EP2906801A1

EP2906801A1 - Method for adjusting an injection behavior of injectors in an internal combustion engine, engine control unit and system for adjusting an injection behavior

Info

Publication number: EP2906801A1
Application number: EP13773618.7A
Authority: EP
Inventors: Jörg REMELE; Aron Toth
Original assignee: MTU Friedrichshafen GmbH
Current assignee: Rolls Royce Solutions GmbH
Priority date: 2012-10-10
Filing date: 2013-10-04
Publication date: 2015-08-19
Also published as: US9664134B2; CN104854331A; WO2014056595A1; US20150211430A1; DE102012020489A1; HK1213618A1; CN104854331B; DE102012020489B4

Abstract

The invention relates to a method for adjusting an injection behavior of injectors in an internal combustion engine, comprising the following steps: switching off an injector; detecting a crank angle signal of the internal combustion engine; transforming the crank angle signal into the frequency range by means of a discrete Fourier transformation; detecting and storing aquantity of the harmonic of the 0.5th order of the Fourier transform of the crank angle signal, and assigning the quantity to the switched-off injector; switching on the switched-off injector; performing the previous steps sequentially for all injectors of the internal combustion engine; forming an average value of the stored quantities with respect to all injectors, and correcting thecontrol of the injectors using a deviation from the average value of a quantity associated with an injector that is to be corrected.

Description

Method for adjusting the injection behavior of injectors in an internal combustion engine, engine control unit and system for adjusting a

Injection behavior

The invention relates to a method for equalizing an injection behavior of injectors in an internal combustion engine according to claim 1, an engine control unit according to the preamble of claim 9, and a system for approximating a

Injection behavior of injectors according to claim 10.

Method for equalizing an injection behavior of injectors in one

Internal combustion engine are known. This is based on the general problem that injectors for injecting fuel into cylinders of the internal combustion engine with identical activation, in particular energization, have production-related variations in their opening behavior. So are the injectors of a

Internal combustion engine with identical energizing parameters, in particular driven with an identical energization duration, they nevertheless inject different amounts of fuel into the individual cylinders. With small injection quantities, the scatter is so great that some injectors inject fuel into their assigned cylinders, while others do not open. Therefore, a pre-injection and a post-injection can not be displayed if the injectors scatter strongly. It is therefore desirable to reduce the scattering in the opening behavior of the injectors in the operation of the internal combustion engine as far as possible.

German Offenlegungsschrift DE 100 55 192 A1 discloses a method for concentricity control for diesel engines, with the aid of which correction factors for

Injection quantities for equality of the individual cylinders with respect to their speed components are obtained. For this purpose, an impulse response spectrum of each cylinder is determined once by the cylinders are switched off individually one after the other, the 2

Speed is recorded above the crank angle. In addition, the speed history of the healthy engine, that is, when all cylinders are working normally, is measured. By differentiating the curve of the healthy engine and the curves for individually deactivated cylinders new curves are generated, which reflect the influence of each cylinder on the overall speed curve. These response curves are Fourier decomposed. In this case, preference is given to low-frequency oscillations, in particular of the 0.5th to third order, and associated spectral impulse responses of the individual cylinders and the individual orders of the harmonics are recorded in a matrix. During operation of the engine, the speed curve of the crankshaft over the angle is constantly recorded and Fourier transformed. The Fourier coefficients, preferably those of the low-frequency oscillations, in particular the harmonics of the 0.5th to third order, are combined as a vector. The correction values for the injection are determined by scaling the vector thus obtained scalar with the matrix representing the impulse responses.

It turns out that the method is based essentially on the assumption that the base vectors associated with the harmonics of the Fourier transformation are linearly independent of each other, so that they form an orthogonal basis of a vector space. In practice, however, it turns out that this assumption is incorrect, with the

corresponding vectors are at least partially collinear or at least not orthogonal to each other. The method is therefore not reliable with the desired result feasible.

The invention has for its object to provide a method which allows safe and efficient injector equalization, so that a series dispersion can be compensated. The process should be simple and inexpensive and in particular during operation of the internal combustion engine to be feasible. The object of the invention is also to provide an engine control unit, by means of which the method is feasible. It is another object of the invention to provide a system with which a

Injection behavior of injectors in an internal combustion engine is gleichgleichbar.

The object is achieved by providing a method with the steps of claim 1. In this case, after a start of the method, first a first injector is switched off. A crank angle signal of the internal combustion engine is detected and transformed into the frequency domain by means of discrete Fourier transformation. From the discreet

Fourier transformation in particular results in an amount and an angle of the

Harmonics of the 0.5th order, in the context of the method, only the amount is detected and stored. The amount is assigned to the injector which is turned off when it is detected alone. Thereafter, the disconnected injector is turned on again. These steps are performed sequentially for all injectors of the internal combustion engine, so that in each step only one injector is switched off. The amounts recorded in the various steps can thus be unambiguously assigned to a switched-off injector. After, for each injector of the internal combustion engine, an amount of the harmonics of the 0.5th order of

Fourier transform the crank angle signal was detected, stored and assigned, an average of the stored amounts is formed over all injectors, that is, all the individual injectors associated, stored amounts are averaged. Actuation of the injectors is now corrected based on a deviation of the amount associated with an injector to be corrected from the mean. This means that for each injector a difference of its associated amount is calculated to the mean, this difference or deviation is a measure of the correction of the control of the injector.

In this way, a kind of regression to the mean value is carried out for virtually all injectors, their injection behavior is thus adjusted so that the measured for each injector amount of the harmonics of the 0.5th order of

Fourier transform of the crank angle signal approximates the mean of all injectors. Unlike the known method is not unique

Pulse response spectrum or an impulse response matrix determined, the correction during ongoing engine operation then exclusively by forming a

Scalar product of the currently measured contributions of different order to

Fourier spectrum of the crank angle signal with the impulse response matrix are calculated, but it is always a comparison of the injection behavior of the individual injectors - based on the magnitude of the harmonics of the 0.5th order - performed to a medium injection behavior. Because of this always performed

Individual comparisons - with switching off individual injectors - with the current mean value, it is possible to dispense with a consideration of higher-order contributions, and to limit itself to the 0.5th order. As a result, the property of the higher harmonics, not aufzuspannen orthogonal vector space, no longer relevant. Therefore, a precise adjustment of the injection behavior of the individual injectors is possible, so that all injectors inject at least approximately the same amount of fuel. It will also be possible to have a pilot injection and / or a

To realize or present post-injection. A pre-injection is advantageous because in this way a softer combustion process and a reduction in the formation of nitrogen oxide can be realized. A post-injection leads to a

Temperature increase of the exhaust gas, which is advantageous for the downstream exhaust aftertreatment.

The method is preferably performed by an engine control unit, wherein the crank angle signal - that is, a speed curve of the crankshaft over the crank angle - preferably detected by a crankshaft sensor and forwarded to the engine control unit. This is anyway in modern internal combustion engines

Crankshaft sensor provided, and it is also an engine control unit includes. For carrying out the method, therefore, only components are used which are present in the internal combustion engine anyway. Thus, there are no additional sensor, device and / or cabling costs for carrying out the method. The algorithm for performing the method is preferably implemented in the engine control unit.

Preferably, the engine control unit is synchronized by the signal of a camshaft sensor to work cycles of the cylinders of the internal combustion engine. This can be done once after or at the start of the internal combustion engine or continuously. Also, a camshaft sensor is usually comprised by an internal combustion engine, and a synchronization of the engine control unit on the power strokes of the cylinder is also carried out for the usual engine control. In that regard, no arises here

Additional effort by the procedure.

A method is preferred, which is characterized in that a correction for an injector is performed only when the deviation of the amount of the harmonic of the 0.5th order detected and stored for it

Fourier transform of the crank angle signal from the average formed across all the injectors exceeds a predetermined threshold. This procedure The idea is based on the idea that not every little deviation from the

Mean value is already relevant in practice. Therefore, in order to keep the injector equalization efficient, a threshold value can be sensibly set which, when exceeded, should actually be corrected by the deviation associated with an injector. It is then first determined for each injector whether the deviation exceeds the predetermined threshold, and only if this is the case, the correction of the control of this injector is actually performed.

A method is also preferred which is characterized in that, for each injector, a difference amount is calculated as the difference of the amount associated with the injector to an amount detected and stored when all of the injectors are turned on, the amounts assigned to the individual injectors Difference between the averaging and the correction.

This approach is based on the idea that the amount of the harmonics of the 0.5th order of the Fourier transform of the crank angle signal in the event that all injectors are turned on, the internal combustion engine is working normally, not necessarily disappears or at least close to zero. Thus, if an amount distinctly different from zero can already be detected for the normally operating internal combustion engine, the amounts measured for the individual injectors that have been switched off are preferably related to this amount, by adding their values for the further method

Differences to this amount are calculated and considered. Also the

Averaging then refers to the difference amounts calculated in this way, and the correction of the control of the injectors is carried out correspondingly on the basis of the deviations of the difference amounts to an average formed therefrom. The

Difference amounts are typically signed, ie no amounts in the strict mathematical sense.

It is possible that the amount used as the reference point for the amounts assigned to the individual injectors, which is measured in the normally running internal combustion engine, is recorded and stored once, for example after a start of the internal combustion engine. It is also possible to enter this amount in

at predetermined intervals or consecutively always to capture and store when no injector is turned off. In such a case, it is preferable the stored in a memory value replaced by a respective current, newly detected value.

It is therefore apparent that the method is preferably performed not on the basis of the absolute values, but rather on the basis of the difference amounts relative to the amount of the harmonics of the 0.5th order as the reference point when the engine is running normally, if this amount, ie Reference point, at least to a significant extent different from zero. If this is not the case, ie if the amount is zero or at least close to zero, the method can be carried out on the basis of the absolute values recorded and stored for the injectors without difference. However, it is readily possible to carry out the method on the basis of the difference amounts in this case, too, in particular because the result does not differ from the method without subtraction if the amount is zero in the case of a normally running engine. Also, the difference amounts are after all "amounts" in the sense of claim 1.

A method is also preferred which is characterized in that at least two iterations of the method are performed. In this case, the method is preferably iterated as long as it is carried out sequentially in succession until the

Deviation of each injector from the average value formed for all injectors does not exceed the predetermined threshold. The method is therefore preferably repeated until the deviation from the mean value for all injectors is less than the predetermined threshold value. Thus, it can be ensured - at least in a range that is relevant to practice - that in fact all the injectors inject substantially the same amount of fuel. A scope relevant to the practice can be determined by setting the predetermined threshold.

A method is also preferred, which is characterized in that the

Control of the injectors is corrected so that when correcting a total power of the engine is not changed. This means that the injectors are virtually corrected in opposite directions. If, therefore, the amount of fuel injected by a first injector is increased, the fuel quantity injected by a second injector is preferably also reduced or the quantities of fuel injected by a plurality of other injectors are correspondingly reduced, so that overall the total output of the internal combustion engine does not change. Injector equalization using the Method is performed, so preferably does not lead to a change in the currently present load point of the internal combustion engine. In particular, it is avoided that the internal combustion engine accelerates suddenly - with negative or positive signs - due to the method. It is possible that this condition is ensured virtually outside the process, for example, by the process speed control is superimposed. However, it is also possible to provide the condition quasi within the method by taking it into account in the correction of the control of the individual injectors inherently.

It is also a method is preferred, which is characterized in that a

Control of the injectors is corrected by a Bestromungsdauer adapted for them. In this case, the energization duration of a single injector is changed so that the desired correction of the injected fuel quantity is achieved.

For example, the energization duration can be extended if the injector is to inject more fuel. It can be shortened if the injector is to inject less fuel.

A method is also preferred, which is characterized in that the

Bestromungsdauer for an injector is adjusted by a current duration difference is credited to the currently existing Bestromungsdauer, which is calculated according to the following equation: = [MW - ABetrag [i]) K (1)

In this case, i is a running variable which runs over the individual injectors and whose value in each case indicates a currently considered injector. ABD [i] means the

Bestromungsdauerdifferenz, which is to be counted on the current Bestromungsdauer for the injector i. It means to add that the - positive or negative - Bestromungsdauerdifferenz is added to the currently available Bestromungsdauer. MW is the mean derived from the individual injectors

Difference amounts of the amounts of the harmonics of the 0.5th order to that for the normal running engine, so when all the injectors are turned on, detected and stored amount is calculated. ABetrag [i] is equal to the difference calculated for injector i. Thus, the mean value MW is so over all injectors the individual difference amounts ABetrag [i] formed mean value. K is a constant which is chosen so that a suitable correction of the energization time is possible.

It is preferably already ensured within the correction of the energization duration that the total power of the internal combustion engine is not changed during the correction. This is ensured by the aforementioned equation (1) preferably under the

condition

£ A5D [/] = 0 (2) is applied. Where Σ is the summation sign, and run variable i runs over all injectors. In the calculation of the energization duration differences for the individual injectors, care is taken that their sum over all injectors always equals 0. If certain energizing periods are increased, others must

Energizing be reduced accordingly, so that the total

Summation condition remains satisfied and cancel the individual Bestromungsdauerdifferenzen against each other.

A method is also preferred which is distinguished in that the constant K is selected as a function of a currently existing load point of the internal combustion engine. Preferably, a table with values for the constant K, which are assigned to different load points of the internal combustion engine, is stored in a memory of the engine control unit. Depending on the currently available load point of the

Combustion engine is then the corresponding value for the constant K to

Implementation of the method used.

The following is also shown: The method is preferably carried out at an operating point of the internal combustion engine by operating under load or at idle. In particular, the method is readily feasible in such operating points. It turns out that in large engines, such as engines, which

Generators drive, in engines for diesel locomotives or ships, or similar, in particular multi-cylinder large engines, a coasting phase, as it is known from the operation of a conventional motor vehicle for road traffic, usually does not exist. The term shear phase refers to an operating state of the

Internal combustion engine by being dragged by a rolling vehicle. In contrast, large engines basically work only under load or at idle. There are Various methods are known whose mode of operation for injector equalization is based on that they are performed in a coasting phase of a motor vehicle. These methods are correspondingly not applicable for large engines, because here generally no shear phase exists. In contrast, the method proposed here is just suitable for large engines, and it is also preferably carried out in a large engine. The particular suitability of the method for large engines results from the fact that it is readily feasible in an operating point of the internal combustion engine, in which this operates under load or at idle.

The object is also achieved by providing an engine control unit for an internal combustion engine having the features of claim 9. The engine control unit is characterized in that it is designed to carry out a method according to one of the previously described embodiments. This means in particular that an algorithm for implementing the method is implemented in the engine control unit. In addition, a connection of a crankshaft sensor to the engine control unit is preferably provided, so that it can detect a crank angle signal and process it in the sense of the method. In addition, preferred are on the

Engine control unit Interfaces for connecting the individual injectors of the

Internal combustion engine provided so that they can be energized by the engine control unit and off individually and can be turned on.

Finally, the object is also achieved by providing a system for equalizing an injection behavior of injectors having the features of claim 10. In this case, the system is used in particular for carrying out a method according to one of the previously described embodiments. The system comprises a switching means, which is designed so that with its help, the individual injectors are selectively switched off and on. It further comprises a detection means, which is designed so that a crank angle signal of the internal combustion engine can be detected. That's it

Detection means preferably designed as a crankshaft sensor. The detection means is operatively connected to a transformation agent such that the one of the

Detected detecting crank angle signal is forwarded to the transformation means. The transformation means is designed so that with its help the

Crank angle signal in the frequency domain by means of discrete Fourier transform is transformable. A memory means is provided which is arranged so that an amount of the harmonics of the 0.5th order of the Fourier transform of the Crank angle signal can be detected and stored with his help. For this purpose, the transformation means and the storage means are preferably correspondingly

operatively connected. The storage means is also adapted to allocate the detected and stored amount to an injector disabled upon detection and storage thereof. Furthermore, an averaging element is provided, which is designed so that with its help a calculation of an average value of the amounts stored in the storage means is possible over all injectors. It is also provided a correction means which is designed so that with his help a

Deviation of the amount associated with an injector to be corrected from the

Average is calculable, wherein a control of the injector based on the

calculated deviation can be corrected.

Preferably, the system comprises an engine control unit, in particular a

Engine control unit according to the embodiment described above.

The engine control unit preferably comprises the switching means, the transformation means, the storage means, the averaging element and the correction means.

A system is preferred in which - preferably also of the

Engine control unit comprises - a difference forming means is provided, by which for each injector a difference amount as the difference of an injector associated amount to an amount that is detected and stored when all injectors are turned on, is calculable. In this case, of course, preferably also a detection and storage means is provided for the amount which is detected and stored when the engine is running normally. In this case, the system is preferably designed such that the difference amounts of the averaging and the correction assigned to the individual injectors are used.

Moreover, the system is preferably designed so that the embodiments described as being preferred in the context of the method can be carried out by the system. In particular, the system is designed so that

Bestromungsdauerdifferenzen which are calculated according to the previously described equation (1), wherein preferably at the same time the previously described condition (2) is maintainable to ensure that the total power of the internal combustion engine is not changed by the injector equalization.

In this case, corresponding means for carrying out the energization duration adaptation according to the said equation (1) and under said condition (2) are preferably provided in the engine control unit.

The invention will be explained in more detail below with reference to the drawing. The single figure shows a flow chart, which represents an embodiment of the method.

The method starts in a step 1, after which a number of cylinders of the internal combustion engine is first determined in a step 3. It is provided in the embodiment of the method shown in the figure that each cylinder is assigned exactly one injector. Therefore, here the number of cylinders also corresponds to the number of injectors. However, it is still possible in another embodiment of the method that the internal combustion engine has more than one injector per cylinder. In this case, preferably not the number of cylinders, but the number of injectors is determined in step 3.

In step 3, a run variable i is also defined and initialized, preferably with the value 0 assigned to it.

In a query 5, the current value of the running variable i is compared with the number of cylinders determined in step 3. In the following, it is assumed that the run variable is first initialized with the value 0, so that the value 0 is also assigned to the run variable i for the first injector for which the method is carried out. It is readily apparent how the procedure would be changed if the

Running variable with another value, for example, would be initialized with the value 1.

In query 5, namely, when the running variable i is initialized with the value 0, it is checked whether the value of the run variable is smaller than the number of cylinders determined in step 3. If so, the process proceeds to a step 7 where the injector associated with the current value of the run variable i is turned off. Subsequently, in a step 9, an amount or difference of the harmonics of the 0.5th order of the Fourier transform of the crank angle signal is detected and stored and assigned to the switched-off injector.

In a step 11, the value of the running variable i is increased by one; at the same time the switched-off injector is switched on again. The method then jumps back to the query 5, in which it is checked again whether the now present, the present value of

Running variable i is even smaller than the number of cylinders. It turns out that in this way a loop 13 is traversed so often until for each injector in step 9 sequentially one after the other an amount or difference was detected. The last injector is assigned a value of the running variable i, which corresponds to the number of cylinders reduced by one. Therefore, after detecting the amount or difference amount for the last injector in step 9, the value of the running variable is increased to a value corresponding to the number of cylinders. If this is found in the query 5, the method proceeds to a step 15.

In turn, the value of the run variable i is initialized in this case, in particular set to 0 in the embodiment of the method discussed here.

In a following step 17, an average value is formed from the detected and stored amounts or difference amounts for the individual injectors.

The method then enters a query 19, in which it is again checked whether the current value of the run variable i is smaller than the number of cylinders determined in step 3. If this is the case, the method proceeds to a step 21, in which a correction of the control of the injector, to which the current value of the running variable i is assigned, is carried out. This preferably takes place on the basis of a difference amount assigned to the solely switched off injector to an amount determined for the normal operation of the internal combustion engine, as well as on the basis of an average value of the difference amounts for the individual injectors. Preferably, a

Bestromungsdauer adapted for the injector, wherein a Bestromungsdauerdifferenz is counted on the currently existing Bestromungsdauer. The

Bestromungsdauerdifferenz is preferably after the above Equation (1) calculated, this is preferably applied under the above condition (2).

In a following step 23, in turn, the value of the running variable i is increased by one. The method then jumps back into the query 19, so that in this respect a loop 25 is realized. This loop is again run through until a correction has been carried out for all injectors, or until the value of the running variable i in the query 19 first corresponds to the number of cylinders determined in step 3, because in the selected embodiment of the method in which the Running variable i is initialized to 0, the last injector to be corrected is assigned a value which is one by one compared to the number of cylinders.

If the value of the running variable i in the query 19 is therefore identical to the number of cylinders determined in step 3 for the first time, the method ends in a step 27.

The correction of the energization duration in the step 21 for the cylinder, to which the current value of the running variable i is assigned, is preferably carried out only if a deviation of the amount or difference amount from the mean value

exceeds the predetermined threshold. Otherwise, no correction is made to the injector, and the process proceeds to step 23.

Preferably, the method is iterated, ie jumps - possibly after a predetermined waiting time - from step 27 back to step 1, wherein this iteration, or provided between steps 27 and 1, not shown in the figure loop is passed through so long until the deviations of the individual amounts or difference amounts for the individual injectors from the mean value are all smaller than a predetermined threshold value. In this case, it is possible for this threshold value to be identical to the threshold value which is selected for the decision as to whether a correction of an individual injector is carried out. However, it is also possible to provide as a termination condition for the iteration of the entire method a threshold deviating from this threshold, which may preferably be greater or smaller than the threshold value for the correction of the individual injectors. Overall, it can be seen that with the aid of the method, a very precise equalization of injectors, especially in large engines and especially during operation under load or idle is possible, so that ultimately the goal can be achieved that the individual injectors essentially the same Inject amount of fuel. Thus, in the internal combustion engine and a

Pre-injection and / or post-injection can be displayed.

Claims

claims

Method for adjusting an injection behavior of injectors in an internal combustion engine, comprising the steps of: switching off an injector; Detecting a crank angle signal of the internal combustion engine; Transform the

Crank angle signal in the frequency domain by means of discrete Fourier transform; Detecting and storing a magnitude of the 0.5th-order harmonic of the Fourier transform of the crank angle signal, and allocating the magnitude to the shut-off injector; Switching on the switched-off injector;

Performing the foregoing steps sequentially for all injectors of the internal combustion engine; Forming an average of the stored amounts over all injectors, and correcting a control of the injectors by means of a

Deviation of the amount associated with an injector to be corrected from the mean.

A method according to claim 1, characterized in that a correction for an injector is performed only when the deviation exceeds a predetermined threshold.

Method according to one of the preceding claims, characterized in that a difference amount is calculated for each injector as a difference of the amount assigned to the injector to an amount which is detected and stored when all the injectors are turned on, wherein the difference amounts associated with the individual injectors Averaging and the correction.

4. The method according to any one of the preceding claims, characterized in that at least two iterations of the method are performed, wherein the method is preferably iterated until the deviation for each injector does not exceed the predetermined threshold.

5. The method according to any one of the preceding claims, characterized in that the control of the injectors is corrected so that the total power of the internal combustion engine is not changed thereby.

6. The method according to any one of the preceding claims, characterized in that a control of the injectors is corrected by a Bestromungsdauer is adjusted for them.

7. Process according to claims 3 and 6, characterized in that the

Bestromungsdauerdifferenz is calculated by the current current duration is calculated by the following formula: ABD [i] = (MW-Abetrag [i]) * K, where ABD [i] on the current Energization duration for an injector i energizing duration difference, MW the mean of the difference amounts of the harmonics of the 0.5th order for the individual injectors, ABetrag [i] of the determined for the injector i difference, and K is a constant, the formula preferably under the Condition IjABD [i] = 0 is applied.

8. The method according to claim 7, characterized in that the constant K

is selected depending on a currently present load point of the internal combustion engine.

9. engine control unit for an internal combustion engine, characterized in that the engine control unit for carrying out a method according to one of

Claims 1 to 8 is set up.

10. System for equalizing an injection behavior of injectors in one

Internal combustion engine, in particular for carrying out a method according to one of claims 1 to 8, with a switching means, designed for selectively switching off and on an injector; a detection means, designed for detection a crank angle signal of the internal combustion engine, wherein the detection means is preferably formed as a crankshaft sensor; a transformation means adapted to transform the crank angle signal into the frequency domain by means of discrete Fourier transformation; a storage means configured to detect and store a magnitude of the 0.5th-order harmonic of the Fourier transform of the crank angle signal and to associate that amount with the shut-off injector; an averaging element configured to calculate an average of the amounts stored in the storage means over all injectors, and a correction means configured to calculate a deviation of the amount associated with an injector to be corrected from the mean value and to correct an actuation of the injectors on the basis of the calculated deviation.