as a postcutter does not differ in its aging behavior from a broadband lambda probe used as a precursor probe. The aging causes in particular a first order filtering, d. H. that a differential first-order differentiator is also connected downstream: it is differentiated and the derivative is added with the product of the provided measured value and a proportional factor in order to obtain the corrected measured value.

The invention is based, as explained above, in the first aspect concerning the method for operating an internal combustion engine (lambda control), on the knowledge that the measured values which the broadband lambda probe would emit if it were fully functional, could be reconstructed, even if she just is not. This is taken into account in the method according to claim 4 in the determination of the storage capacity. In particular, it is also possible to take this into account if, in addition to the lambda probe connected upstream of the oxygen storage device, the lambda probe connected downstream of this is also a broadband lambda probe.

As mentioned above, the correction of the output signals may be performed by differentiating the provided measurement value, adding the derivative thus formed to the multiplied by a factor provided measurement value, and this sum represents the corrected measurement value used hereinafter.

Hereinafter, preferred embodiments of the invention will be described in its various aspects with reference to the drawings. It shows:

1 2 shows schematically the construction of the components playing a role in the invention in a motor vehicle,

2 in symbol representation, a control loop from which the invention proceeds as prior art,

3 schematically a plurality of variables measured and used in the control loop during operation of the control loop 2 .

4 a section of a control loop according to the invention according to an embodiment of the invention is illustrated, and

5 by using the control loop 4 obtained, measured and used in the control loop analogous to 3 shows,

6 illustrates a portion of a control loop according to another embodiment, and

7 by using the control loop 6 obtained, measured and used in the control loop analogous to 3 shows,

8th a section of a control circuit according to the invention according to another embodiment illustrated and

9 some sizes in the loop 8th measured and determined, and

10 a control circuit according to the invention in analogy to the control loop 2 according to another embodiment of the invention,

11 some in a loop according to 10 measured and used sizes to illustrate the further aspect of the invention illustrated.

An in 1 shown and in whole with 10 designated system is typically used in a motor vehicle when an exhaust gas control is to take place. The motor vehicle is powered by an internal combustion engine 12 driven. This fuel is supplied. A control loop 14 includes a control unit 16 that the supply of fuel, possibly also of air in the internal combustion engine 12 regulates. Through the control unit 16 is set as the air-fuel ratio also in the exhaust gas from the internal combustion engine 12 looks. The exhaust gas is via a pipe 18 to a broadband lambda probe 20 passed, which emits an output signal depending on the air-fuel ratio. In the present case, when using a Nernst probe, it is assumed that an associated current control unit is part of the broadband lambda probe. The output signal of this control unit, which is typically provided digitally as a measured value, is regarded here as an output signal or provided measured value of the broadband lambda probe. This output signal is now via a connection line 22 the control loop 14 fed. In the exhaust system of the broadband lambda probe 20 downstream is a catalyst 24 in particular an oxygen storage 26 includes. The catalyst 24 and the oxygen storage 26 is another lambda probe 28 , typically a jump probe, but optionally also another broadband lambda probe, downstream. This allows the air-fuel ratio behind the catalyst 24 and the oxygen storage 26 be measured. A corresponding voltage signal or output signal (measured value) is sent via a connecting line 30 the control loop 14 fed.

In the lambda control, a desired behavior for the air-fuel ratio is set. The broadband lambda probe 20 and the lambda sensor 28 determine the actual behavior, and the control loop 14 responds to this and causes a drive of the internal combustion engine 12 or the supply of fuel to it by the control unit 16 for setting the desired air-fuel ratio λ.

2 now explains the control loop 14 :
The representation is symbolic, so that for better understanding, individual elements of the control loop with the same reference numbers as the components 1 are designated, without this means in any case a concrete arrangement. However, the signal paths always correspond to interconnections.

There is then an outer and an inner control loop. The outer loop is the slower one. It is the control circuit where the lambda probe 28 behind the catalyst 24 measured values λ _n-is to _be compared with a setpoint λ _n-soll . In particular, that of the lambda probe 28 measured value λ _n-is over the line 30 an adder 32 of the control loop 14 supplied, which has an inverting input. A nominal value λ _n-soll , which is usually constant, is supplied via the non-inverting input. The adder thus calculates the difference Δλ _n = -λ _n-soll -λ _n-ist . This difference becomes one unit 34 fed. This unit 34 sets a time-dependent function λ _v-soll for the air-fuel ratio before the catalyst 24 and the oxygen storage 26 firmly. This size is the non-inverting input of an adder 36 fed. The adder 36 has another, inverting input. This is what the broadband lambda probe does 20 measured signal λ _v-ist supplied. Thus, the difference Δλ _v = λ _v-soll - λ _{v-ist is} formed. This size becomes a regulator 38 fed. Typically, in the lambda control, a proportional-integral controller (PI controller) acts as a controller 38 used. In the PI-Regier 38 is the signal Δλ _v , which is time-dependent as Δλ _v (t), on the one hand multiplied by a proportionality factor p. At the same time, the variable Δλ _v (t) is constantly integrated and normalized to the reset time T _N. This is also multiplied by the factor p, thus obtaining for the output signal U (t) of the controller 38 :

The output of the regulator is thus composed of a quantity P which goes back to the proportional gain and a quantity I which corresponds to the normalized integral amplified by the gain p.

The controller sends the output signal U (t) to the control unit 16 further. This controls the internal combustion engine 12 depending on him from the regulator 38 of the supplied signal and thus causes a determination of λ _v-ist . Behind the catalyst 24 is through the broadband lambda probe 28 the signal λ _n-is measured, as described above via the line 30 the adder 33 is supplied.

The value of λ _n-ist only changes if too much oxygen is withdrawn from the oxygen storage device at permanently incorrectly set value λ _n-soll or, conversely _, the oxygen storage tank is completely filled. Ideally, the oxygen storage is about half full. The lambda values λ _v-should oscillate around such a value that is due to the filling of the oxygen storage 26 did not change anything with oxygen. Therefore, the outer loop is very slow.

The following is based on the 3 to 8th described effects and measures only affect the inner loop. Based on 9 and 10 the invention will be described in an aspect related to the external control circuit.

3 shows a plurality of curves. It has been assumed that λ _{n-soll is} constant. The curve 40 shows the behavior of λ _v-soll (t) as determined by the unit 34 is given. It has now been assumed that the broadband lambda probe 20 a strong aging is inferior. It therefore does not react directly to a change in the air-fuel ratio λ according to curve 40 : While the curve 40 goes through the value λ = 1 at time t ₁ , goes through a curve 42 which is the behavior of λ _v- dependent on time when passing through the curve 40 given situation, the value λ = 1 only at a later time t ₂ . An aged broadband lambda probe shows the effect of low-pass filtering of the actual signal to be measured. As between the times t ₁ and t ₂ until the time t _3, in which λ _{v is} exactly equal to λ _v-soll, λ _{v-is <v λ-soll} is the PI tries Regier 38 to compensate for this effect. The curve 44 shows the P component in the output signal of the controller 38 , the curve 46 the I component in the output signal. It can be seen that in particular the I component has a large amplitude. This reaches its maximum at the time t ₃ .

The sum U = P + I is a curve 48 played. Until time t ₄ , the curve exceeds 48 the curve 40 , This means that the control unit 16 too much fuel supplies. This causes in the curve 42 a climb over the curve 40 to be noted.

The behavior when switching from fat to lean is analogous. A lambda sensor may also have an asymmetric effect on aging. In this case, a similar behavior is seen only when changing from rich to lean or change from lean to fat. In the present reference to 3 the case of a symmetrical effect of aging, the signal λ _v-ist varies according to the curve 42 between the values λ = 0.9 and λ = 1.1. According to the curve 40 however, it should only fluctuate between the values λ = 0.95 and λ = 1.05. The aging of the probe thus causes an amplification of the desired effects, with the result that overall too much fuel is consumed and also the exhaust gas composition is worse, in particular, possibly not compliant with statutory emissions regulations.

In the present case is based on three embodiments, in the 4 . 6 respectively. 8th are shown, which measures in the control loop 14 can be taken to even when not fully functional broadband lambda probe 20 to obtain a still reasonably well-functioning control system, which ensures in particular compliance with the emission regulations.

In a first embodiment, based on 4 is explained, unlike the control loop of the prior art that of the broadband lambda probe 20 measured signal λ _v-is not directly to the inverting input of the adder 36 fed. Rather, it is a unity 50 interposed, in which a correction of λ _v-ist . In the context of the description of the present invention, it is irrelevant whether a correction to the determined value λ _{v is} actually or to the output signals of the broadband lambda probe 20 and thus the measured values, because the latter are correlated with the value λ _v-ist . In the correction unit 50 the value λ _v-ist is differentiated on the one hand. On the other hand, it is multiplied by a proportional factor smaller than 1. The derivative as well as the product are then added. The unit 50 Thus, a first-order differentiator is known to offset a first-order filter having the effect of integration. However, an aged broadband lambda probe shows exactly the effect as if the signals of a fully functional broadband lambda probe had been subjected to a first order filter. By the unit 50 Thus, the effect of aging is compensated. It becomes the value λ _v-istkorr the inverting input of the adder 36 fed. This is from the adder's point of view 36 provided a signal corresponding to that of a fully functional broadband lambda probe.

In 5 are the curve 40 as well as the curves 42 . 44 . 46 and 48 out 3 corresponding curves 42b . 44b . 46b . 48b in the case of a fully functional broadband lambda probe. This case is through the unit 50 simulated.

At time t ₂ ( 3 ) is at the aged lambda probe in the curve 42 to see a relatively steep rise. Therefore, by adding the derivative a fairly large distance between the curve 42 and 40 be balanced at this time. At time t ₄ , the curve drops 42 from. This will correct that curve 42 over the curve 40 is because the derivative is negative, and the sum of the derivative with the proportional component then approaches that through the curve again 40 given value.

This can be done by means of the unit 50 the curve 42b ( 5 ), which is almost complete with the curve 40 is identical. It is achieved a stable control. The amplitude in the I component of the controller output of the controller 38 is like by curve 46b to recognize, low. At the controller output comes according to the curve 48b just to a little overshoot just behind the respective jumps in the curve 40 ,

In the other embodiment, the value λ _v-is from the broadband lambda probe 20 a unit 52 fed. Identification of the controller used 38 ' it is that in the regulator 38 ' used factor p and the reset time T _{N are} variable. They are by a unit 52 established. The unit 52 now evaluates the values λ _v-ist . The result of the evaluation is the definition of the quantities p and T _N.

Based on 3 it has been explained that in particular the I component in the output of the regulator 38 an overshoot of the curve 48 and therefore also the curve 42 causes. If this I component is reduced (T _N increased), if necessary, the factor p is also changed (reduced), this can be determined from the curves 40 . 42c . 44c . 46c and 48c in 7 Achieve behaviors shown: As with the curves 44c and 46c to recognize, the controller responds 38 ' due to the change of the quantities p and T _N relative to the curves 3 underlying situation less strong: all rashes in the curves 44c and 46c are smaller than in the curves 44 and 46 , Accordingly, there is also in the curve 48c a lower rash than in a curve 48 , and this causes the value λ _v-is according to curve 42c less fluctuates than according to curve 42 , While in turn 42 the values between λ = 0.9 and λ = 1.1 fluctuate, fluctuates in the curve 42c the value of λ is only between 0.93 and 1.07. A variation between the values 0.95 and 1.05 is desired. The small deviation is still acceptable.

The setting of the values p and T _N should be generated depending on the behavior of. It should be recognized from λ _v-ist how strongly the correct signal is corrupted by aging of the probe. The easiest way to do this is to get out of the bend 42c Thus, from the behavior of λ _v-is over time, a quantity is derived. For example, the time offset .DELTA.t ₆₆ can be determined. This is the time offset between a jump in the setpoint signal according to the curve 40 and in the actual signal according to the curve 42c , Here it is assumed that the jump takes place starting from λ = 1. The time t _{5 is} determined at which 66% of the jump has taken place, ie in the present case a jump to 1.033 has taken place during the jump from 1.0 to 1.05. Then it becomes the curve 42c determines when it has reached the same value λ = 1.033. This is the time t ₆ . The difference t ₆ -t ₅ is now exactly the time offset .DELTA.t ₆₆ to be determined.

The size .DELTA.t ₆₆ was in this case based on 7 explained, so the state in which already an adjustment of the variables p and T _N by the unit 52 is done. Preferably, this size is constantly determined. For example, given unchanged quantities of p and T _{N, it could} also be based on the curves 40 and 42 out 3 be determined. Then, a determination of the size .DELTA.t _{66 would have} led to a change in the sizes p and T _N.

Will Δt ₆₆ as in 7 As shown, this indicates that the values for p and T _N should remain permanently as they are set.

By further aging of the broadband lambda probe, Δt _{66 can} especially increase. Then also p and T _{N would} be further adapted if necessary. The unit 52 , which is the size λ _v-is supplied, in the determination of the size .DELTA.t ₆₆ and the size λ _v -soll supplied. However, other quantities can also be derived from λ _v-ist , so that the supply of λ _v-soll to the unit 52 does not have to be necessary. For example, the size I could also be out of the regulator 38 ' the unit 52 and to the amplitude A1 in the oscillation according to the curve 46 be examined. The quantity T _N could then be increased if the amplitude A1 exceeds a threshold value. The quantity T _N can also be a continuous function of the amplitude A1. The same applies to p. Even with the derivation of other determinants, such as the above-mentioned time offset .DELTA.t ₆₆ , it is possible, the sizes p and T _N by the unit 52 as a function in To define dependence of these determinants or provide only a limited number of options that are selected by exceeding thresholds by the determinants. Instead of the values of λ _v-can evaluate the unit 52 also evaluate the controller output signals, eg. B. the shares P and I individually or only one of these shares.

In a further embodiment, an intervention takes place in the area of fixing the value λ _v-soll (t). This is a unit 54 provided that the signals λ _v-is from the broadband lambda probe 20 be supplied. The unit 54 outputs a signal that determines how large the amplitude a of the fluctuation should be in the reference signal. Without intervention shows the curve 40 a rectangular behavior with jumps of a = 0.5 above and below λ = 1 (cf. 3 ). Should the unit 54 notice that the lambda probe has aged too much, the effect of overshoot can be as it is in the curve 42 indicates that the amplitude a is reduced. In 3 you can see that the curve 42 between the values 0.90 and 1.10 fluctuates, while the setpoint curve fluctuates between 0.95 and 1.05. If now the amplitude a halved, the curve would become 40 So jump between the values 0,975 and 1,025 back and forth, the curve measures 42 Accordingly, only fluctuate as desired between about 0.95 and 1.05.

The measures described above were, as already mentioned, the broadband lambda probe 20 So in the exhaust system of the internal combustion engine 12 in front of the catalyst 24 and the oxygen storage 26 arranged probe. However, one aspect of the invention is also concerned with the aging of the lambda probe 28 if she were a broadband lambda probe.

In 9 three curves are shown: Shown is how a fully functional probe upon long-term exposure of the exhaust line with alternately lean and rich exhaust according to a curve 56 responding. The curve 56 shows the signals that the broadband lambda probe 20 gives off when switching between lean and rich exhaust gas with a much larger time scale than the fast control. A curve according to the type of curve 56 is not normally run during operation of the motor vehicle, but only for testing purposes. It is present only to discuss the behavior of the broadband lambda probe 20 shown.

Is the lambda probe 28 As a broadband lambda probe fully functional, it gives readings according to the curve 58 again. Is the lambda probe 28 aged, it gives readings according to the curve 60 again. Usually, the curves 58 and 60 like the curve 56 not completely gone through. Rather, a control to a value of about λ = 1. This is exactly an area in which in the curve 58 a big change is taking place. However, this is for the curve 60 not the case.

Therefore, the curve is preferred 60 corrected. Instead of as in the prior art (see 2 ) is the value λ _{v-which is} the broadband lambda probe 28 measures, immediately to the inverting output of the adder 32 supply, is provided in the fourth embodiment of the present invention, that the lambda probe 28 and the adder 32 one unity 62 is subordinate. The unit 62 works the same way as the unit 50 out 4 : It differentiates the value λ _n-ist and adds the lambda value λ _n-ist multiplied by a factor smaller than 1. This can be compensated for that the lambda probe 28 possibly aged.

The findings described above can also be used advantageously in a method for determining the storage capacity of the oxygen storage 26 be used.

As is known, the exhaust gas stream is initially charged with rich exhaust gas to determine the oxygen storage capacity. The excess fuel in the exhaust gas then extracts oxygen from the oxygen storage. The rich exhaust gas is supplied until no more oxygen is stored. Now, the oxygen storage is subjected to lean exhaust gas. Through the broadband lambda probe 20 in front of the oxygen storage 26 is determined when the exhaust gas, the air-fuel ratio λ = 1 is exceeded. Once this is exceeded, it will pass through the oxygen storage 26 Absorbed oxygen. Is the broadband lambda probe 20 fully functional, for example, it measures the curve when changing from rich to lean exhaust gas 64 in 11 , The time t ₇ is then the time from which the oxygen storage 26 Oxygen is supplied. For the lambda probe 28 is an example of the curve 66 represented in the event that it is a jumping probe. The voltage remains constant for a long time after the value λ = 1 has been exceeded at time t ₇ , because initially the oxygen storage decreases 26 Oxygen from the lean exhaust gas, so there is no lean exhaust gas at the broadband lambda probe 28 at. At some point, the oxygen storage is 26 filled. Then there is a jump in the curve 66 , At time t ₈ , the value 0.5 V is determined by the output voltage of the lambda probe 28 below. This value corresponds exactly to the value λ = 1. It can now from the curve 64 be determined, such as the oxygen storage capacity of the oxygen storage 26 is: For this, only the area of the area between the curve must be 64 and the axis-parallel, on the value λ = 1 constant curve are determined. At the same time, the exhaust gas mass flow must be determined during the entire process. It then applies:

Is the broadband lambda probe 20 fully functional, so does the curve 64 the actual value and the oxygen _{storage capacity} m _osc are measured correctly. It is different when the broadband lambda probe 20 aged. Then it shows the behavior according to the curve 68 , Even when using the times t ₇ and t ₈ , the integral according to the above formula would be significantly smaller than correct. The value λ = 1 is in particular also passed through at a later time t ₉ . This further falsifies the integral.

It has been described above that it is possible to differentiate the measured value or the output signal of the broadband lambda probe 20 and adding the same signal multiplied by a factor less than 1, the signal 64 as it is a fully functional broadband lambda probe 20 would show to reconstruct. When this is done, a corrected value for λ will be off the curve 68 determined. This corrected value for λ must then be used in the above formula. The integral must also be at time t ₇ , which is from the reconstructed curve 64 was determined, begin.

Just like the signal 68 correcting the first broadband lambda probe, the signal can also be corrected 66 correct if the relevant lambda probe 28 a functional impairment, z. B. is subject to aging. Then, if appropriate, the time t _{8 is determined} by means of a corrected function. A corresponding signal to the signal 66 could also be corrected if the lambda probe 28 a broadband lambda probe would be.

By referring to the 3 to 10 described aspects of the invention, it is possible to subject a combustion engine of a lambda control, even if a broadband lambda probe and possibly also a second broadband lambda probe due to aging, poisoning or other reasons is no longer fully functional. The embodiments according to the 4 . 6 . 8th and 10 may be provided in any combination with each other.

Based on 11 It has been explained that when using such a controlled system it is also possible to correctly the oxygen storage capacity of the oxygen storage 26 to determine.

As part of a lambda control with a broadband lambda probe 20 and a lambda probe 28 For the first time, a correction is made in the control circuit and in particular at the output signals of the broadband lambda probe 20 and optionally the lambda probe 28 , The latter may be for the purpose of regulating or correctly determining the oxygen storage capacity of the oxygen storage 26 respectively.

Claims (5)

Method for operating an internal combustion engine ( 12 ) in a system ( 10 ), in which a control unit ( 16 ) the amount of the internal combustion engine ( 12 ) and fuel, in which at least one broadband lambda probe ( 20 ) Exhaust gas is supplied from the internal combustion engine, wherein the broadband lambda probe ( 20 ) provides an output signal which is corrected in dependence upon the air-fuel ratio in the exhaust gas, wherein a difference between a desired value defined by the desired value function and the corrected output signal is formed and fed to a regulator input of a regulator which is connected to a control circuit ( 14 ) for generating control signals for the control unit ( 16 ) that belongs to the Determine air-fuel ratio, characterized in that the provided output signal is differentiated, the derivative thus formed is added to the multiplied by a multiplied output signal provided and this sum represents the corrected output signal. Method for operating an internal combustion engine ( 12 ) in a system ( 10 ), in which a control unit ( 16 ) the amount of the internal combustion engine ( 12 supplied air and fuel, in which a first and a second broadband lambda probe ( 20 ) Exhaust gas is supplied from the internal combustion engine, wherein the first broadband lambda probe ( 20 ) provides in the exhaust gas depending on the air-fuel ratio, an output signal in a control loop ( 14 ) together with a setpoint for determining a desired value function for a second broadband lambda probe ( 20 ) is used, wherein the control loop ( 14 ) for generating control signals for the control unit ( 16 ), which determine the air-fuel ratio, characterized in that that of the first broadband lambda probe ( 28 ) and the derivative thus formed is added to the multiplied by a multiplied output signal, and thus represents a corrected output, forming a difference between the setpoint and the corrected output that determines the setpoint function. Motor vehicle with a by a control unit ( 16 ) controlled internal combustion engine ( 12 ), the exhaust of a broadband lambda probe ( 20 . 28 ), wherein a signal output of the broadband lambda probe ( 20 . 28 ) with a correction unit ( 50 . 62 ) and the correction unit ( 50 . 62 ) and a desired signal supply line with an adder ( 36 . 32 ) are coupled to an inverting input, wherein the correction unit ( 50 . 62 ): - means for differentiating to generate a derivative from across the signal output of the broadband lambda probe ( 20 . 28 ) means for multiplying to produce a product starting from the signal output of the broadband lambda probe ( 20 . 28 ), - means for adding the derivative and the product. Method for determining the storage capacity of an oxygen storage device ( 26 ) in an exhaust line, in front of the oxygen storage ( 26 ) an upstream broadband lambda probe ( 20 ) and behind the oxygen storage ( 26 ) a downstream lambda probe ( 28 ), comprising the steps of: - applying a defined first air-fuel ratio to the exhaust gas line and then changing to a defined second air-fuel ratio, - determining a first detection time point (t ₇ ) based on output signals of the upstream broadband lambda probe ( 20 ), which are corrected to compensate for possible malfunctions of the same previously by the broadband lambda probe ( 20 ) is added, the derivative thus formed is added to the multiplied by a multiplied output signal provided and this sum represents the corrected output signal, - determining a second detection time due to outputs of the downstream lambda probe ( 28 ), Calculating an integral using output signals of the upstream broadband lambda probe measured between the first and the second detection time ( 20 ), which are corrected to compensate for possible malfunctions of the same previously by the broadband lambda probe ( 20 ) is differentiated, the derivative thus formed is added to the multiplied by a multiplied output signal provided and this sum represents the corrected output signal; and further deriving the storage capacity from this integral. A method according to claim 4, characterized in that the downstream lambda probe is a broadband lambda probe ( 28 ) and corrected by the output signals provided to compensate for possible malfunction of the same by the downstream of the lambda probe ( 28 ) is differentiated, the derivative thus formed is added to the multiplied by a multiplied output signal provided and this sum represents the corrected output signal.

Images