EP1693553B1 - Method for controlling an oil pump of an engine - Google Patents

Method for controlling an oil pump of an engine Download PDF

Info

Publication number
EP1693553B1
EP1693553B1 EP20060001076 EP06001076A EP1693553B1 EP 1693553 B1 EP1693553 B1 EP 1693553B1 EP 20060001076 EP20060001076 EP 20060001076 EP 06001076 A EP06001076 A EP 06001076A EP 1693553 B1 EP1693553 B1 EP 1693553B1
Authority
EP
European Patent Office
Prior art keywords
neural network
oil pressure
control
oil pump
engine
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
EP20060001076
Other languages
German (de)
French (fr)
Other versions
EP1693553A2 (en
EP1693553A3 (en
Inventor
Peter Scholl
Oliver Grunwald
Hubert Dr. Keller
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
KELLER, HUBERT B.
Dr Ing HCF Porsche AG
Original Assignee
Dr Ing HCF Porsche AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Dr Ing HCF Porsche AG filed Critical Dr Ing HCF Porsche AG
Publication of EP1693553A2 publication Critical patent/EP1693553A2/en
Publication of EP1693553A3 publication Critical patent/EP1693553A3/en
Application granted granted Critical
Publication of EP1693553B1 publication Critical patent/EP1693553B1/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F01—MACHINES OR ENGINES IN GENERAL; ENGINE PLANTS IN GENERAL; STEAM ENGINES
    • F01M—LUBRICATING OF MACHINES OR ENGINES IN GENERAL; LUBRICATING INTERNAL COMBUSTION ENGINES; CRANKCASE VENTILATING
    • F01M1/00—Pressure lubrication
    • F01M1/16—Controlling lubricant pressure or quantity

Definitions

  • the invention relates to a device for controlling the lubricating oil pressure of an internal combustion engine, as for example from the DE 101 41 786 A1 is known.
  • this known device for controlling the lubricating oil pressure of an internal combustion engine of an oil pump is associated with an oil pressure control valve, the valve body is acted upon by a front side by the oil pressure in an opening direction, wherein the opening characteristic of the oil pressure control valve in response to various operating parameters and operating conditions is adjustable.
  • the opening characteristic of the valve can be changed in a simple manner by applying oil pressure to the piston rear side via a hydraulic line leading from the pressure side to the rear side of the piston.
  • the inventive method has the advantage that by an intelligent control of the oil pressure with the involvement of an intelligent mitvertenden structure, the control of the oil pump a variety of factors such as the speed, the instantaneous speed curve, the oil viscosity, the oil temperature, manufacturing tolerances within the oil pump and the engine and considered possible leakage due to wear and aging of the engine.
  • an intelligent learning structure By using an intelligent learning structure, a high quality of control with low computing power is guaranteed.
  • a neural network such as a Kohonen network
  • the control of the oil pump valve motor unit via a neural network offers the advantage of a self-learning algorithm with the ability to make permanent corrections in the nodes of such a network, to make the input area multi-dimensional and allow interpolation over several manipulated variables, creating a certain smoothness in the control is given and no jumps in the control of the oil pump, which have a negative effect on the behavior occur.
  • the use of neural networks to control various methods is already known in principle, as for example the DE 43 33 698 A1 can be seen.
  • the advantage of the present method is to integrate this methodology of processing and determining various control variables in the engine control and in the control of the engine oil pump so that there is a technical improvement of the behavior during driving.
  • the method for controlling an engine oil pump by incorporating an intelligent mitvertenden structure also has the advantage that the change in the parameters can be logged during operation.
  • the detection of state changes is to be evaluated in terms of a diagnosis and the quality of the control process is improved.
  • the inventive method for controlling an oil pump with an intelligent mitlindenden structure for controlling the oil pump consists essentially of two phases. A first phase, in which the intelligent learning structure is determined, and a second phase, in which this learning structure is implemented in the control process in the vehicle.
  • FIG. 1 shows the first phase, which can also be called the preparation phase or training phase.
  • This training phase for the intelligent learning structure is carried out on a computer and not directly on the vehicle.
  • input variables 10 such as the rotational speed n, the rotational speed difference ⁇ n, the oil temperature T (oil) and the engine torque Mm, are entered from predefined reference measurements. These various operating parameters were recorded in advance in the operation of an internal combustion engine, for example on a test bench or by measurements during various journeys and stored as representative quantities. They indicate the driving behavior of the internal combustion engine under various conditions.
  • the input variables 10 are fed to a neural network 11.
  • the neural network 11 is in an initial state before it is calculated for use in the vehicle.
  • the Kohonen network 11 outputs, after a corresponding evaluation and weighting of the input variables 10, an output quantity AG which is fed to the input of the MLP network 12.
  • the MLP network 12 simulates the behavior of the internal combustion engine, so that at the output of this MLP network 12 is a size for the corresponding actual oil pressure is available, as in a real operation of an internal combustion engine with the corresponding inputs and the neural network 11 output size for the control of the engine oil pump, would occur.
  • This quantity of the actual oil pressure is compared with the desired oil pressure to be taken from the characteristic map 13 on the basis of the corresponding operating variables. This is in the FIG. 1 with the Processing stage 14 shown.
  • an adaptation of the Kohonen network 12 is performed.
  • This adaptation essentially represents an iteration method, in which, by the feedback of the reactions to the previously output output variable AG, the Kohonen network 12 learns to react to changes in the operation of the internal combustion engine.
  • the neural network is an intelligent learning structure whose points of support are not even defined and then adapted by applying a characteristic field, but the neural network itself learns and outputs directly the optimal control variable.
  • This learning or training phase is the basis for the later implementation of this neural network 12 in the control of the engine oil pump.
  • FIG. 2 It is shown how the control of the engine oil pump works using the neural network 11.
  • the input variables 20 essentially correspond to the input variables 10 FIG. 1 , Only these input variables 20 are not measured here by a computer, but directly by the sensors of the internal combustion engine, not shown, and provided for example via a data bus 21 to the method for controlling the engine oil pump.
  • the data supplied by the data bus 21 are processed in a preparation unit 22, so that they are then forwarded directly as input variables 20 to the Kohonen network 11 and to the characteristic map 13 for determining the desired oil pressure.
  • the Kohonen network 11 determines the control variable SG, which is then output for controlling the engine oil pump, on the basis of the operating parameters available as input variables 20.
  • the control of the engine oil pump in turn influences the operation of the internal combustion engine and the actual oil pressure.
  • an adaptation of the neural network takes place analogously to the training phase, so that this neural network is also constantly learning in actual operation Medium represents.
  • the actual calculation within the neural network concerns already known computational algorithms and methodologies.
  • Essential to the invention is the integration of a known neural network in the processes within the control of an internal combustion engine for controlling the engine oil pump.

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Lubrication Of Internal Combustion Engines (AREA)
  • Lubrication Details And Ventilation Of Internal Combustion Engines (AREA)

Description

Die Erfindung geht aus von einer Einrichtung zur Regelung des Schmieröldruckes einer Brennkraftmaschine, wie sie beispielsweise aus der DE 101 41 786 A1 bekannt ist.The invention relates to a device for controlling the lubricating oil pressure of an internal combustion engine, as for example from the DE 101 41 786 A1 is known.

Bei dieser bekannten Einrichtung zur Regelung des Schmieröldruckes einer Brennkraftmaschine ist einer Ölpumpe ein Öldruckregelventil zugeordnet, dessen Ventilkörper von einer Vorderseite durch den Öldruck in eine Öffnungsrichtung beaufschlagt wird, wobei die Öffnungscharakteristik des Öldruckregelventils in Abhängigkeit verschiedener Betriebsparameter und Betriebszustände einstellbar ist. Die Öffnungscharakteristik des Ventils kann auf einfache Art und Weise dadurch geändert werden, dass über eine von der Druckseite auf die Rückseite des Kolbens führende Hydraulikleitung Öldruck auf die Kolbenrückseite aufgebracht wird. Durch eine entsprechende Steuerung des über die erhöhte Hydraulikleitung in den Gehäuseraum des Kolbenventils geführten Schmieröls können entsprechend den verschiedenen Betriebszuständen des Motors unterschiedliche Öldrücke eingestellt werden.In this known device for controlling the lubricating oil pressure of an internal combustion engine of an oil pump is associated with an oil pressure control valve, the valve body is acted upon by a front side by the oil pressure in an opening direction, wherein the opening characteristic of the oil pressure control valve in response to various operating parameters and operating conditions is adjustable. The opening characteristic of the valve can be changed in a simple manner by applying oil pressure to the piston rear side via a hydraulic line leading from the pressure side to the rear side of the piston. By a corresponding control of the guided over the elevated hydraulic line in the housing space of the piston valve lubricating oil different oil pressures can be adjusted according to the different operating conditions of the engine.

Aus der US 6,269,788 B1 ist eine Ölpumpe für eine Brennkraftmaschine bekannt, die von einer programmierbaren Recheneinheit angesteuert wird. Die programmierbare Recheneinheit verarbeitet verschiedene Eingangsgrößen nach einem vorgegebenen Algorithmus.From the US 6,269,788 B1 an oil pump for an internal combustion engine is known, which is controlled by a programmable arithmetic unit. The programmable arithmetic unit processes various input variables according to a predetermined algorithm.

Das erfindungsgemäße Verfahren hat den Vorteil, dass durch eine intelligente Regelung des Öldruckes unter Einbindung einer intelligenten mitlernenden Struktur die Ansteuerung der Ölpumpe die verschiedensten Faktoren, wie die Drehzahl, der momentane Drehzahlverlauf, die Ölviskosität, die Öltemperatur, Fertigungstoleranzen innerhalb der Ölpumpe und des Motors und eine mögliche Leckage durch Verschleiß und Alterung des Motors berücksichtigt. Durch das Nutzen einer intelligenten mitlernenden Struktur ist eine hohe Regelgüte mit geringer Rechenleistung gewährleistet.The inventive method has the advantage that by an intelligent control of the oil pressure with the involvement of an intelligent mitlernenden structure, the control of the oil pump a variety of factors such as the speed, the instantaneous speed curve, the oil viscosity, the oil temperature, manufacturing tolerances within the oil pump and the engine and considered possible leakage due to wear and aging of the engine. By using an intelligent learning structure, a high quality of control with low computing power is guaranteed.

Es hat sich ferner als vorteilhaft erwiesen, als mitlernende intelligente Struktur ein neuronales Netz, wie beispielsweise ein Kohonen-Netz einzusetzen. Die Ansteuerung der Ölpumpe-Ventil-Motor-Einheit über ein neuronales Netz bietet den Vorteil eines selbstlernenden Algorithmus mit der Möglichkeit, permanent Korrekturen in den Stützstellen eines solchen Netzes vorzunehmen, den Eingangsbereich mehrdimensional zu gestalten und eine Interpolation über mehrere Stellgrößen zuzulassen, wodurch eine gewisse Glattheit in der Ansteuerung gegeben ist und keine Sprünge bei der Ansteuerung der Ölpumpe, die sich negativ auf das Verhalten auswirken, auftreten.It has also proven to be advantageous to use a neural network, such as a Kohonen network, as a learning intelligent structure. The control of the oil pump valve motor unit via a neural network offers the advantage of a self-learning algorithm with the ability to make permanent corrections in the nodes of such a network, to make the input area multi-dimensional and allow interpolation over several manipulated variables, creating a certain smoothness in the control is given and no jumps in the control of the oil pump, which have a negative effect on the behavior occur.

Die Verwendung neuronaler Netze zur Steuerung verschiedener Verfahren ist prinzipiell bereits bekannt, wie dies beispielsweise der DE 43 33 698 A1 zu entnehmen ist. Der Vorteil bei dem vorliegenden Verfahren besteht darin, diese Methodik der Verarbeitung und Bestimmung verschiedener Steuergrößen so in die Motorsteuerung und in die Ansteuerung der Motorölpumpe zu integrieren, dass sich eine technische Verbesserung des Verhaltens im Fahrbetrieb ergibt.The use of neural networks to control various methods is already known in principle, as for example the DE 43 33 698 A1 can be seen. The advantage of the present method is to integrate this methodology of processing and determining various control variables in the engine control and in the control of the engine oil pump so that there is a technical improvement of the behavior during driving.

Das Verfahren zur Ansteuerung einer Motorölpumpe unter Einbindung einer intelligenten mitlernenden Struktur hat ferner den Vorteil, dass die Veränderung der Parameter im laufenden Betrieb protokolliert werden kann. Das Erfassen von Zustandsänderungen ist im Sinne einer Diagnose auszuwerten und die Qualität des Regelprozesses wird verbessert.The method for controlling an engine oil pump by incorporating an intelligent mitlernenden structure also has the advantage that the change in the parameters can be logged during operation. The detection of state changes is to be evaluated in terms of a diagnosis and the quality of the control process is improved.

Ein Ausführungsbeispiel der Erfindung ist in der Zeichnung dargestellt und in der nachfolgenden Beschreibung näher erläutert.
Es zeigt

Fig. 1
die Trainingsphase zum Anlernen des Kohonen-Netzes,
Fig. 2
das im Fahrzeug implementierte Regelverfahren mit dem integrierten Kohonen-Netz.
An embodiment of the invention is illustrated in the drawing and explained in more detail in the following description.
It shows
Fig. 1
the training phase for learning the Kohonen network,
Fig. 2
the implemented in the vehicle control method with the integrated Kohonen network.

Das erfindungsgemäße Verfahren zur Regelung einer Ölpumpe mit einer intelligenten mitlernenden Struktur zur Regelung der Ölpumpe besteht im Wesentlichen aus zwei Phasen. Einer ersten Phase, in welcher die intelligente mitlernende Struktur bestimmt wird, und einer zweiten Phase, in welcher diese mitlernende Struktur in das Regelverfahren im Fahrzeug implementiert wird.The inventive method for controlling an oil pump with an intelligent mitlernenden structure for controlling the oil pump consists essentially of two phases. A first phase, in which the intelligent learning structure is determined, and a second phase, in which this learning structure is implemented in the control process in the vehicle.

Figur 1 zeigt die erste Phase, welche auch als Vorbereitungsphase oder Trainingsphase bezeichnet werden kann.
Diese Trainingsphase für die intelligente mitlernende Struktur wird auf einem Rechner und nicht direkt am Fahrzeug durchgeführt. Hierzu werden von vordefinierten Referenzmessungen verschiedenen Eingangsgrößen 10, wie die Drehzahl n, die Drehzahldifferenz Δn, die Öltemperatur T(öl) und das Motormoment Mm, eingegeben. Diese verschiedenen Betriebsparameter wurden im Vorfeld beim Betrieb einer Brennkraftmaschine beispielsweise auf einem Prüfstand oder durch Messungen während verschiedenster Fahrten erfasst und als repräsentative Größen abgespeichert. Sie geben das Fahrverhalten der Brennkraftmaschine unter den verschiedensten Bedingungen an. Wie der Figur 1 zu entnehmen, werden die Eingangsgrößen 10 an ein neuronales Netz 11 geführt. In dieser in Figur 1 dargestellten Trainingsphase befindet sich das neuronale Netz 11 in einem Ausgangszustand, bevor es für den Einsatz im Fahrzeug berechnet wird.
Parallel zu der Weiterleitung der Betriebsparameter 10 an das neuronale Netz 11 sind diese Eingangsdaten an ein MLP-Netz 12 geführt, welches den Betrieb der Brennkraftmaschine simuliert, und an ein Kennfeld 13 für den Soll-Öldruck geführt.
Als neuronales Netz 11 wird bei diesem Verfahren ein Kohonen-Netz verwendet, wobei ein solches Kohonen-Netz bereits bekannt ist (vgl. "Maschinelle Intelligenz" Hubert B. Keller, ISBN 35280548911) und hier nicht im einzelnen erläutert werden soll. FIG. 1 shows the first phase, which can also be called the preparation phase or training phase.
This training phase for the intelligent learning structure is carried out on a computer and not directly on the vehicle. For this purpose, input variables 10, such as the rotational speed n, the rotational speed difference Δn, the oil temperature T (oil) and the engine torque Mm, are entered from predefined reference measurements. These various operating parameters were recorded in advance in the operation of an internal combustion engine, for example on a test bench or by measurements during various journeys and stored as representative quantities. They indicate the driving behavior of the internal combustion engine under various conditions. Again FIG. 1 can be seen, the input variables 10 are fed to a neural network 11. In this in FIG. 1 shown training phase, the neural network 11 is in an initial state before it is calculated for use in the vehicle.
Parallel to the forwarding of the operating parameters 10 to the neural network 11, these input data are fed to an MLP network 12, which simulates the operation of the internal combustion engine, and guided to a map 13 for the desired oil pressure.
As a neural network 11, a Kohonen network is used in this method, such Kohonen network is already known (see "Machine Intelligence" Hubert B. Keller, ISBN 35280548911) and will not be explained in detail here.

Das Kohonen-Netz 11 gibt nach einer entsprechenden Bewertung und Wichtung der Eingangsgrößen 10 eine Ausgangsgröße AG aus, die an den Eingang des MLP-Netzes 12 geführt ist. Das MLP-Netz 12 simuliert das Verhalten der Brennkraftmaschine, so dass am Ausgang dieses MLP-Netzes 12 eine Größe für den entsprechenden Ist-Öldruck zur Verfügung steht, wie er bei einem realen Betrieb einer Brennkraftmaschine mit den entsprechenden Eingangsgrößen und der von dem neuronalen Netz 11 ausgegebenen Größe für die Ansteuerung der Motorölpumpe, auftreten würde. Diese Größe des Ist-Öldruckes wird mit dem aufgrund der entsprechenden Betriebsgrößen aus dem Kennfeld 13 zu entnehmenden Soll-Öldruck verglichen. Dies ist in der Figur 1 mit der Verarbeitungsstufe 14 dargestellt. Aufgrund der in der Verarbeitungsstufe 14 ermittelten Abweichung zwischen Ist-Öldruck und Soll-Öldruck wird eine Adaption des Kohonen-Netzes 12 durchgeführt. Diese Adaption stellt im Wesentlichen ein Iterationsverfahren dar, in welchem durch die Rückkopplung der Reaktionen auf die vorher ausgegebene Ausgangsgröße AG das Kohonen-Netzes 12 lernt, auf Veränderungen im Betrieb der Brennkraftmaschine zu reagieren. Nach der Trainingsphase ist das neuronale Netz eine intelligente mitlernende Struktur, dessen Stützstellen nicht einmal festgelegt und dann durch Beaufschlagen eines Kennfeldes dieses adaptieren, sondern das neuronale Netz selber lernt und gibt jeweils direkt die optimale Steuergröße aus. Diese Lern- oder Trainingsphase ist die Grundlage für die spätere Implementierung dieses neuronalen Netzes 12 in die Steuerung der Motorölpumpe.The Kohonen network 11 outputs, after a corresponding evaluation and weighting of the input variables 10, an output quantity AG which is fed to the input of the MLP network 12. The MLP network 12 simulates the behavior of the internal combustion engine, so that at the output of this MLP network 12 is a size for the corresponding actual oil pressure is available, as in a real operation of an internal combustion engine with the corresponding inputs and the neural network 11 output size for the control of the engine oil pump, would occur. This quantity of the actual oil pressure is compared with the desired oil pressure to be taken from the characteristic map 13 on the basis of the corresponding operating variables. This is in the FIG. 1 with the Processing stage 14 shown. Due to the deviation between the actual oil pressure and the target oil pressure determined in the processing stage 14, an adaptation of the Kohonen network 12 is performed. This adaptation essentially represents an iteration method, in which, by the feedback of the reactions to the previously output output variable AG, the Kohonen network 12 learns to react to changes in the operation of the internal combustion engine. After the training phase, the neural network is an intelligent learning structure whose points of support are not even defined and then adapted by applying a characteristic field, but the neural network itself learns and outputs directly the optimal control variable. This learning or training phase is the basis for the later implementation of this neural network 12 in the control of the engine oil pump.

In Figur 2 ist dargestellt, wie die Ansteuerung der Motorölpumpe unter Nutzung des neuronalen Netzes 11 arbeitet. Die Eingangsgrößen 20 entsprechen im Wesentlichen den Eingangsgrößen 10 aus Figur 1. Nur werden diese Eingangsgrößen 20 hier nicht von einem Rechner, sondern direkt von den nicht dargestellten Sensoren der Brennkraftmaschine gemessen und beispielsweise über einen Datenbus 21 dem Verfahren zur Ansteuerung der Motorölpumpe zur Verfügung gestellt. Zur Vorbereitung werden die vom Datenbus 21 gelieferten Daten in einer Vorbereitungseinheit 22 aufbereitet, so dass sie anschließend direkt als Eingangsgrößen 20 an das Kohonen-Netz 11 und an das Kennfeld 13 zur Bestimmung des Soll-Öldruck weitergeleitet werden. Das Kohonen-Netz 11 bestimmt aufgrund der als Eingangsgrößen 20 zur Verfügung stehenden Betriebsparameter die Steuergröße SG, die dann zur Ansteuerung der Motorölpumpe ausgegeben wird.
Die Ansteuerung der Motorölpumpe beeinflusst wiederum den Betrieb der Brennkraftmaschine und den Ist-Öldruck. Durch die Erfassung des Ist-Ölsruckes und einen Vergleich des Ist-Öldruckes mit dem Soll-Öldruck in der Vergleichsstufe 14 erfolgt analog zur Trainingsphase eine Adaption des neuronalen Netzes (Kohonen-Netz), so dass dieses neuronale Netz auch im eigentlichen Betrieb ein ständig mitlernendes Medium darstellt.In FIG. 2 It is shown how the control of the engine oil pump works using the neural network 11. The input variables 20 essentially correspond to the input variables 10 FIG. 1 , Only these input variables 20 are not measured here by a computer, but directly by the sensors of the internal combustion engine, not shown, and provided for example via a data bus 21 to the method for controlling the engine oil pump. For preparation, the data supplied by the data bus 21 are processed in a preparation unit 22, so that they are then forwarded directly as input variables 20 to the Kohonen network 11 and to the characteristic map 13 for determining the desired oil pressure. The Kohonen network 11 determines the control variable SG, which is then output for controlling the engine oil pump, on the basis of the operating parameters available as input variables 20.
The control of the engine oil pump in turn influences the operation of the internal combustion engine and the actual oil pressure. By detecting the actual oil pressure and comparing the actual oil pressure with the desired oil pressure in the comparison stage 14, an adaptation of the neural network (Kohonen network) takes place analogously to the training phase, so that this neural network is also constantly learning in actual operation Medium represents.

Wesentlich ist sowohl in der Trainingsphase als auch bei der eigentlichen Umsetzung des Verfahrens zur Ansteuerung einer Motorölpumpe, dass eine Rückinformation nach einer Steuergrößenausgabe über eine Erfassung des Ist-Öldruckes und des Soll-Öldruckes erfolgt.It is essential both in the training phase and in the actual implementation of the method for controlling an engine oil pump, that a return information after a control variable output via a detection of the actual oil pressure and the target oil pressure.

Die eigentliche Berechnung innerhalb des neuronalen Netzes betrifft bereits bekannte Rechenalgorithmen und Methodiken. Erfindungswesentlich ist die Einbindung eines an sich bekannten neuronalen Netzes in die Vorgänge innerhalb der Ansteuerung einer Brennkraftmaschine zur Ansteuerung der Motorölpumpe.The actual calculation within the neural network concerns already known computational algorithms and methodologies. Essential to the invention is the integration of a known neural network in the processes within the control of an internal combustion engine for controlling the engine oil pump.

Die vorstehende Beschreibung führt aus, dass die intelligente mitlernende Struktur während der Trainingsphase angelernt und dann in die Steuereinheit implementiert ist. Es ist jedoch genauso möglich, diese Trainingsphase im laufenden Betrieb der Brennkraftmaschine durchzuführen. Die verschiedenen Eingangsgrößen zur Bestimmung der Ansteuerung des Motorölpumpe sowie die dazu gehörende Reaktion der Brennkraftmaschine sind bekannt, so dass mit diesen Größen ein Nachlernen des neuronalen Netzes durchgeführt werden kann. Eine Protokollierung dieses Nachlernprozessen ermöglicht es außerdem, Veränderungen zu diagnostizieren um so Rückschlüsse auf beispielsweise Alterungsprozesse und/oder Defekte der einzelnen Komponenten zu erfassen.The above description states that the intelligent learning structure is taught during the training phase and then implemented in the control unit. However, it is equally possible to perform this training phase during operation of the internal combustion engine. The various input variables for determining the control of the engine oil pump and the associated reaction of the internal combustion engine are known, so that with these variables a re-learning of the neural network can be performed. A logging of these Nachlernprozessen also makes it possible to diagnose changes so as to draw conclusions about, for example, aging processes and / or defects of the individual components.

Claims (5)

  1. Method for controlling an engine oil pump of an internal combustion engine which has at least one control unit to which the operating parameters (20) which are detected by sensors are fed, wherein a neural network (11) for determining the control variable to be output to the engine oil pump is implemented in the control unit, wherein the neural network is trained in a training phase during the application before its implementation in the control unit to the effect that it constitutes an intelligent simultaneously learning structure which, during the controlling of the engine oil pump during the driving mode, closes a control loop which adapts the control variable (SG) by feeding back the output control variable and detecting the actual oil pressure and carrying out a comparison with a predetermined setpoint oil pressure in such a way that in each case the control variable which is to be output at a particular time is continuously recalculated in the neural network (11).
  2. Method according to Claim 1, characterized in that the control variables (SG) which are to be output are not permanently stored adaptation values.
  3. Method according to Claim 1, characterized in that a Kohonen network is used as the neural network.
  4. Method according to Claim 1, characterized in that during the training phase the neural network is connected to a further neural network for simulating the operation of an internal combustion engine, and this further neural network outputs a value for the actual oil pressure to be expected, wherein the actual oil pressure is compared with the setpoint oil pressure which is obtained from a characteristic diagram (13), and the result of this evaluation is fed to the neural network, with the result that the neural network actively learns to react to changes.
  5. Method according to Claim 1, characterized in that the adaptation of the control variable (SG) which is recalculated in each case on an up-to-date basis in the neural network (11) is logged and can be read out for diagnostics.
EP20060001076 2005-02-18 2006-01-19 Method for controlling an oil pump of an engine Expired - Fee Related EP1693553B1 (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
DE200510007406 DE102005007406A1 (en) 2005-02-18 2005-02-18 Method for controlling an engine oil pump

Publications (3)

Publication Number Publication Date
EP1693553A2 EP1693553A2 (en) 2006-08-23
EP1693553A3 EP1693553A3 (en) 2010-05-26
EP1693553B1 true EP1693553B1 (en) 2011-07-27

Family

ID=36369911

Family Applications (1)

Application Number Title Priority Date Filing Date
EP20060001076 Expired - Fee Related EP1693553B1 (en) 2005-02-18 2006-01-19 Method for controlling an oil pump of an engine

Country Status (2)

Country Link
EP (1) EP1693553B1 (en)
DE (1) DE102005007406A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
ES2775249T3 (en) * 2014-07-01 2020-07-24 Fpt Motorenforschung Ag Lubricating oil system for a combustion engine, in particular for industrial and commercial vehicles
DE102019220501A1 (en) * 2019-12-21 2021-06-24 Robert Bosch Gesellschaft mit beschränkter Haftung Method for controlling a hydraulic cylinder of a work machine
JP7424328B2 (en) * 2021-02-26 2024-01-30 トヨタ自動車株式会社 Transmission control device
CN113464235B (en) * 2021-06-16 2023-03-14 东风汽车集团股份有限公司 Oil pump becomes row device and engine
DE102022209422A1 (en) 2022-09-09 2024-03-14 Volkswagen Aktiengesellschaft Operating method for operating an oil pump control valve

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AT1265U1 (en) * 1995-07-19 1997-01-27 Matthias Budil LEARNING RULE FOR NEURONAL NETWORKS
JPH11343916A (en) * 1998-06-02 1999-12-14 Yamaha Motor Co Ltd Data estimating method in engine control
AT407563B (en) * 1998-02-26 2001-04-25 Tcg Unitech Ag OIL PUMP FOR AN INTERNAL COMBUSTION ENGINE
DK173533B1 (en) * 1999-01-18 2001-02-05 Man B & W Diesel As Method of lubricating a cylinder in an internal combustion engine as well as cylinder lubrication system and connecting element
DE19915737A1 (en) * 1999-04-08 2000-10-12 Bayerische Motoren Werke Ag Method for regulating the lubrication, preferably in internal combustion engines and arrangement for regulating according to the method
US6269788B1 (en) 2000-03-13 2001-08-07 Robert L. Kachelek Programmable computer controlled electric oil pump drive for engines
DE10113538B4 (en) * 2001-03-20 2012-03-01 Bayerische Motoren Werke Aktiengesellschaft Regulating device and control method
DE10141786B4 (en) 2001-08-25 2008-12-24 Dr. Ing. H.C. F. Porsche Aktiengesellschaft Device for controlling the lubricating oil pressure of an internal combustion engine

Also Published As

Publication number Publication date
EP1693553A2 (en) 2006-08-23
EP1693553A3 (en) 2010-05-26
DE102005007406A1 (en) 2006-08-24

Similar Documents

Publication Publication Date Title
EP1693553B1 (en) Method for controlling an oil pump of an engine
DE102016117560B4 (en) TOOL MACHINE FOR PRODUCING A SPEED DISTRIBUTION
DE102016008994A1 (en) TOOLING MACHINE, SIMULATING DEVICE AND MACHINE INCOMING DEVICE
DE112008001457T5 (en) Torque control with feedback control
DE102011103594A1 (en) Method for controlling technical processes and methods for carrying out tests on test benches
DE3400711C2 (en)
DE19524669C2 (en) Control and regulating device for commissioning a vehicle with an automotive hydrostatic transmission
EP3713813B1 (en) Method for configuring and controlling a steering system, steering system, and vehicle
DE102007006616B3 (en) Electronically controlled, automatically shifted gearbox optimizing method for e.g. car, involves determining weighting factor for quality of shifting process from recognition parameter
EP1095319B1 (en) Method for operating a control system and device for carrying out said method
DE3932763C1 (en)
DE3931143C2 (en) Method for monitoring the operation of a machine tool
DE4005973A1 (en) DIAGNOSTIC METHOD FOR CHECKING ACTUATORS FOR CONTROLLING INTERNAL COMBUSTION ENGINES
DE102019001177A1 (en) Controller, machine learning device and system
DE10035005A1 (en) Method and appliance for automatic control of gear ratio with continuously variable ratio change involves use of family of performance characteristics obtained from tests
EP1262760A1 (en) Method for determining the viscosity of a working fluid of a combustion engine
EP2366567B1 (en) Method for controlling a dampener for a motor vehicle
WO2015135995A1 (en) Motor vehicle, control unit and method for controlling a phase angle of a camshaft
DE10351957B4 (en) Damping system with a LOLIMOT model against drive train vibrations
DE19807763A1 (en) Phase adjustment setting device for i.c. engine camshaft
EP3292444A1 (en) Method for operating an electric or fluidic actuator and movement controller for an actuator
DE60302429T2 (en) Method and device for determining the operating state in a Brenkraftmaschine with cam profile change
DE112012003139T5 (en) A method of controlling a vehicle steering apparatus
DE10312840B4 (en) Adaptive controller and method for determining its parameters
EP3489773A1 (en) Method for the computer-based control of a technical system, in particular an energy generation plant

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A2

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU LV MC NL PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA HR MK YU

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: KELLER, HUBERT B.

Owner name: DR. ING. H.C. F. PORSCHE AKTIENGESELLSCHAFT

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: KELLER, HUBERT B.

Owner name: DR. ING. H.C. F. PORSCHE AKTIENGESELLSCHAFT

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: KELLER, HUBERT B.

Owner name: DR. ING. H.C. F. PORSCHE AG

PUAL Search report despatched

Free format text: ORIGINAL CODE: 0009013

AK Designated contracting states

Kind code of ref document: A3

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU LV MC NL PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA HR MK YU

RIC1 Information provided on ipc code assigned before grant

Ipc: F02D 41/14 20060101ALI20100419BHEP

Ipc: F01M 1/16 20060101AFI20060524BHEP

17P Request for examination filed

Effective date: 20101126

AKX Designation fees paid

Designated state(s): DE FR GB IT

RIC1 Information provided on ipc code assigned before grant

Ipc: F02D 41/14 20060101ALI20110111BHEP

Ipc: F01M 1/16 20060101AFI20110111BHEP

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): DE FR GB IT

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

Free format text: NOT ENGLISH

REG Reference to a national code

Ref country code: DE

Ref legal event code: R096

Ref document number: 502006009878

Country of ref document: DE

Effective date: 20110915

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20120502

REG Reference to a national code

Ref country code: DE

Ref legal event code: R097

Ref document number: 502006009878

Country of ref document: DE

Effective date: 20120502

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 11

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 12

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 13

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: GB

Payment date: 20180119

Year of fee payment: 13

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: IT

Payment date: 20180129

Year of fee payment: 13

Ref country code: FR

Payment date: 20180119

Year of fee payment: 13

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20190119

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20190131

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20190119

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IT

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20190119

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20200115

Year of fee payment: 15

REG Reference to a national code

Ref country code: DE

Ref legal event code: R119

Ref document number: 502006009878

Country of ref document: DE

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20210803