DE102011106544A1 - Method for controlling auto-ignition in e.g. Otto engine used in motor vehicle, involves executing evaluation of combustion in one previous cycle according to setting of operating parameter for cylinder of combustion engine - Google Patents

Abstract

The method involves performing combustion process of internal combustion engine (1) in an operating range out by a process of space ignition. The evaluation of the combustion in one previous cycle is executed according to the setting of operating parameter for cylinder (2) of internal combustion engine so as to obtain combustion cycle. The transition cycle is determined during load hop so as to prevent excitation of controlled ignition. The evaluation of each cylinder of internal combustion engine is performed by setting next combustion cycle for cylinder. An independent claim is included for internal combustion engine.

Description

The present invention relates to a method for controlled self-ignition in an internal combustion engine and an internal combustion engine with a switchover between conventional engine operation and a controlled auto-ignition.

The injection and the residual gas rate are suitable influencing factors for the regulation of an Otto engine self-ignition, abbreviated also CAI from English for Controlled Auto Ignition. This is also reflected in an analysis of pending patents again. A large number of these publications treat the control of the residual gas rate, a variable valve train required for this purpose and methods for controlling the fuel injection or the associated charge stratification.

The research activities that can be found in pamphlets can be divided into two areas:

a) Investigation of slow, quasi-stationary effects:

• - Determination of optimal manipulated variables as a function of external, slowly changing boundary conditions such as speed, intake manifold pressure, fuel properties, component and air temperatures.
• - These manipulated variables are stationary or change comparatively slowly with time constants >> 1 cycle. The corresponding work is usually carried out on stationary engine test benches.
• - For stabilization, closed-loop controls are used, but they deliberately have large time constants to minimize system excitation.
• - These research activities are very advanced, the influences of the individual manipulated variables in the stationary case have been described in detail by numerous authors.

b) Investigations of the highly dynamic effects characteristic of CAI.

• - Current research focuses on highly dynamic effects that occur from cycle to cycle.
• - This includes cycle-to-cycle dependencies, which were basically described by some authors.
• - So far, there are only basic research, the relationships are presented qualitatively.
• - The highly dynamic cycle-to-cycle effects are the main reason why the firing process is described as difficult to control by many authors.

A selection of different research directions is given in the following publications:
DE 10 2007 030 280 A1 Bosch, method for operating an internal combustion engine
DE 10 2008 004 360 A1 Bosch, method for controlling a self-igniting internal combustion engine
DE 11 2006 000 529 T5 GM, method for controlling transient loads between lean and stoichiometric combustion modes
DE 10 2008 004 362 A1 Bosch, method for controlling an internal combustion engine, computer program and control unit
DE 10 2008 004 361 A1 Bosch, method for controlling an internal combustion engine, computer program and control unit
DE 10 2008 004 360 A1 Bosch, method and apparatus for controlling a self-igniting internal combustion engine
DE 10 2008 000 552 A1 Bosch, method for operating a self-igniting internal combustion engine and corresponding control device
DE 10 2007 051 250 A1 Bosch, method for controlling a self-igniting internal combustion engine and corresponding control device
DE 10 2006 034 806 A1 Bosch, method for operating an internal combustion engine
DE 10 2004 038 122 B4 Siemens, method and apparatus for controlling an internal combustion engine
DE10237328B4 Siemens, method for controlling the combustion process of an HCCI internal combustion engine
DE10233612B4 Siemens, method and apparatus for controlling the combustion process of an HCCI engine
DE 10 2006 027 571 A1 Bosch, method for operating an internal combustion engine
DE 10 2006 052 631 A1 Bosch, method for operating an internal combustion engine
DE 10 2006 041 467 A1 Daimler, control concepts in gasoline engines with homogeneous compression-ignition combustion
DE 10 2007 001 237 A1 Ford, system and procedures to control auto-ignition

The benefits of controlled auto-ignition, CAI, for high efficiency and low emissions have been demonstrated in numerous laboratory tests. However, the controllability of the combustion process with changing boundary conditions has proved problematic. Depending on the operating conditions, unstable system conditions may occur which lead to unwanted acoustic behavior, the extinguishing of the Burn or even damage the engine.

The object of the present invention is therefore to reliably apply the advantages demonstrated on test benches in the vehicle.

This object is achieved with a method having the features of claim 1 and with an internal combustion engine having the features of claim 6. Further advantageous embodiments and further developments are specified in the subclaims as well as in the following description. Individual features of the subclaims, the description as well as the figures can be linked to other features of other embodiments to further developments, without these being enumerated.

A method is proposed for controlled self-ignition in an internal combustion engine in which combustion takes place in an operating region by a method of a space ignition, wherein an evaluation of the combustion of at least one previous cycle is carried out in order to derive at least one setting for at least one operating parameter of a coming, preferably to win currently pending combustion cycle.

The space ignition allows ignition of a combustion mixture, preferably by compression heat. In particular, the proposed method can be used in CAI and HCCI combustion processes. HCCI is the English abbreviation for Homogeneous Charge Compression Ignition, ie homogeneous compression ignition. In addition to use on a gasoline engine or on a diesel engine, the proposed method can also be used in a Miller or Atkinson method.

The combustion time can be directly influenced by the ignition spark in conventional gasoline engines and by the injection in the case of diesel engines. For CAI and HCCI this direct influence possibility is missing, instead the combustion center of gravity results indirectly from the ignition delay time, which in turn depends on numerous boundary conditions such as mixture composition, fuel properties, pressure and temperature in the combustion chamber. Among other things, the combustion focus position is precisely controlled by the proposed method in order to reliably represent the controlled auto-ignition in transient vehicle use. The influencing of the combustion characteristics such as the center of gravity is hereby preferably carried out indirectly via the aforementioned variables injection and residual gas rate. In particular, the proposed method is based on the fact that at least one predecessor cycle is evaluated to such an extent that its effects on a subsequent cycle are thereby captured. Since in controlled auto-ignition the mixture contains a significant portion of the residual gas of the previous cycle and its properties affect the ignition delay time, an analysis of the previous cycle enables improved and in particular safe controlled auto-ignition. In addition to the analysis of the previous cycles but also other boundary conditions are used. This concern in particular the speed, component and gas temperatures but also other boundary conditions, which exert an influence on the motor process.

It is preferred that an evaluation for each cylinder of the internal combustion engine takes place individually and for each cylinder an adjustment for at least one combustion cycle is made. The individual cylinder-adapted evaluation makes it possible to compensate for existing fluctuations of each individual cylinder. In particular, an expired combustion is not identical to each other in each cylinder, but varies. The individual evaluation ensures that there is no suppression of fluctuations due to averaging, but that the possibility exists of being able to react immediately in a targeted manner when fluctuations occur.

It is preferable if the method is performed cascading to predict several future cycles. In particular, not only a single completed combustion activity of an earlier cycle, but also several predecessor cycles can be considered in this way. This allows, among other things, to determine a trend formation. Furthermore, there is the possibility to extend the horizon of a model-predictive optimization by the prediction of several future cycles. In this way, several future cycles can be included in the optimization.

It is advantageous if the method is carried out in real time. This allows to prevent in particular expected outliers which may be caused by, for example, incomplete burns or other disturbances in a cylinder during a predecessor cycle. If these are not avoided, component damage may occur. The proposed real-time calculation makes it possible to carry out an immediate adaptation of the pending injection and thus to prevent irregularities in the combustion. Real-time, for example, provides that the calculation is completed within defined time criteria. For example, the Calculations within a period of time to be completed, which refers to crankshaft degree, abbreviated ° CA and thus the rotational speed. Thus, depending on the crankshaft speed and thus the load or transient case, the defined period for the optimization to change. A further embodiment provides that a fixed predetermined period of time is predetermined, which is sufficient for all load cases. For example, in one embodiment it can be provided that the real-time calculation is completed within a maximum of 30 ° CA. Another embodiment provides a time window of a maximum of 10 milliseconds for the calculations.

A further development of the method provides, taking into account one or more boundary conditions, to determine one or more parameters of the coming cycle, wherein the prediction takes into account a plurality of characteristics from one or more earlier cycles, optimizing one or more manipulated variables by varying them in real time by means of preferably a cost function. The prediction of the parameters can be done by means of physical calculation, artificial neural networks or maps

• – Drehzahl
• – Kühlwassertemperatur
• – Öltemperatur
• – Temperatur der angesaugten Luft
• – Kraftstoffqualität und Kraftstofftemperatur Berücksichtigen.

The prediction can have one or more constraints such as:

• - Rotation speed
• - Cooling water temperature
• - oil temperature
• - Temperature of the intake air
• - Consider fuel quality and fuel temperature.

• – Der Zylinderdruckverlauf und dessen Historie
• – Aus dem Druckverlauf generierte Kenngrößen
• – Die eingespritzte Kraftstoffmenge
• – Schwerpunktlage
• – Temperatur der Zylinderladung
• – Homogenität der Zylinderladung
• – Ladungsbewegung der Zylinderladung
• – Akustische Signale, beispielsweise eines Klopfsensors
• – Signale eines Ionenstromsensors
• – Eine Analyse der Drehungleichförmigkeit oder daraus generierte Kenngrößen

For example, according to one embodiment, the characteristics to be taken into account in the prediction of one or more previous cycles are the load and the center of gravity of the previous combustion. Another embodiment provides that the gas temperature and the mixture composition after completion of the previous combustion are used. Also, for example, these four parameters can be selected together. Other parameters that may become of the past are, for example

• - The cylinder pressure history and its history
• - Characteristics generated from the pressure profile
• - The injected fuel quantity
• - center of gravity
• - Temperature of the cylinder charge
• - Homogeneity of the cylinder charge
• - Charge movement of the cylinder charge
• - Acoustic signals, such as a knock sensor
• - Signals of an ion current sensor
• - An analysis of the rotational irregularity or parameters generated from it

By predicting characteristics of one or more future cycles based on constraints and characteristics of past combustion cycles, one or more manipulated variables can be optimized in real time by varying them by means of preferably a cost function. This real-time optimization is preferably carried out individually for each cylinder. The optimization in this case provides that preferably a plurality of manipulated variables are varied within the available calculation time and for each manipulated variable combination according to the selected modeling, a prediction of the resulting combustion parameters for one or more combustion cycles takes place. By selecting a suitable cost function or another method of calculation, it is possible to evaluate the predicted combustion and thus determine an optimum of the manipulated variables. For this purpose, for example, when using a corresponding cost function within this weighting can be provided, which takes into account the individual boundary conditions. There is also the possibility that the boundary conditions specify fixed limits within which the manipulated variables can be varied. By optimizing the manipulated variables from cycle to cycle, predefined combustion target values can be achieved very precisely.

If the proposed method is stimulated, for example, with a jump of the target load, a transition cycle is automatically determined to avoid stimulation of the controlled auto-ignition, so that optimized control values determined only in a cycle following the current cycle later, preferably a second cycle, are used , before that, however, based on it, close to coming manipulated variables. This transitional cycle is not determined by expert knowledge or system knowledge, but is the result of automatic optimization. This transition cycle allows jumps to the stationary manipulated variables immediately in the event of a load jump. This jumping of the control variables involves the risk of excitation of the system and thus a rocking. In case of a load jump to high load combustion with residual gas of the much colder predecessor cycle is ignited. Since the preconditions clearly differ from the stationary state under high load, the manipulated variables must be adapted accordingly, which is realized fully automatically by the proposed method.

Furthermore, according to an additional idea, a method is proposed for controlled self-ignition in a self-igniting internal combustion engine with homogenization provided there, wherein an evaluation of combustion takes place in an earlier cycle in order to derive at least one setting for at least to win an upcoming, preferably currently pending combustion cycle. Here, in particular, a diesel engine combustion, which is abbreviated under the name HCCI known, can be improved. In this case, the method can resort to those features described above, which are also used in the spark-ignited internal combustion engine. The same applies to features that are still set out below in connection with the controlled auto-ignition of the spark-ignited internal combustion engine and explained in more detail below with reference to the individual figures. The same also applies to the proposed in detail below, proposed internal combustion engine.

Furthermore, an internal combustion engine having a plurality of cylinders is proposed, wherein each cylinder is assigned at least one ignition device, an injection device and a sensor and a switchover is provided by an ignition means of the ignition device and a controlled auto-ignition per cylinder, the internal combustion engine having an engine control unit, and the internal combustion engine is equipped with a real-time optimization of an upcoming operation of at least each injection device based on an evaluation of a completed combustion cycle. Preferably, the method described above is implemented here in the engine control unit. The required arithmetic operations can be realized by the microprocessor.

A further embodiment of the internal combustion engine provides that it has at least one field programmable gate array, abbreviated FPGA, for the optimization. An FPGA has the advantage over processor systems that a large number of parallel arithmetic operations can be performed. This saves time, so that an optimization as proposed above by means of this electronic integrated circuit can be carried out in the necessary real-time consideration.

An FPGA preferably has a flash memory for configuration. Preferably, this flash memory is housed in the chip housing, so that no external flash memory is required for this purpose. Furthermore, there is the possibility that the FPGA is housed as an integrated circuit with multiple chips in a housing as a so-called multi-dye. Preferably, it is provided that a synchronous clock is provided on the FPGA, so that the calculation optimization in parallel calculation within the provided time window in all calculation positions is thereby also completed in a timely manner. According to one embodiment it can be provided that the FPGA used has implemented as a memory cell an SRAM, an EPROM, an EEPROM, a flash or another memory module. Another embodiment, however, provides that the FPGA is programmable only once and is thus protected against external interference.

It is preferred that the FPGA is part of the engine control unit and is used for the optimization described above. In addition, however, the FPGA can also be used for other optimizations or parallel computing tasks. Due to the number of cylinders available, it may be sufficient that a single FPGA is provided for the optimization. Due to the firing order, there is a sufficient time window for the FPGA to optimize each cylinder individually. Another embodiment provides that the FPGA is redundant. In this way, if the FPGA fails, the other FPGA can do its job. It is also possible to control the FPGA by detecting deviations of results between the two FPGA devices.

A further embodiment provides that the internal combustion engine for optimization has a system-on-A chip, which has an FPGA. In this case, for example, the engine control unit may have an FPGA module including CPU, bus system, RAM, ROM, controller, timer and the like integrated together. Due to the time-critical functionality is further provided that the necessary parameters for the calculation or the calculated control variables are not transmitted via a bus or similar, present in the motor vehicle transmission path. This may be sufficient at a sufficient data transfer rate. However, it has also proved to be advantageous if there is a direct wiring to the individual injectors of the individual cylinders and the engine control unit, so that the results of the FPGA immediately take into account in the upcoming injection, in particular a pilot injection, their rate, timing, history and with respect to other parameters or a post-injection.

According to one embodiment, it is provided, for example, that the average cylinder internal pressure p _mi and the center of gravity position are determined as a parameter of the preceding cycle in order to carry out the optimization. It can act as a sensor, a pressure sensor which is arranged in the cylinder, for example in a spark plug. Furthermore, there is the possibility that a knock sensor is used as the sensor. There is also the possibility that the sensor is on Ion current sensor is. By means of the sensor, the previous combustion or a parameter resulting therefrom can be determined and evaluated. Therefore, it is preferable that each cylinder has at least one of these sensors to obtain the required combustion characterizing value. Furthermore, there is the possibility that several of these sensors are assigned to each cylinder and therefore can also enter into a variety of parameters in the optimization.

Further advantages and embodiments will be apparent from the following figures. However, the resulting from the individual figures features are not limited to these. The same applies to the associated figure descriptions. Rather, one or more of these features can be linked to one or more features from other figures or other description passages, in particular the general description for further embodiments. Furthermore, one or more embodiments as described below represent only exemplary embodiments of the invention, without, however, restricting them as such, but rather only explain. Show it:

1 a comparison of a time course of center of gravity positions for a terminally stable and a stable operating point,

2 a normalized autocovariance function of three different measuring points,

3 a dependence of a current mean in-cylinder pressure on the previous,

4 a comparison of measured data and forecasts,

5 an exemplary course of a replacement cost function as a function of selected manipulated variables of the injection,

6 a reaction in the event of a sudden change in an injection quantity,

7 resulting curves of the average internal cylinder pressure and a combustion center position with optimized selection of an injection strategy,

8th an independently generated by a selected optimization algorithm transition strategy at a load step, and

9 a schematic view of an internal combustion engine of a vehicle.

1 represents a comparison of the time profiles of the center of gravity for an unstable and a stable operating point. Depending on the operating point, there are always deviations in the combustion process of successive cycles. As a measure of the stability of the combustion, the standard deviation of the mean pressure σ _pmi is often used.

In 1 has cylinders 3 a high standard deviation on, cylinder 1 however, a small standard deviation. The normalized autocovariance function shows a significant negative correlation of successive cycles, as it also explains below 3 clearly shows. Consequently, qualitatively an alternating behavior of the center of gravity situations is to be expected. This correlation is stronger the more unstable the combustion and the higher σ _pm . The larger σ _pm , the larger the coefficients of the normalized autocovariance cf. also 3 together with the associated description.

2 shows normalized autocovariance functions of three different measurement points at n = 2000 1 / min, pmi = 1.5bar. The past combustion cycles have a significant impact on the current combustion. The Otto engine self-ignition can thus be represented as autoregressive process, which depends not only on the manipulated variables, but also on the past output variables. On the basis of characteristic quantities of the direct predecessor cycles, a prediction of the next combustion can take place with knowledge of the manipulated variables and the boundary conditions. To describe the influence of the past, several cycles or only the direct predecessor cycle can be used. As parameters to describe the influence of the past, for example, the use of the load and the center of gravity of the previous one is used. However, it is also possible to use other variables such as the gas temperature or the mixture composition.

3 shows the dependence of the current pmi on the pmi and the center of gravity of the predecessor cycle. In 3 .27 represents the xy-plane the two parameters of the predecessor cycle, ie the internal mean pressure p _mi (z-1) and the center of gravity α ₅₀ (z-1). On the z-axis the resulting internal mean pressure p _mi (z ). This level now indicates the expected value of the mean internal pressure p _mi (z) for a set of manipulated variables as a function of parameters of the predecessor cycle α ₅₀ (z-1) and p _mi (z-1). In order to make a prediction of parameters of the following cycle on the basis of the past, depending on the manipulated variables and boundary conditions, artificial neural networks or other prediction methods can alternatively be used. By cascading the mechanisms shown, a prediction of several upcoming burns in Dependence of the known manipulated variables and boundary conditions are performed.

4 shows a comparison of the measured data and the prediction of the gradients of pmi and center of gravity. in 4 Measurements and predictive values are compared for a strongly varying measurement with Otto engine self-ignition.

• – Die Restgasrate über externe Abgasrückführung
• – Die Steuerzeiten der Ladungswechselventile
• – Der Einspritzzeitpunkt oder die Verwendung von weiteren Einspritzungen
• – Die Verwendung eines unterstützenden Zündfunkens
• – Ladungsbewegungsorgane wie beispielsweise Drallklappen

5 represents the course of the replacement cost function depending on the selected manipulated variables of the injection. In the example shown, two control variables have been selected for the optimization, namely mass and distribution of the injection, but it is also possible a higher number of degrees of freedom. Alternatively or widening, the following manipulated variables can also be varied:

• - The residual gas rate via external exhaust gas recirculation
• - The timing of the charge exchange valves
• - The injection time or the use of further injections
• - The use of a supporting spark
• - Charge movement organs such as swirl flaps

With the aid of the illustrated method, it is possible to predict a plurality of combustion parameters, such as, for example, load, center of gravity, knock intensity, emissions, peak pressure and efficiency. Depending on the driving situation, different target values must be observed for these parameters. The target values are, for example, a defined mean pressure, which corresponds to the driver's desired torque, low fuel consumption and low emissions. To solve the multi-size optimization, the determination of a scalar replacement cost function is proposed. x '= f (z ₁ , z ₂ , ..., z _c )

Various methods are possible for this, for example a linear combination of the individual cost functions can be carried out with weighted factors.

To determine the minimum of the replacement cost function, known optimization methods can be used. A single cycle is suggested as the forecast horizon, but several cycles can be used to provide model predictive optimization.

When realizing the developed method on a real internal combustion engine hard time criteria must be met, therefore, a real-time capable system is needed. This is implemented in a control unit. Preferably, an engine control unit is used for this purpose. An injection control device, which also controls or regulates a variable valve train, can also be used. In the example shown, the selected manipulated variables, ie the two injection parameters, are determined between completion of a combustion and the following injection. For the calculation, therefore, a defined, speed-dependent limit must be adhered to.

Measurements on the engine test bench

The key challenge in realizing CAI is mastering transient behavior. The most demanding case is an abrupt change in the target values within a cycle. Therefore, real-time optimization is validated in the case of highly dynamic processes by load jumps, which can be performed on the engine test bench. As an example of the pilot operation, jumps are made between two points of different amounts of fuel. The injection _quantity is periodically from m _Kr.1 = 6.45 mg to m _Kr.2 = 10; 36 mg and back varied.

6 shows the curves of p _mi and α ₅₀ in the described jumps of the injection quantity. The measured values of σ _pm and α ₅₀ are marked with circles, which in real-time predicted values are indicated by circles. In the course of the internal mean pressure p _mi a characteristic, alternating behavior after the load jumps is recognizable. The step responses to cycle 3 and cycle 13 show a very similar, damped transient response. This trend is predicted qualitatively correctly by the real-time optimization. The validation of the developed algorithm is analogous to the stationary measurements in the form that not the manipulated variables, but the target values are given. The target values are based on the previously measured mean values of the controlled load jumps. The developed real-time algorithm can now independently decide from cycle to cycle which injection strategy comes closest to the desired target values.

In 7 the resulting curves of p _mi and α _{50 are shown} with an optimized choice of injection strategy. In the case of the load jumps, the new target variables are only operated with the stationary values during the second cycle. In order to avoid excitation of the system, a transition cycle is determined by the real-time optimization. This transition cycle allows both the increase and the reduction of the load a significant reduction in system excitation and an almost immediate achievement of steady state.

In 8th are the ones chosen independently by the optimization algorithm Transition strategies and the timing of the load jumps shown. The contour lines show the load achieved during the respective injection in stationary case. It is significant that the transition strategies shown were not determined by expert knowledge or iterative attempts. The traces have been determined in real time as a result of determining the optimum and are based solely on the models underlying the system.

Out 9 is a schematic view of a section of an internal combustion engine 1 showing several cylinders 2 having. Every cylinder 2 is a cylinder device 3 assigned. As well as an injection device 4 , The cylinder also has a variable valve train 5 on. By means of the variable valve train 5 different timing at an intake 6 and exhaust 7 Valve of the cylinder 2 be provided. Likewise, the opening times as well as the opening stroke can be variably set. Furthermore, the cylinder 2 a sensor 8th assigned. This sensor may be an ion current sensor, a pressure sensor, a knock sensor or any other sensor by means of which it is possible to determine one or more parameters related to the elapsed combustion of a previous cycle. The sensor 8th stands as well as the ignition device 3 , the injector 4 and the variable valve train with an engine control unit 9 in connection. These connections are by appropriately wavy lines 10 indicated. It is preferred if the injection device 4 at least connected via a hard-wired connection with the engine control. Due to the fact that the engine control unit 9 On the one hand, a bus connection on the other hand, but also has a direct wiring, a real-time operation as described above can be guaranteed. The engine control unit 9 has a system On A Chip 11 on. This system-on-a-chip comprises at least one CPU, one or more memories and in particular also a field programmable gate array 12 , abbreviated FPGA. With this it is possible to carry out parallel a variety of arithmetic operations in order to enable an optimization of the upcoming injection process. In addition, the engine control unit has 9 also implements a switchover 13 on. Switching 13 is able to switch between ignition by means of the ignition device 3 and a controlled auto-ignition per cylinder 2 to switch over. Switching 13 in particular, evaluates parameters such as the load, load range and the like in order to determine whether such switching is possible and, in particular, also advantageous under predefinable criteria. The switching is preferred 13 in the engine control unit 9 implemented, preferably in the system-on-a-chip 11 ,

Furthermore, for the embodiment of the internal combustion engine, for the method of controlled auto-ignition in a gasoline engine or for a HCCI process in a diesel engine to the above cited in the prior art documents within the scope of the disclosure of the respective options for design and necessary Components referred to in the disclosure of the invention.

• – Aufzeigen der Korrelation aufeinander folgender Verbrennungszyklen.
• – Darstellung der motorischen Verbrennung als autoregressiven Prozess.
• – Eine Prognose kommender Verbrennungen auf Basis der vergangenen Verbrennungszyklen.
• – Nutzung von charakteristischen Größen wie zum Beispiel Schwerpunktlage der Verbrennung, Last, aktuelle Abgastemperatur, Zylinderdruck usw. der vergangenen Zyklen, um eine Vorhersage der kommenden Verbrennung zu machen.
• – Echtzeit-Mehrgrößenoptimierung, um für jeden Zyklus die Verbrennung exakt an definierte Zielgrößen anzupassen.
• – Darstellung der aktiven Führung der Verbrennung bei Lastsprüngen durch Echtzeitoptimierung

The proposed invention makes possible in particular the following:

• - Demonstrate the correlation of consecutive combustion cycles.
• - Representation of engine combustion as an autoregressive process.
• - A forecast of upcoming burns based on past combustion cycles.
• - Use of characteristic quantities such as center of gravity of combustion, load, current exhaust gas temperature, cylinder pressure, etc. of the past cycles to make a prediction of the coming combustion.
• - Real-time multi-variable optimization to precisely match combustion to defined target values for each cycle.
• - Representation of the active management of combustion at load jumps through real-time optimization

• – Die ottomotorische Selbstzündung bietet im Teillastbetrieb einen vergleichbar guten Wirkungsgrad wie geschichtete Magerkonzepte, jedoch mit deutlich geringeren Emissionen.
• – Auch der dieselmotorische Betrieb kann durch Homogenisierung in Form eines HCCl weiter optimiert werden, jedoch mit vergleichbaren Problemen bei der Kontrollierbarkeit des Brennverfahrens, die durch die Erfindung gelöst werden.
• – Insbesondere bei zahlreichen Fahrzeugen kann die Erfindung ihre Vorteile zeigen.
• – Besonders vorteilhaft wird die Erfindung zur ottomotorischen Selbstzündung und zur Homogenisierung des dieselmotorischen Prozesses bei der Kontrollierbarkeit des transienten Verhaltens eingesetzt.
• – Mit Hilfe dieser Erfindung, insbesondere des dargestellten Verfahrens, kann genau dieser kritische Punkt der Brennverfahren sicher beherrscht werden.

An application of the invention is preferably provided as follows:

• - The ottomotor auto-ignition offers in part-load operation, a comparable good efficiency as stratified lean concepts, but with significantly lower emissions.
• - The diesel engine operation can be further optimized by homogenization in the form of a HCCl, but with comparable problems in the controllability of the combustion process, which are solved by the invention.
• - Especially in many vehicles, the invention can show its advantages.
• The invention is used particularly advantageously for Otto engine autoignition and for the homogenization of the diesel engine process in the controllability of the transient behavior.
• - With the help of this invention, in particular the illustrated method, precisely this critical point of the combustion process can be safely controlled.

The proposed methods are characterized, in relation to the particular combustion method, in particular by the type of combustion characteristic of auto-ignition and not by the absence of a spark.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102007030280 A1 [0004]
• DE 102008004360 A1 [0004, 0004]
• DE 112006000529 T5 [0004]
• DE 102008004362 A1 [0004]
• DE 102008004361 A1 [0004]
• DE 102008000552 A1 [0004]
• DE 102007051250 A1 [0004]
• DE 102006034806 A1 [0004]
• DE 102004038122 B4 [0004]
• DE 10237328 B4 [0004]
• DE 10233612 B4 [0004]
• DE 102006027571 A1 [0004]
• DE 102006052631 A1 [0004]
• DE 102006041467 A1 [0004]
• DE 102007001237 A1 [0004]

Claims (10)

Method for controlled auto-ignition in an internal combustion engine, in which combustion in an operating range is carried out by a method of a space ignition, wherein an evaluation of the combustion is carried out in at least one previous cycle, to at least one setting for an operating parameter of at least one upcoming, preferably currently pending To win combustion cycle. A method according to claim 1, characterized in that an evaluation for each cylinder of the internal combustion engine takes place individually and for each cylinder an adjustment is made for a coming combustion cycle. A method according to claim 1 or 2, characterized in that the method is performed cascading to predict further future cycles. A method according to claim 1, 2 or 3, characterized in that in real time an optimization of one or more manipulated variables by variations thereof by means of preferably a cost function taking into account one or more boundary conditions, to determine one or more characteristics of the coming cycle, wherein the optimization takes into account multiple characteristics from one or more previous cycles. Method according to one of the preceding claims, characterized in that at a load step to avoid excitation of the controlled auto-ignition, a transition cycle is determined, so that only in an after the clock cycle later following, preferably second cycle determined optimized control variables are used, but before it ajar close to manipulated variables. Internal combustion engine based on a gasoline engine having a plurality of cylinders, wherein each cylinder is associated with at least one ignition device, an injection device and a sensor and a switch between ignition by means of the ignition device and a controlled auto-ignition per cylinder is provided, wherein the internal combustion engine having an engine control unit, and Internal combustion engine is equipped with a real-time optimization of an upcoming operation of at least each injector based on an evaluation of a successful combustion cycle. Internal combustion engine according to claim 6, characterized in that the internal combustion engine for the optimization comprises at least one field programmable gate array. Internal combustion engine according to claim 6 or 7, characterized in that the engine control unit for optimization comprises a field programmable gate array. Internal combustion engine according to one of the preceding claims, characterized in that the internal combustion engine for optimization has a system on a chip having a field programmable gate array. Method for controlled auto-ignition in a self-igniting internal combustion engine with homogenization, wherein an evaluation of combustion takes place in an earlier cycle in order to obtain at least one setting for at least one upcoming, preferably currently pending combustion cycle.

Images