EP0442873B1 - A process and device for lambda control - Google Patents

Abstract

A device for lambda control operates on an internal combustion engine (11) with a catalyst (12) and lambda probes, one (13.v) arranged in front and the other (13.h) behind it. Using integration means (15), the device integrates the difference between the actual value of lambda measured by the rearward probe and the nominal one on which adjustment is to be made. The integration figure is used as a regulating rating for a means (16) of lambda control. This device and the relevant process make it possible to regulate to the actually desired lambda figure even when the forward lambda probe is making false measurements, e.g. owing to hydrocarbons in the exhaust gases in front of the catalyst or, on constant control, the faulty linearisation of the probe characteristic.

Description

The invention relates to a method and a device for regulating the air / fuel mixture to be supplied to an internal combustion engine with the aid of the actual lambda value measured by a lambda probe arranged in front of a catalytic converter. The invention also relates to an apparatus for performing such a method (compare claims 1.6, first part, see US-A-3 939 654).

State of the art

It is known from internal combustion engines with catalytic converters to arrange a lambda probe in front of and behind the catalytic converter. The front measures an actual lambda value and the rear measures an actual lambda value. The front lambda actual value is subtracted from the control lambda setpoint to which control is to take place. The control deviation formed in this way is converted by a means for lambda control into a manipulated value which is dimensioned such that the control deviation is to be eliminated by it. The rear lambda actual value serves to monitor the catalyst activity.

It is known that the rear lambda actual value fluctuates less than the front lambda actual value and that it provides more precise information about the actual lambda value. This is because the lambda value measured by a lambda probe depends not only on the oxygen content of the measured mixture, but also on the content of unburned hydrocarbons. Residual combustion and compensation of fluctuations take place in the catalytic converter, as a result of which the rear lambda probe can determine the actual lambda value of the air / fuel mixture supplied to the internal combustion engine very precisely.

Due to the high accuracy of the rear lambda actual value, it is desirable to use this actual value to form the control deviation. However, this cannot lead to practical results since a very long dead time elapses between the provision of an air / fuel volume and the point in time at which this volume, now as a burned mixture, reaches the rear catalytic converter. This makes sensible regulation impossible. However, it would be possible to correct a manipulated value formed with the aid of the rear lambda actual value by means of lambda control with a manipulated value which is formed with the aid of the front lambda actual value by a second, faster means for lambda control. With such an arrangement, however, stability problems would arise.

From US 3 939 654 a method for regulating the composition of the fuel / air mixture for an internal combustion engine is known, which is equipped with an exhaust gas probe arranged upstream and downstream of a catalytic converter. The signal Z2 of the rear probe is compared with a target value, the difference is integrated and used as a target value for a comparison with the signal of the front exhaust probe. The comparison result is used for the integral control of the mixture composition.

However, this document is limited to a single lambda setpoint which corresponds to the stoichiometric fuel / air ratio.

The invention is based on the object of specifying a method for lambda control which operates stably and allows a desired lambda setpoint to be set as precisely as possible. The invention is also based on the object of specifying a device for performing such a method.

Advantages of the invention

The invention is for the method by the features of claim 1 and for the device by the features of claim 5 given. Advantageous further developments and refinements of the method are the subject of subclaims 2-4.

The method according to the invention is characterized in that a control lambda setpoint, to which the means for lambda control regulates, is formed with the aid of the rear lambda actual value and a default lambda setpoint to which control is ultimately to be used. The setpoint / actual value comparison therefore takes place against the reliable actual lambda value, which enables the lambda value to be set exactly to the actually desired target lambda setpoint. The fact that the difference between the rear lambda actual value and the specified lambda setpoint is not used as a control deviation for a means for lambda control, but rather that the common control deviation between the control lambda setpoint and the front lambda actual value is influenced by an integration value formed with the aid of the difference value a fast, yet stable control behavior.

A device for performing such a method has a means for lambda control, a means for forming the difference between a specified lambda setpoint and the rear lambda actual value, a means for integrating the difference and a means for forming the control lambda setpoint using the integration value . The device is preferably designed as a correspondingly programmed microcomputer.

drawing

Fig. 1

ein Funktionsblockschaltbild einer Vorrichtung zur Lambdaregelung auf einen einzigen Vorgabe-Lambdasollwert mit Hilfe zweier Lambdasonden, wie sie prinzipiell aus der US 3 939 654 bekannt ist;

Fig. 2

ein Teil-Funktionsblockschaltbild betreffend einen Zusammenhang von Funktionsgruppen, der abweichend vom entsprechenden Zusammenhang gemäß Fig. 1 ausgebildet ist, um auf von Betriebspunkt zu Betriebspunkt unterschiedliche Vorgabe-Lambdasollwerte einstellen zu können; und

Fig. 3

ein Teil-Funktionsblockschaltbild entsprechend dem von Fig. 2, jedoch mit einem zusätzlichen Vordersonden-Lambdasollwert-Kennfeld.

The invention is explained in more detail in the following exemplary embodiments illustrated by figures. Show it:

Fig. 1

a functional block diagram of a device for lambda control to a single default lambda setpoint using two lambda probes, as is known in principle from US 3,939,654;

Fig. 2

1 shows a partial functional block diagram relating to a relationship between function groups, which is designed differently from the corresponding relationship according to FIG. 1 in order to be able to set different lambda setpoints from operating point to operating point; and

Fig. 3

a partial functional block diagram corresponding to that of FIG. 2, but with an additional predensor lambda setpoint map.

Description of exemplary embodiments

The device for lambda control explained below with reference to FIG. 1 is arranged on an internal combustion engine 11 with a catalytic converter 12, a front lambda probe 13.v in front of the catalytic converter and a rear lambda probe 13.h behind the catalytic converter. As functional groups, it has a front subtraction means 14.v, a rear subtraction means 14.h, an integration means 15 and a means for lambda control 16. The control value of the means for lambda control 16 is passed to a multiplication means 17, where it is multiplicatively linked to a preliminary injection time tiv to form an injection time signal ti. The injection time signal is fed to an injection arrangement 18.

A rear lambda actual value λ _{actual-h is} measured from the rear lambda probe 13.h, which is subtracted in the rear subtraction means 14.h from the actually desired lambda value, the specified lambda _target value λ _target-v . The difference is integrated in the integration means 15 and serves as the control lambda _setpoint λ _{setpoint R} for the control in the means 16 for Lambda control. The _actual lambda actual value λ _{actual-v is} subtracted from the control lambda setpoint in the front subtraction means 14.v, as measured by the front lambda probe 13.v. The control deviation formed in this way is converted by means 16 for lambda control into the control value already mentioned, a control factor FR. This procedure leads to the following control behavior.

It is assumed that the default lambda target value is 1 and that at a point in time at which the observation begins, an air / fuel mixture is currently being provided by the injection arrangement 18, which leads to the desired target lambda target value of 1. However, the internal combustion engine 11 operates at an operating point in which a relatively high percentage of hydrocarbons are generated. These hydrocarbons in the exhaust gas cause the front lambda sensor 13.v to display a richer mixture than is actually present. The measured actual lambda value is z. B. 0.99. In contrast, the actual lambda value, ie the actual lambda value, is exactly 1. The integration means 15 is at the value 1. In this case, the difference between the specified lambda target value and the actual lambda value is zero, which is why the integration means 15 does not change the set integration value. The control lambda setpoint value delivered to the front subtraction means 14.v is therefore 1. The lower front lambda actual value is subtracted from this. Because of this control deviation, the means 16 for lambda control ensures that the mixture becomes leaner. The actual lambda actual value increases in direction 1 and the actual lambda actual value rises above 1. The difference value formed by the rear subtraction means 14.h thereby becomes negative, as a result of which the integration value, that is to say the control lambda setpoint, is lowered by the integration means 15. If there has been a decrease down to 0.99, the following conditions exist. The injection arrangement 18 again ensures an air / fuel mixture with the lambda value 1 Front lambda probe 13.v measures the actual lambda value in front 0.99. This corresponds exactly to the control lambda setpoint, which is why the lambda control 16 leaves the control value unchanged, so that the injection arrangement continues to ensure a mixture with the specified lambda value 1. The rear lambda probe 13.h measures the lambda value 1. Since this corresponds to the specified lambda target value, the integration value from the integration means 15 remains unchanged at 0.99.

In this way, the aforementioned coupling of signals ensures that the means for the lambda control 16 exactly reaches the desired target lambda setpoint, although the actual lambda value used for the control incorrectly measures the actual lambda value. However, regulation to the correct value takes place at a relatively slow speed. This is because, due to the dead time already mentioned, the speed at which the integration means 15 integrates must not be very high. You will z. B. selected so that the oscillation of the actual lambda value around an average is approximately 1/5 to 1/10 of the control oscillation in the control circuit with the means 16 for lambda control.

In Fig. 1, a means 21 for integration release is shown, which acts on the integration means 15. It is used to block the integration process when special conditions exist in which there is no regulation to a desired lambda value, e.g. B. in overrun cut-off mode or in full load operation.

In practice, the same lambda value is not continuously regulated, but different lambda values are desired for different operating states. In particular, the oil is enriched with increasing load in order to counteract an increase in nitrogen oxides in the exhaust gas. Accordingly, one does not become a single one when practicing the invention Use default lambda setpoint as assumed in FIG. 1 to explain the basic principle, but different default lambda setpoints will be specified for different operating points. It is expedient to store such target values in a map, which can be evaluated as address values with the aid of values of operating variables. An arrangement with such a map is shown in FIG. 2.

The arrangement according to FIG. 2 has a default lambda setpoint map 19 which can be addressed via values of the speed n and a load-dependent variable L. The default lambda _target value λ _{target V} read in each case is in turn given to the rear subtraction means 14.h. At the same time, it arrives at an addition means 20, to which the integration value is also supplied by the integration means 15. The rest of the arrangement essentially corresponds to that of FIG. 1. Only the means for enabling integration 21 are missing. The reason for this is explained further below.

The purpose of the addition means 20 will be explained using an example. It is initially assumed that this addition means is missing, that is to say the structure according to FIG. 1 is present, but gives the default lambda target values to the rear subtraction means 14.h with a default lambda target value map. First of all, let the output value be 1. The state explained with reference to FIG. 1 then exists, in which the actual lambda actual value is 0.99. Now the operating point changes, which results in a new default lambda setpoint of 0.98. The actual lambda actual value measured at this lambda value is 0.97. The integration means 15 must then integrate in the embodiment according to FIG. 1 from 0.99 to 0.97, which takes up a lot of time. In the embodiment according to FIG. 2, the integration means 15 integrates to - 0.001 if the specified lambda setpoint 1 and the lambda actual value front is 0.99. The default lambda setpoint jumps from 1 to 0.98 with an associated actual lambda value of 0.97, the new value of 0.98 is given directly to the addition means 20. The integration value remains at 0.01. A change in the default lambda setpoint thus directly affects the means for lambda control 16 without the integration means 15 having to be active. It only has to take action if there is a different difference between the actual lambda value rear and the actual lambda value for the new operating point than for the operating point that previously existed.

Even if the last-mentioned aggravating condition exists that different differences between the actual lambda value rear and the actual lambda value belong to different operating points, it can be avoided that the means for integration 15 must compensate for such a difference by integration with each operating point jump. This can be achieved through structural adaptation. With regard to methods for structural adaptation, reference is made to DE 36 03 137 A1 (US Ser. No. 6696). The possibility of adaptation is indicated in FIG. 2 in that the integration means 15 are supplied with values of operating variables, namely values of the speed n and values of a load-dependent variable L. The integration means 15 is constructed as a map. An integration value that was learned in the past is stored in each map point. The integration value corresponds to the difference between the actual lambda value rear and the actual lambda value for the relevant operating point. If there is a change from one operating point to another, the new default lambda setpoint from the default lambda setpoint map 19 and the associated integration value from the associated map point of the adder 15 arrive at the addition means 20. There are no map points for different values of the addressing variables. No integration value is output for these points, which corresponds to the blocking of integration by the means for integration release 21 in the embodiment according to FIG. 1.

An embodiment will now be explained with reference to FIG. 3, which allows a very quick adjustment to a new lambda value after a change in operating point, even without structural adaptation. However, adaptation is also possible, which can then be easily broken down into a global and a structural part.

The embodiment according to FIG. 3 differs from that according to FIG. 2 in that it is not the default lambda target value from the default lambda target value map 19 that is given to the addition means 20 as the lambda target value, but a pre-probe lambda target value from a pre-probe lambda target value. Map 22. The content of this pre-probe lambda setpoint map 22 is identical to the content of a conventional lambda setpoint map. One of these has already taken into account that the lambda probe arranged in front of the catalytic converter increasingly measures incorrectly with increasing hydrocarbon content in the exhaust gas. Is before a certain operating point, for. For example, if the lambda value 0.98 is desired, but it is known that the front lambda probe measures 0.96 at this lambda value, the value 0.96 is stored for the operating point in question in the conventional map and thus also in the pre-probe lambda setpoint map. The lambda value 0.98 is actually set with this setpoint.

The pre-probe lambda setpoints and the default lambda setpoints are recorded for all operating points using a measurement setup. The values are stored in the maps. If an engine used in practice exactly matches the engine with which the measurement was made and if this also applies to the lambda probes used, the integration means 15 never need to be integrated, since for each operating point the predefined lambda setpoint value is accurate for each operating point the associated default lambda setpoint results. Soak the characteristics of the engine or probes, however, on the properties of the parts that were used when recording the characteristic maps, be it due to manufacturing-related scattering or be it due to aging, the integration means 15 compensates for the deviation. For the most important errors, especially for deviations in the probe properties, the compensating integration value is the same for all operating points. The integration means 15 can accordingly be set to a very slow integration speed. Rapidly changing differences from operating point to operating point in the difference between the actual lambda value front and the actual lambda value rear are compensated for by the different lambda setpoints from the two maps. Long-term changes or differences in scatter are eliminated by the initial value of the integration means 15. If it is to be taken into account that changes in aging or differences in variation can be dependent on the operating point, this can be done by adaptively changing the values in the pre-probe lambda setpoint map 22. This is indicated in Fig. 3 by the fact that the output signal from the integrator 15 acts on the mentioned map. Structural adaptation takes place by changing the map values. A part of the integration value from the integration means 15 can be used for global adaptation. With regard to applicable adaptation methods, reference is again made to the above-mentioned patent application.

The previous statements were valid for means 16 for lambda control with two-point behavior, as well as for those with constant behavior. If the consideration is specified for means for continuous lambda control, there is yet another advantage of the described method. It should be noted that the lambda value-voltage characteristic of a lambda probe is non-linear in all its areas. However, it can be linearized in various areas with fairly good accuracy, e.g. B. in a range of about +/- 3% around the lambda value 1. With the help of the linearized characteristic, a relatively simple control procedure can be carried out. However, due to the small differences between the actual characteristic curve and the linearized characteristic curve, there are slight deviations between the actual lambda value and the measured value. It is then slightly incorrectly regulated. The integration means 15 is also able to correct this error, as described above with reference to the hydrocarbon error.

The linearization error just described has a particularly negative effect if the lambda sensor is operated temporarily at a temperature that is relatively far from the temperature for which the actual characteristic curve was determined, from which the linearization was then carried out. The characteristic curve changes depending on the temperature. Now, however, the rate of change of the probe temperature is lower than the rate of integration of the integration means 15. If the actual lambda value is incorrectly measured on the front lambda probe 13.v due to the shift in the characteristic curve, this error is also eliminated with the aid of the rear lambda probe 13. h and the integration means 15 balanced. This is possible because the temperature behind the catalyst 12 fluctuates significantly less than in front of it.

As long as the integration is permitted, it is advantageous to integrate at a speed that is proportional to the difference between the rear lambda actual value and the lambda setpoint. As a result, the integration means 15 changes the control lambda setpoint for the means 16 for lambda control the faster the further the actual lambda value deviates from the lambda setpoint. This ensures that the desired lambda setpoint is reached as quickly as possible. However, the speed of integration must not be too high, since, due to the dead time mentioned at the beginning, a control oscillation could otherwise be built up. It is therefore advisable to limit the integration speed upwards. A method in which the speed of integration remains constant regardless of the value of the difference mentioned is simpler to carry out. This integration speed is chosen to be as high as possible, but only so high that even in the worst case there are no control oscillations with an impermissibly high amplitude.

In all the embodiments described so far, it is assumed that the difference between the actual lambda value rear and the preset lambda setpoint corresponds to the difference between these two variables. However, it is also sufficient to simply determine whether one value is greater than the other or not, and to integrate in one direction or the other, depending on the comparison result.

Claims (6)

1. Method for lambda control, in which the lambda value of the air/fuel mixture to be fed to an internal combustion engine is controlled to a control set lambda value (λ set-R) with the aid of the front actual lambda value (λ act-v) measured by a lambda probe arranged upstream of a catalytic converter, and in which the rear actual lambda value (λ act-h) is measured by a second lambda probe downstream of the catalytic converter, in which method a differential value between the rear actual lambda value and a prespecified set lambda value (λ set-V) is formed, with which differential value an integration value is formed and the control set lambda value is formed as a function of the integration value and a further prespecified set lambda value (λ set-VS), characterised in that the prespecified set lambda values are varied as a function of the operating characteristic variables, one prespecified set lambda value (λ set-V) being, if appropriate, identical with the further prespecified set lambda value (λ set-VS).
2. Method according to Claim 1, characterised in that the control set lambda value (λ set-R) is formed by adding the integration value to the prespecified set lambda value (λ set-V) (Fig. 2).
3. Method according to Claim 1, characterised in that the control set lambda value (λ set-R) is formed by adding the integration value to a front-probe set lambda value (λ set-VS) (Fig. 3).
4. Method according to one of Claims 1 to 3, characterised in that the integration values are used for adaptation.
5. Device for lambda control having

   - a means (16, 1f4v) for lambda control to a control set lambda value (λ set-R), which means is fed, as an actual value, to the front actual lambda value (λ act-v) as it is measured by a lambda probe (13.v) to be arranged upstream of a catalytic converter (12),

   - a means (14.h) for forming a differential value between a prespecified set lambda value (λ set-V) and a rear actual lambda value (λ act-h) as it is measured by a lambda probe (13.h) to be arranged downstream of the catalytic converter,

   - a means (15) for integrating the differential value and

   - a means (15, 20) for forming the control set lambda value with the aid of the integration value

   characterised by a prespecified set lambda value characteristic diagram (19).
6. Device according to Claim 5, characterised by a front-probe set lambda value characteristic diagram (22) and an addition means (20) which forms the control set lambda value from the respective integration value and the respective front-probe set lambda value (λ set-VS).

Images