EP1041264A2 - Hybrid model for the modelling of a whole process in a vehicle - Google Patents

Hybrid model for the modelling of a whole process in a vehicle Download PDF

Info

Publication number
EP1041264A2
EP1041264A2 EP00106509A EP00106509A EP1041264A2 EP 1041264 A2 EP1041264 A2 EP 1041264A2 EP 00106509 A EP00106509 A EP 00106509A EP 00106509 A EP00106509 A EP 00106509A EP 1041264 A2 EP1041264 A2 EP 1041264A2
Authority
EP
European Patent Office
Prior art keywords
model
physical
hybrid
neural
simulated
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.)
Ceased
Application number
EP00106509A
Other languages
German (de)
French (fr)
Other versions
EP1041264A3 (en
Inventor
Heiko Dr. Konrad
Gerd Krämer
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.)
Bayerische Motoren Werke AG
Original Assignee
Bayerische Motoren Werke 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 Bayerische Motoren Werke AG filed Critical Bayerische Motoren Werke AG
Publication of EP1041264A2 publication Critical patent/EP1041264A2/en
Publication of EP1041264A3 publication Critical patent/EP1041264A3/en
Ceased legal-status Critical Current

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F01—MACHINES OR ENGINES IN GENERAL; ENGINE PLANTS IN GENERAL; STEAM ENGINES
    • F01L—CYCLICALLY OPERATING VALVES FOR MACHINES OR ENGINES
    • F01L1/00—Valve-gear or valve arrangements, e.g. lift-valve gear
    • F01L1/34—Valve-gear or valve arrangements, e.g. lift-valve gear characterised by the provision of means for changing the timing of the valves without changing the duration of opening and without affecting the magnitude of the valve lift
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F01—MACHINES OR ENGINES IN GENERAL; ENGINE PLANTS IN GENERAL; STEAM ENGINES
    • F01L—CYCLICALLY OPERATING VALVES FOR MACHINES OR ENGINES
    • F01L9/00—Valve-gear or valve arrangements actuated non-mechanically
    • F01L9/20—Valve-gear or valve arrangements actuated non-mechanically by electric means
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D41/00—Electrical control of supply of combustible mixture or its constituents
    • F02D41/02—Circuit arrangements for generating control signals
    • F02D41/14—Introducing closed-loop corrections
    • F02D41/1401—Introducing closed-loop corrections characterised by the control or regulation method
    • F02D41/1405—Neural network control
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F01—MACHINES OR ENGINES IN GENERAL; ENGINE PLANTS IN GENERAL; STEAM ENGINES
    • F01L—CYCLICALLY OPERATING VALVES FOR MACHINES OR ENGINES
    • F01L2800/00—Methods of operation using a variable valve timing mechanism
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D13/00—Controlling the engine output power by varying inlet or exhaust valve operating characteristics, e.g. timing
    • F02D13/02—Controlling the engine output power by varying inlet or exhaust valve operating characteristics, e.g. timing during engine operation
    • F02D13/0203—Variable control of intake and exhaust valves
    • F02D13/0215—Variable control of intake and exhaust valves changing the valve timing only
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D41/00—Electrical control of supply of combustible mixture or its constituents
    • F02D41/0002—Controlling intake air
    • F02D2041/001—Controlling intake air for engines with variable valve actuation
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D41/00—Electrical control of supply of combustible mixture or its constituents
    • F02D41/02—Circuit arrangements for generating control signals
    • F02D41/14—Introducing closed-loop corrections
    • F02D41/1401—Introducing closed-loop corrections characterised by the control or regulation method
    • F02D2041/1433—Introducing closed-loop corrections characterised by the control or regulation method using a model or simulation of the system
    • F02D2041/1436—Hybrid model
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D2200/00—Input parameters for engine control
    • F02D2200/02—Input parameters for engine control the parameters being related to the engine
    • F02D2200/04—Engine intake system parameters
    • F02D2200/0402—Engine intake system parameters the parameter being determined by using a model of the engine intake or its components

Definitions

  • the invention relates to a hybrid model for modeling an overall process in a vehicle consisting of at least one physical and one neural sub-model.
  • the filling of cylinders in engines with variable valve train measured with a very delayed air mass sensor. It will therefore expediently from different input variables, which are directly at the inlet be measured and determined with the help of a model.
  • the Filling of the individual cylinders influenced by several manipulated variables, some of them are interdependent or independent.
  • Empirical methods such as Maps.
  • empirical methods are usually imprecise and require a high level Coordination effort.
  • Another possibility are physical functions, at which the process behavior from the consideration of the physical relationships is derived.
  • physical functions are sometimes difficult to create.
  • the overall system and the Dependencies to be known within the system.
  • the effort for the creation of physical models with increasing model complexity disproportionately too.
  • different concepts e.g. Direct injection, electronic valve train, variable valve train, etc.
  • DE 197 06 750 A1 describes a method for controlling the mixture in a Internal combustion engine and a device for performing this method known.
  • the Combustion chamber of the internal combustion engine air mass coming from a Input size determined.
  • the amount of fuel to be supplied in Determined as a function of this input variable.
  • the neural network is used to describe the Control variable for the fuel path depending on the engine operating state and the driver-influenced control variable for the air path.
  • the control variable for the fuel path is exclusive in this embodiment set on the neural network.
  • neural networks are outside the Work area in which the training data are determined, an implausible Can have extrapolation behavior and therefore in safety-critical Processes, e.g. in motor vehicles, are difficult to use.
  • the object of the present invention is to develop a hybrid model for modeling a To specify the overall process in a vehicle, with which physical have difficult to describe processes modeled without the implausible Extrapolation behavior must be accepted.
  • the overall process (for example the filling of the Cylinder) is divided into sub-processes, which are of different sub-models described and then combined into an overall model.
  • the neural model takes over the description of a process part, which is physical is difficult to grasp.
  • the modeling of the air mass filling can be used as a concrete application Specify internal combustion engines, for example with variable valve train. At this Application could determine the basic filling using a physical model become. However, the influence of camshaft spreading could neural network are described. Especially when describing the Influence of camshaft spreading is only possible with a high physical model Create effort.
  • the modeling of the basic model with a physical process description has the advantage that the share of the neural sub-model in the overall model can be deliberately restricted. This ensures that Overall model shows no implausible extrapolation behavior.
  • the merging of the different sub-models can be additive, for example and / or multiplicative.
  • the use of others is also logical or arithmetic links when the Results of the sub-models possible.
  • neural sub-model neural network
  • Continuous adaptation of the network parameters is also optional possible during the operation of the vehicle. For example Series tolerances are recorded and included.
  • hybrid models presented can also be used for other concepts can be reused by, for example, the input quantities of the neural Network can be relearned.
  • both the tax times can be included an electronic valve train and the spread in a motor with variable Model the valve train with the hybrid model presented.
  • Physical models sometimes use different maps or Characteristic curves that usually require a large amount of memory. In particular in the case of complicated processes, physical modeling is a big one Number of maps and characteristic curves required. In the present Overall, the use of a physical-neuronal hybrid model is less Storage space is required because the neural networks require elaborate maps and Characteristic curves can be avoided. Rather, the lesser need Network parameters in neural networks require less memory.
  • the only drawing shows a simple schematic block diagram in which an overall model for modeling the air mass filling at one Internal combustion engine with variable valve timing with a physical model for basic filling and a neural network model for the influence of spreading is described.
  • the basic filling is physical and depending on the speed N, the cylinder stroke (stroke) and the pressure difference D_P and the Suction temperature T_Ans described. These parameters are the physical model as input variables and determine accordingly a map stored in it and some thermodynamic Basic equations the initial quantity of the physical model.
  • the influence of the camshaft spread is determined using the neural network model described, since it is difficult to create a physical model here.
  • input variables for the neural network model serve (Stroke) the spreads of the intake and exhaust valves (E_Spr, A_Spr).
  • E_Spr, A_Spr the spreads of the intake and exhaust valves
  • Cylinder filling are determined and output.
  • This influence becomes multiplicative coupled with the output from the physical model, which leads to the then total determined air mass ML_Mod leads.
  • the proportion of the neuronal Partial model limited to the overall model. In the present case, the restriction is given in Dependence on the initial value of the physical sub-model.
  • a hybrid model can also be used to describe other overall processes such as an electronic valve train, turbocharged engines, direct injection engines or a synchronization control can be used, whereby each Sub-processes describe their own mostly completed processes and at least one sub-process is represented with a neural network.

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Combined Controls Of Internal Combustion Engines (AREA)
  • Feedback Control In General (AREA)
  • Output Control And Ontrol Of Special Type Engine (AREA)

Abstract

The model consists of at least one physical sub-model and at least one neural sub-model. One process component of the overal process is simulated with the physical model. A further process component of the total process is exclusively simulated with the neural model and the overall process is described by a combination of the separately simulated processes.

Description

Die Erfindung betrifft ein Hybridmodell zur Modellierung eines Gesamtprozesses in einem Fahrzeug bestehend aus je zumindest einem physikalischen und einem neuronalen Teilmodell.The invention relates to a hybrid model for modeling an overall process in a vehicle consisting of at least one physical and one neural sub-model.

Es ist bekannt, physikalische Zusammenhänge und Abläufe bei Prozessen modellhaft zu beschreiben. Mit den Modellen kann einerseits eine Diagnose vorhandener Sensoren durchgeführt werden. Andererseits können auch nicht meßbare Signale modellhaft erfaßt bzw. vorhandene Sensorik eingespart werden.It is known physical relationships and processes in processes to describe as a model. On the one hand, the models can be used for diagnosis existing sensors can be carried out. On the other hand, neither can Measurable signals are modeled or existing sensors are saved.

Beispielsweise kann die Füllung von Zylindern bei Motoren mit variablen Ventiltrieb über einen Luftmassensensor nur stark verzögert gemessen werden. Sie wird daher sinnvollerweise aus verschiedenen Eingangsgrößen, die direkt am Einlaß gemessen werden, und unter Zuhilfenahme eines Modells bestimmt. Dabei ist die Füllung der einzelnen Zylinder durch mehrere Stellgrößen beeinflußt, die teilweise voneinander abhängig oder auch unabhängig sind.For example, the filling of cylinders in engines with variable valve train measured with a very delayed air mass sensor. It will therefore expediently from different input variables, which are directly at the inlet be measured and determined with the help of a model. Here is the Filling of the individual cylinders influenced by several manipulated variables, some of them are interdependent or independent.

Eine Möglichkeit zur Modellierung sind empirische Verfahren, wie z.B. Kennfelder. Empirische Verfahren sind jedoch meist ungenau und erfordern einen hohen Abstimmungsaufwand. Eine weitere Möglichkeit sind physikalische Funktionen, bei denen das Prozeßverhalten aus der Betrachtung der physikalischen Zusammenhänge abgeleitet wird. Allerdings sind für mache Prozesse physikalische Funktionen manchmal schwierig zu erstellen. Insbesondere müssen das Gesamtsystem und die Abhängigkeiten innerhalb des Systems bekannt sein. Auch nimmt der Aufwand für die Erstellung physikalischer Modelle mit zunehmender Modellkomplexität überproportional zu. Darüber hinaus sind für verschiedene Konzepte (z.B. Direkteinspritzer, elektronischer Ventiltrieb, variabler Ventiltrieb, etc.) immer neue Modelle zu erstellen.Empirical methods, such as Maps. However, empirical methods are usually imprecise and require a high level Coordination effort. Another possibility are physical functions, at which the process behavior from the consideration of the physical relationships is derived. However, for some processes, physical functions are sometimes difficult to create. In particular, the overall system and the Dependencies to be known within the system. Also the effort for the creation of physical models with increasing model complexity disproportionately too. In addition, for different concepts (e.g. Direct injection, electronic valve train, variable valve train, etc.) always new To create models.

Aus der DE 197 06 750 A1 ist ein Verfahren zur Gemischsteuerung bei einem Verbrennungsmotor sowie eine Vorrichtung zur Durchführung dieses Verfahrens bekannt. Gemäß dem darin beschriebenen Ausführungsbeispiel wird die in einen Brennraum des Verbrennungsmotors gelangende Luftmasse aus einer Eingangsgröße bestimmt. Ferner wird die zuzuführende Kraftstoffmenge in Abhängigkeit von dieser Eingangsgröße ermittelt. Bei der Ermittlung der Kraftstoffmenge wird ein neuronales Netzwerk verwendet, welches lernfähig ist. Bei dem vorgestellten Verfahren dient das neuronale Netzwerk zur Beschreibung der Steuergröße für den Kraftstoffpfad in Abhängigkeit des Motorbetriebszustandes und der fahrerbeeinflußten Steuergröße für den Luftpfad. Bei der Bildung der Steuergröße für den Kraftstoffpfad wird bei dieser Ausführungsform ausschließlich auf das neuronale Netzwerk gesetzt.DE 197 06 750 A1 describes a method for controlling the mixture in a Internal combustion engine and a device for performing this method known. According to the exemplary embodiment described therein, the Combustion chamber of the internal combustion engine air mass coming from a Input size determined. Furthermore, the amount of fuel to be supplied in Determined as a function of this input variable. When determining the The amount of fuel used is a neural network that is capable of learning. At In the method presented, the neural network is used to describe the Control variable for the fuel path depending on the engine operating state and the driver-influenced control variable for the air path. In the formation of the The control variable for the fuel path is exclusive in this embodiment set on the neural network.

Ein wesentlicher Nachteil von neuronalen Netzen liegt darin, daß sie außerhalb des Arbeitsbereiches, in dem die Trainingsdaten ermittelt werden, ein unplausibles Extrapolationsverhalten aufweisen können und dafür in sicherheitskritischen Prozessen, z.B. bei Kraftfahrzeugen, nur schwer einsetzbar sind.A major disadvantage of neural networks is that they are outside the Work area in which the training data are determined, an implausible Can have extrapolation behavior and therefore in safety-critical Processes, e.g. in motor vehicles, are difficult to use.

Aufgabe der vorliegenden Erfindung ist es, ein Hybridmodell zur Modellierung eines Gesamtprozesses in einem Fahrzeug anzugeben, mit welchem sich physikalisch schwierig zu beschreibende Prozesse modellieren lassen, ohne das unplausible Extrapolationsverhalten in Kauf genommen werden müssen.The object of the present invention is to develop a hybrid model for modeling a To specify the overall process in a vehicle, with which physical have difficult to describe processes modeled without the implausible Extrapolation behavior must be accepted.

Diese Aufgabe wird durch die im Anspruch 1 genannten Merkmale gelöst.This object is achieved by the features mentioned in claim 1.

Erfindungswesentlich ist, daß der Gesamtprozeß (beispielsweise die Befüllung der Zylinder) in Teilprozesse aufgeteilt wird, welche von verschiedenen Teilmodellen beschrieben und dann zu einem Gesamtmodell zusammengeführt werden. Vorliegend wird zumindest ein Prozeßanteil mit einem physikalischen Modell und ein Prozeßanteil mit einem neuronalen Model beschrieben. Das neuronale Model übernimmt dabei die Beschreibung eines Prozeßanteils, welcher physikalisch schwierig zu fassen ist.It is essential to the invention that the overall process (for example the filling of the Cylinder) is divided into sub-processes, which are of different sub-models described and then combined into an overall model. At least one process component with a physical model and described a process part with a neural model. The neural model takes over the description of a process part, which is physical is difficult to grasp.

Als konkrete Anwendung läßt sich die Modellierung der Luftmassenfüllung bei Verbrennungsmotoren, beispielsweise mit variablem Ventiltrieb, angeben. Bei dieser Anwendung könnte die Basisfüllung über ein physikalisches Modell bestimmt werden. Der Einfluß der Nokkenwellenspreitzung hingegen könnte über das neuronale Netzwerk beschrieben werden. Gerade bei der Beschreibung des Einflusses der Nockenwellenspreitzung ist ein physikalisches Modell nur mit hohem Aufwand zu erstellen.The modeling of the air mass filling can be used as a concrete application Specify internal combustion engines, for example with variable valve train. At this Application could determine the basic filling using a physical model become. However, the influence of camshaft spreading could neural network are described. Especially when describing the Influence of camshaft spreading is only possible with a high physical model Create effort.

Die Modellierung des Basismodells mit einer physikalischen Prozeßbeschreibung hat den Vorteil, daß der Anteil des neuronalen Teilmodells am Gesamtmodell gezielt beschränkt werden kann. Auf diese Weise wird gewährleistet, daß das Gesamtmodell kein unplausibles Extrapolationsverhalten zeigt.The modeling of the basic model with a physical process description has the advantage that the share of the neural sub-model in the overall model can be deliberately restricted. This ensures that Overall model shows no implausible extrapolation behavior.

Bei einer Anwendung des Hybridmodells auf die Beschreibung der Befüllung von Zylindern bei einem Verbrennungsmotor kann die Basisfüllung mit dem physikalischen Modell in Abhängigkeit von Fahrbetriebsbedingungen, wie der Drehzahl, einem Zylinder-Hub und/oder der Druckdifferenz in einem Zylinder beschrieben werden.When applying the hybrid model to the description of the filling of Cylinders in an internal combustion engine can be filled with the basic filling physical model depending on driving operating conditions, such as the Speed, a cylinder stroke and / or the pressure difference in a cylinder to be discribed.

Die Zusammenführung der verschiedenen Teilmodelle kann beispielsweise additiv und/oder multiplikativ gewählt werden. Natürlich ist auch die Verwendung anderer logischer oder arithmetischer Verknüpfungen bei einer Zusammenführung der Ergebnisse der Teilmodelle möglich.The merging of the different sub-models can be additive, for example and / or multiplicative. Of course, the use of others is also logical or arithmetic links when the Results of the sub-models possible.

Natürlich kann die Belernung des neuronalen Teilmodelles (neuronales Netzwerk) gezielt durch Vorgabe von Lernwerten vor der konkreten Anwendung erstellt werden. Optional ist aber auch eine kontinuierliche Adaption der Netzparameter während des Betriebs des Fahrzeugs möglich. So können beispielsweise Serientoleranzen erfaßt und miteinbezogen werden. Of course, learning the neural sub-model (neural network) created specifically by specifying learning values before concrete application become. Continuous adaptation of the network parameters is also optional possible during the operation of the vehicle. For example Series tolerances are recorded and included.

Als Vorteile des Hybridmodelles gegenüber einem rein physikalischen Vollmodell ist eine deutliche Reduzierung des Modellierungsaufwandes anzugeben. Durch die Vermeidung eines neuronalen Vollmodells kann ein (unplausibels) Extrapolationsverhalten ausgeschlossen werden.The advantages of the hybrid model over a purely physical full model is specify a significant reduction in modeling effort. Through the Avoiding a full neural model can be an (implausible) Extrapolation behavior can be excluded.

Überdies können die aufgestellten Hybridmodelle auch bei anderen Konzepten wiederverwendet werden, indem zum Beispiel die Eingangsgrößen des neuronalen Netzwerkes neu belernt werden. Vorliegend lassen sich sowohl die Steuerzeiten bei einem elektronischen Ventiltrieb und die Spreizung bei einem Motor mit variablem Ventiltrieb mit dem vorgestellten Hybridmodell modellieren.In addition, the hybrid models presented can also be used for other concepts can be reused by, for example, the input quantities of the neural Network can be relearned. In the present case, both the tax times can be included an electronic valve train and the spread in a motor with variable Model the valve train with the hybrid model presented.

Physikalische Modelle bedienen sich teilweise verschiedener Kennfelder oder Kennlinien, die in der Regel einen großen Speicherbedarf erfordern. Insbesondere bei komplizierten Prozessen ist für die physikalische Modellierung eine große Anzahl von Kennfeldern und Kennlinien erforderlich. Bei der vorliegenden Verwendung eines physikalisch-neuronalen Hybridmodelles wird insgesamt weniger Speicherplatz benötigt, da mit den neuronalen Netzen aufwendige Kennfelder und Kennlinen vermieden werden können. Vielmehr benötigen die geringeren Netzparameter bei neuronalen Netzwerken einen geringeren Speicherbedarf.Physical models sometimes use different maps or Characteristic curves that usually require a large amount of memory. In particular in the case of complicated processes, physical modeling is a big one Number of maps and characteristic curves required. In the present Overall, the use of a physical-neuronal hybrid model is less Storage space is required because the neural networks require elaborate maps and Characteristic curves can be avoided. Rather, the lesser need Network parameters in neural networks require less memory.

Die vorliegende Erfindung wird anhand eines speziellen Ausführungsbeispiels und mit Bezug auf die einzige nachfolgende Zeichnung näher erläutert.The present invention is based on a specific embodiment and explained in more detail with reference to the only drawing below.

Die einzige Zeichnung zeigt ein einfaches schematisches Blockdiagramm, bei dem ein Gesamtmodell zur Modellierung der Luftmassenfüllung bei einem Verbrennungsmotor mit variabler Ventilsteuerung mit einem physikalischen Modell für die Basisbefüllung und einem neuronalen Netz-Modell für den Spreitzungseinfluß beschrieben ist. Die Basisfüllung wird physikalisch und in Abhängigkeit von der Drehzahl N, dem Zylinder-Hub (Hub) und der Druckdifferenz D_P sowie der Ansaugtemperatur T_Ans beschrieben. Diese Parameter werden dem physikalischen Modell als Eingangsgrößen zugeführt und bestimmen entsprechend einem darin abgelegten Kennfeld sowie einiger thermodynamischer Grundgleichungen die Ausgangsgröße des physikalischen Modells. The only drawing shows a simple schematic block diagram in which an overall model for modeling the air mass filling at one Internal combustion engine with variable valve timing with a physical model for basic filling and a neural network model for the influence of spreading is described. The basic filling is physical and depending on the speed N, the cylinder stroke (stroke) and the pressure difference D_P and the Suction temperature T_Ans described. These parameters are the physical model as input variables and determine accordingly a map stored in it and some thermodynamic Basic equations the initial quantity of the physical model.

Der Einfluß der Nockenwellenspreizung wird mittels des neuronalen Netzmodells beschrieben, da hier ein physikalisches Modell nur schwer zu erstellen ist. Als Eingangsgrößen für das neuronale Netzmodell dienen neben dem Zylinder-Hub (Hub) die Spreizungen der Einlaß- und der Auslaßventile (E_Spr, A_Spr). Durch das Belernen der Kopplungen des neuronalen Netzes kann am Ausgang des neuronalen Modells der Einfluß der Nockenwellenspreitzung auf die Zylinderbefüllung ermittelt und ausgegeben werden. Dieser Einfluß wird multiplikativ mit dem Ausgang aus dem physikalischen Modell gekoppelt, was zu der dann insgesamt ermittelten Luftmasse ML_Mod führt. Dabei ist der Anteil des neuronalen Teilmodells am Gesamtmodell beschränkt. Die Beschränkung erfolgt vorliegend in Abhängigkeit vom Ausgangswert des physikalischen Teilmodells.The influence of the camshaft spread is determined using the neural network model described, since it is difficult to create a physical model here. As In addition to the cylinder stroke, input variables for the neural network model serve (Stroke) the spreads of the intake and exhaust valves (E_Spr, A_Spr). By learning the couplings of the neural network can be at the exit of the the influence of camshaft spreading on the neuronal model Cylinder filling are determined and output. This influence becomes multiplicative coupled with the output from the physical model, which leads to the then total determined air mass ML_Mod leads. The proportion of the neuronal Partial model limited to the overall model. In the present case, the restriction is given in Dependence on the initial value of the physical sub-model.

Damit wird gewährleistet, daß das Gesamtmodell kein unplausibles Extrapolationsverhalten zeigt. Versuche haben ergeben, daß sich die mittleren Fehler bei einer Realisierung der Modellierung der Frischluft-Zylinderbefüllung bei Verbrennungsmotoren mit variablen Ventilsteuerungen mit dem physikalisch-neuronalen Hybridmodell deutlich reduzieren lassen.This ensures that the overall model is not implausible Shows extrapolation behavior. Experiments have shown that the middle Error when realizing the modeling of the fresh air cylinder filling Internal combustion engines with variable valve controls with the physical-neuronal Have the hybrid model significantly reduced.

Natürlich kann ein Hybridmodell auch zur Beschreibung anderer Gesamtprozesse wie eines elektronischen Ventiltriebes, turboaufgeladener Motoren, Direkteinspritzermotoren oder einer Gleichlaufregelung verwendet werden, wobei jeweils Teilprozesse eigene zumeist abgeschlossene Vorgänge beschreiben und zumindest ein Teilprozeß mit einem neuronalen Netzwerk dargestellt wird.Of course, a hybrid model can also be used to describe other overall processes such as an electronic valve train, turbocharged engines, direct injection engines or a synchronization control can be used, whereby each Sub-processes describe their own mostly completed processes and at least one sub-process is represented with a neural network.

Claims (9)

Hybridmodell zur Modellierung eines Gesamtprozeßes in einem Fahrzeug bestehend aus zumindest einem physikalischen und einem neuronalen Teilmodell,
dadurch gekennzeichnet, daß ein Prozeßanteil aus dem Gesamtprozeß ausschließlich mit dem physikalischen Modell simuliert wird und ein weiterer Prozeßanteil aus dem Gesamtprozeß ausschließlich mit dem neuronalen Modell simuliert wird und der Gesamtprozeß durch eine Zusammenführung der jeweils separat simulierten Prozesse beschriebenen wird. Hybrid model for modeling an overall process in a vehicle, consisting of at least one physical and one neuronal sub-model,
characterized, that a part of the process from the overall process is only simulated with the physical model and another part of the process from the overall process is simulated exclusively with the neural model and the overall process is described by merging the separately simulated processes. Hybridmodell nach Anspruch 1,
dadurch gekennzeichnet, daß ein Basismodell physikalisch simuliert wird. Hybrid model according to claim 1,
characterized, that a basic model is physically simulated. Hybridmodell nach Anspruch 1 oder 2,
dadurch gekennzeichnet, daß eine Modellierung der Luftmassenfüllung bei Verbrennungsmotoren mit variabler Ventilsteuerung durchgeführt wird. Hybrid model according to claim 1 or 2,
characterized, that a modeling of the air mass filling in internal combustion engines with variable valve control is carried out. Hybridmodell nach Anspruch 3,
dadurch gekennzeichnet, daß die Basisfüllung mit dem physikalischen Modell und in Abhängigkeit von der Drehzahl, von einem Zylinder-Hub und/oder der Druckdifferenz und/oder der Ansaugtemperatur in einem Zylinder beschrieben wird. Hybrid model according to claim 3,
characterized, that the basic filling with the physical model and depending on the speed, a cylinder stroke and / or the pressure difference and / or the intake temperature in a cylinder is described. Hybridmodell nach Anspruch 3 oder 4,
dadurch gekennzeichnet, daß der Einfluß der Nockenwellenspreizung mit dem neuronalen Modell simuliert wird. Hybrid model according to claim 3 or 4,
characterized, that the influence of the camshaft spread is simulated with the neural model. Hybridmodell nach einem der vorhergehenden Ansprüche,
dadurch gekennzeichnet, daß bei der Zusammenführung beider Teilmodelle eine aditive oder eine multiplikative Kopplung gewählt wird. Hybrid model according to one of the preceding claims,
characterized, that an aditive or a multiplicative coupling is selected when merging the two sub-models. Hybridmodell nach einem der vorhergehenden Ansprüche,
dadurch gekennzeichnet, daß der Einfluß des neuronalen Modells auf das Gesamtmodell beschränkt ist. Hybrid model according to one of the preceding claims,
characterized, that the influence of the neural model on the overall model is limited. Hybridmodell nach einem der vorhergehenden Ansprüche,
dadurch gekennzeichnet, daß eine Belernung des neuronalen Modells vor dem Betrieb des Fahrzeugs erfolgt. Hybrid model according to one of the preceding claims,
characterized, that the neural model is learned before the vehicle is operated. Hybridmodell nach einem der vorhergehenden Ansprüche,
dadurch gekennzeichnet, daß eine Belernung des neuronalen Modells adaptiv während des Betriebes des Fahrzeugs erfolgt. Hybrid model according to one of the preceding claims,
characterized, that the neural model is learned adaptively during the operation of the vehicle.
EP00106509A 1999-04-01 2000-03-25 Hybrid model for the modelling of a whole process in a vehicle Ceased EP1041264A3 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE19914910 1999-04-01
DE19914910A DE19914910A1 (en) 1999-04-01 1999-04-01 Hybrid model for modeling an overall process in a vehicle

Publications (2)

Publication Number Publication Date
EP1041264A2 true EP1041264A2 (en) 2000-10-04
EP1041264A3 EP1041264A3 (en) 2002-08-07

Family

ID=7903278

Family Applications (1)

Application Number Title Priority Date Filing Date
EP00106509A Ceased EP1041264A3 (en) 1999-04-01 2000-03-25 Hybrid model for the modelling of a whole process in a vehicle

Country Status (2)

Country Link
EP (1) EP1041264A3 (en)
DE (1) DE19914910A1 (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1253491A2 (en) * 2001-04-24 2002-10-30 Bayer Aktiengesellschaft Hybrid model and method for determining the mechanical properties and processing properties of an injection moulded article
EP1342899A1 (en) * 2000-12-12 2003-09-10 Toyota Jidosha Kabushiki Kaisha Controller of internal combustion engine
WO2006000474A1 (en) * 2004-06-24 2006-01-05 Siemens Aktiengesellschaft Method for determining the air mass in a cylinder
FR2876152A1 (en) * 2004-10-06 2006-04-07 Renault Sas IMPROVED METHOD AND SYSTEM FOR ESTIMATING EXHAUST GAS TEMPERATURE AND INTERNAL COMBUSTION ENGINE EQUIPPED WITH SUCH A SYSTEM
DE102004055313A1 (en) * 2004-11-16 2006-05-18 Volkswagen Ag Cylinder pressure sensor diagnosis/reinforcement adaptation performing method for combustion engine, involves receiving sensors signals, and determining diagnosis value/adaptation parameter based on moment difference of model torques
WO2006114550A1 (en) * 2005-04-28 2006-11-02 Renault S.A.S Method for controlling a motor vehicle using a network of neurones

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE10113538B4 (en) * 2001-03-20 2012-03-01 Bayerische Motoren Werke Aktiengesellschaft Regulating device and control method
DE10203919A1 (en) * 2002-01-31 2003-08-21 Bayerische Motoren Werke Ag Reconstructing physical magnitudes for further processing in association with engine controller using neural network, produces models for individual system sections
DE10237328B4 (en) * 2002-08-14 2006-05-24 Siemens Ag Method for controlling the combustion process of an HCCI internal combustion engine
AT6293U1 (en) * 2002-12-05 2003-07-25 Avl List Gmbh METHOD FOR CONTROLLING OR CONTROL OF AN INTERNAL COMBUSTION ENGINE WORKING IN A CIRCUIT PROCESS
DE10328015A1 (en) * 2003-06-23 2005-01-13 Volkswagen Ag Virtual lambda sensor for road vehicle internal combustion engine has computer connected to engine control module for regulating air-fuel mixture
DE102014000397A1 (en) 2014-01-17 2015-07-23 Fev Gmbh Model-based cylinder fill detection for an internal combustion engine
DE102021204544A1 (en) 2021-05-05 2022-11-10 Robert Bosch Gesellschaft mit beschränkter Haftung Method for operating a hydraulic cylinder of a working machine

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0445555A2 (en) * 1990-03-06 1991-09-11 Bayerische Motoren Werke Aktiengesellschaft Method for the regulation of the camshaft phasing, continuously variable according to RPM
DE19706756A1 (en) 1997-02-20 1998-09-03 Siemens Ag Gradient amplifier for magnetic resonance tomography

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE4338607B4 (en) * 1993-11-11 2005-10-06 Siemens Ag Method and device for managing a process in a controlled system
DE19547496C2 (en) * 1995-12-19 2003-04-17 Dierk Schroeder Process for regulating internal combustion engines
US5877954A (en) * 1996-05-03 1999-03-02 Aspen Technology, Inc. Hybrid linear-neural network process control
JPH10122017A (en) * 1996-10-14 1998-05-12 Yamaha Motor Co Ltd Engine control system
US5714683A (en) * 1996-12-02 1998-02-03 General Motors Corporation Internal combustion engine intake port flow determination
DE19706750A1 (en) * 1997-02-20 1998-08-27 Schroeder Dierk Prof Dr Ing Dr Method for controlling the mixture in an internal combustion engine and device for carrying it out
DE19709955C2 (en) * 1997-03-11 2003-10-02 Siemens Ag Method and device for controlling an internal combustion engine

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0445555A2 (en) * 1990-03-06 1991-09-11 Bayerische Motoren Werke Aktiengesellschaft Method for the regulation of the camshaft phasing, continuously variable according to RPM
DE19706756A1 (en) 1997-02-20 1998-09-03 Siemens Ag Gradient amplifier for magnetic resonance tomography

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1342899A1 (en) * 2000-12-12 2003-09-10 Toyota Jidosha Kabushiki Kaisha Controller of internal combustion engine
EP1342899A4 (en) * 2000-12-12 2012-04-25 Toyota Motor Co Ltd Controller of internal combustion engine
EP2527630A3 (en) * 2000-12-12 2014-07-23 Toyota Jidosha Kabushiki Kaisha Controller for an internal combustion engine with variable valve mechanism
EP2527631A3 (en) * 2000-12-12 2014-08-27 Toyota Jidosha Kabushiki Kaisha Controller for an internal combustion engine with variable valve mechanism
EP2570637A3 (en) * 2000-12-12 2014-07-23 Toyota Jidosha Kabushiki Kaisha Controller for an internal combustion engine with variable valve mechanism
EP1253491A2 (en) * 2001-04-24 2002-10-30 Bayer Aktiengesellschaft Hybrid model and method for determining the mechanical properties and processing properties of an injection moulded article
EP1253491B1 (en) * 2001-04-24 2006-08-02 Bayer MaterialScience AG Hybrid model and method for determining the mechanical properties and processing properties of an injection moulded article
US7357127B2 (en) 2004-06-24 2008-04-15 Siemens Aktiengesellschaft Method for determining the air mass in a cylinder
WO2006000474A1 (en) * 2004-06-24 2006-01-05 Siemens Aktiengesellschaft Method for determining the air mass in a cylinder
US7664593B2 (en) 2004-10-06 2010-02-16 Renault S.A.S. Method and system for estimating exhaust gas temperature and internal combustion engine equipped with such a system
WO2006037926A1 (en) * 2004-10-06 2006-04-13 Renault S.A.S Improved method and system for estimating exhaust gas temperature and internal combustion engine equipped with such a system
FR2876152A1 (en) * 2004-10-06 2006-04-07 Renault Sas IMPROVED METHOD AND SYSTEM FOR ESTIMATING EXHAUST GAS TEMPERATURE AND INTERNAL COMBUSTION ENGINE EQUIPPED WITH SUCH A SYSTEM
DE102004055313B4 (en) * 2004-11-16 2017-06-22 Volkswagen Ag Method and device for diagnosis or gain adaptation of cylinder pressure sensors
DE102004055313A1 (en) * 2004-11-16 2006-05-18 Volkswagen Ag Cylinder pressure sensor diagnosis/reinforcement adaptation performing method for combustion engine, involves receiving sensors signals, and determining diagnosis value/adaptation parameter based on moment difference of model torques
FR2885175A1 (en) * 2005-04-28 2006-11-03 Renault Sas METHOD FOR CONTROLLING A VEHICLE ENGINE USING A NEURON NETWORK
CN101198783B (en) * 2005-04-28 2010-10-13 雷诺股份公司 Method for controlling a vehicle motor using a network of neurones
US7774127B2 (en) 2005-04-28 2010-08-10 Renault S.A.S. Method for controlling a motor vehicle using a network of neurones
WO2006114550A1 (en) * 2005-04-28 2006-11-02 Renault S.A.S Method for controlling a motor vehicle using a network of neurones

Also Published As

Publication number Publication date
DE19914910A1 (en) 2000-10-26
EP1041264A3 (en) 2002-08-07

Similar Documents

Publication Publication Date Title
DE102007021592B4 (en) METHOD FOR CREATING A MACHINE FIELD AND MODEL DURING A DEVELOPMENT PROCESS OF A COMBUSTION ENGINE
EP0170018B1 (en) Process and apparatus for the self-testing of control levers
AT520179B1 (en) Test bench and method for carrying out a test
EP1041264A2 (en) Hybrid model for the modelling of a whole process in a vehicle
DE102008001081A1 (en) Method and engine control unit for controlling an internal combustion engine
DE102019127482B4 (en) Control device
WO2013131836A2 (en) Method for optimizing the emissions of internal combustion engines
DE102017107271A1 (en) Method for determining a driving cycle for driving tests for determining exhaust emissions from motor vehicles
EP3698036A1 (en) Calculation of exhaust emissions from a motor vehicle
DE102005019017A1 (en) Method and device for fault diagnosis for internal combustion engines
AT520827B1 (en) A method of determining a vehicle parameter of a vehicle record of a vehicle and using the vehicle parameter on a test bench
EP1623284B1 (en) Method for optimizing vehicles and engines used for driving such vehicles
DE102007020355B4 (en) An engine control system and method for detecting a malfunction in a torque control path
AT515154A2 (en) Method of creating a model ensemble
AT523093A1 (en) Method and system for analyzing and / or optimizing a configuration of a vehicle type
EP3374618B1 (en) System and method for calibrating a vehicle component
WO2009095333A1 (en) Method for controlling an internal combustion engine
EP1273782A2 (en) Method for determining characteristic mapping data for controlling the characteristic map of an internal combustion engine, and a method for controlling an internal combustion engine
DE102017106943A1 (en) Method and arrangement for simulating driving tests
DE102008004218B4 (en) Procedure for determining the dynamic soot emission
DE69918639T2 (en) Control method for controlling the injection of an internal combustion engine as a function of fuel quality
DE102015207270A1 (en) Method and apparatus for simulation coupling of an event-driven controller subsystem and a plant subsystem
DE10219797B4 (en) Method for optimizing a model for controlling an internal combustion engine
EP3871052B1 (en) Method for calibrating a technical system
DE102016103643B4 (en) Method and device for checking a software of a control unit of a vehicle

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 CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE

AX Request for extension of the european patent

Free format text: AL;LT;LV;MK;RO;SI

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 CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE

AX Request for extension of the european patent

Free format text: AL;LT;LV;MK;RO;SI

17P Request for examination filed

Effective date: 20020829

AKX Designation fees paid

Designated state(s): DE ES FR GB IT SE

17Q First examination report despatched

Effective date: 20040712

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

Free format text: STATUS: THE APPLICATION HAS BEEN REFUSED

18R Application refused

Effective date: 20060416