DE102015200965A1 - Optimal parameterization of complex components and associated functional structures, especially internal combustion engines and hybrid drives - Google Patents

Abstract

The invention relates to a method for the optimal parameterization of complex components and associated complex functional structures, in particular of internal combustion engines and hybrid drives and their associated functional structures, in which at least one target value is parametrized step by step within a calibration and optimization process by a measurement data set on a test bench for a drive cycle is detected and a multi-dimensional engine simulation model is created on the basis of the measured data set determined on the test bench and therefrom a parameter data set. It is envisaged that the measured data set of the driving cycle is optimized with the aid of an extended engine simulation environment and at least one optimum target value is determined, according to which d) the optimized measured data record automatically taking into account the optimum target value in an optimized engine control unit data set for an engine control unit (MSG) of the internal combustion engine and Hybrid drive is transferred.

Description

The invention relates to a method for the optimal parameterization of complex components and associated complex functional structures, in particular of internal combustion engines and hybrid drives and their associated functional structures. Within the method, at least one target value is parameterized step by step within a calibration and optimization process by recording a measurement data record on a test bench for a drive cycle, and creating a multi-dimensional engine simulation model on the basis of the test data record determined on the test bench. Subsequently, a parameter data record to be optimized is created from the measurement data record. Currently, the optimization of the parameter data set already takes place with the aid of a parameterization software in a specific engine simulation environment.

The parameterization of complex component and functional structures as well as operating strategies in internal combustion engines and hybrid drives is possible without parameterization software only with a lot of test bench and personnel deployment.

The existing parameterization software currently only allows parameterization of simple function structures. For complex nested function structures there is currently no suitable parameterization software available.

The well-known parameterization software is also only suitable for special parameterization scopes, whereby the parameterization software can only be used for special ECUs.

Another disadvantage is that software extensions for taking into account new functions of so-called functional structures can only be created by the manufacturer of the parameterization software itself.

Another disadvantage is that the functional structures, once created, can no longer be integrated into the programming software of other ECU suppliers. In addition, the present programming software is disadvantageously not able to provide a holistic - one speaks of a global - optimization for all component and functional structures and operating strategies in internal combustion engines and hybrid drives.

Up to now, an attempt has been made to get as close to a global optimum as possible by means of technical parameterization on the engine test bench and variation of individual parameters.

This currently requires high engine test bench and vehicle roller run times.

Up to now, parameters or parameter data sets have been determined on the basis of measurement data on the engine test bench and adapted manually via certain correction values in order to ensure optimum parameters or parameter sets for all component and functional structures and operating strategies of internal combustion engines and hybrid drives.

The object of the invention is to ensure an improved parameterization of complex component and functional structures as well as operating strategies in internal combustion engines and hybrid drives.

The object is achieved by a method for optimal parameterization of complex components and associated complex functional structures, in particular internal combustion engines and hybrid drives and their associated functional structures.

• a) für einen Fahrzyklus ein Messdatensatz auf einem Prüfstand erfasst wird, und
• b) ein multi-dimensionales Motorsimulationsmodell auf der Basis des auf dem Prüfstand ermittelten Messdatensatzes und daraus ein Parameterdatensatz erstellt wird.

According to the invention, at least one target value is parameterized stepwise within a calibration and optimization process by

• a) a measured data set is recorded on a test bench for a driving cycle, and
• b) a multi-dimensional engine simulation model is created on the basis of the measured data record determined on the test bench and from this a parameter data record.

• c) der Parameterdatensatz des Fahrzyklusses mit Hilfe einer erweiterten Motorsimulationsumgebung optimiert und mindestens ein optimaler Zielwert ermittelt wird.

According to the invention, it is further provided that

• c) the parameter data set of the driving cycle is optimized with the aid of an extended engine simulation environment and at least one optimum target value is determined.

• – mindestens ein Motorparameter und/oder
• – Umgebungsparameter und und/oder
• – Betriebsstrategieparameter

In a preferred embodiment of the method according to method step c), the optimization of the parameter data set of the driving cycle with the help of the extended engine simulation environment for determining the at least one optimal target value is as follows: The method is preferably characterized in that in the extended engine simulation environment when determining the at least one optimal target value for the parameter data set of the driving cycle optimized in method step c),

• At least one engine parameter and / or
• - Environmental parameters and and / or
• - Operating strategy parameters

is considered in at least one dynamic model and subsequently in an emission model and / or a consumption model and the at least one optimal target value is optimized and determined by means of at least one optimization algorithm.

According to the invention, the engine simulation environment is substantially extended by the aforementioned simulation parameters of the parameter data set (engine parameters, environmental parameters, operating strategy parameters) compared to the conventional engine simulation environment, whereby a significantly more efficient calibration and optimization process can be carried out.

The advantages are explained in more detail in the following detailed description with reference to accompanying figures.

In detail, it is preferably provided that, within the extended engine simulation environment, the engine parameters are used as simulation parameters of the parameter data set, in particular a turbocharger charging system and / or an injection system and / or a rail pressure line and / or an exhaust gas recirculation system and / or a high and low pressure exhaust gas recirculation system and / or or a flap system and / or a hybrid system is / are considered.

Furthermore, it is preferably provided in detail that within the engine simulation environment, the environmental parameters are taken into account as simulation parameters of the parameter data set, in particular an ambient pressure and / or an intake temperature.

In addition, it is preferably provided in detail that within the extended engine simulation environment the operating strategy parameters are considered as simulation parameters of the parameter data set, in particular different injection patterns of the injection system and / or exhaust gas recirculation modes of the exhaust gas recirculation system.

It is preferably proceeded such that for the just mentioned simulation parameters of the parameter data set in an associated engine map setpoint values are formed, which serve as input variables for the dynamic model, within the dynamic model a control deviation over a controller actual value is modeled over time , wherein the simulation parameters thus modeled again serve as input variables for the models for target value determination of the optimal target value, in particular for the emission model and / or the consumption model.

Subsequently, it is preferred that the simulation parameters of the parameter data set modeled within the dynamic model be adopted as input variables, according to which at least one integrated time-resolved cumulated optimal target value, in particular an emission target value, and the simulation parameters based on the simulation parameters over the drive cycle / or a consumption target value is formed.

It is particularly provided that within the extended engine simulation environment in the optimization of the driving cycle - automatically - for the respective simulation parameters of the parameter data set (engine parameters, environmental parameters and operating strategy parameters) by means of the at least one optimization algorithm, an optimized parameter data set is created.

The optimization of the driving cycle within the extended engine simulation environment for determining the at least one target value is preferably carried out using a combination of an optimization algorithm with local convergence and a global optimization algorithm following the local optimization algorithm.

• d) der derart optimierte Parameterdatensatz unter Berücksichtigung des optimalen Zielwertes automatisch in einen optimierten Motorsteuergeräte-Datensatz für ein Motorsteuergerät des Verbrennungsmotors und Hybridantriebs überführt wird.

According to the invention, it is provided in a final step of the method that

• d) the thus optimized parameter data set, taking into account the optimum target value, is automatically converted into an optimized engine control unit data set for an engine control unit of the internal combustion engine and hybrid drive.

The method is performed on a data processing system.

The data processing system is according to the inventive method a) for acquiring the measurement data on a test bench, and b) for creating a multi-dimensional engine simulation model based on the measured data set determined on the test bench and for creating the parameter data set, and c) for optimizing the Parameter data set and for determining an optimal target value and d) for transferring the optimized parameter data set in the optimized engine control unit data record for an engine control unit suitable. The patent application further relates to a computer program product comprising program code means stored on a computer-readable medium for carrying out the described method when the program is executed on the data processing system, in particular a computer.

The invention will be explained in more detail with reference to the accompanying figures.

Show it:

1 a schematic representation of a conventional process flow of a calibration process for vehicle cycles of an internal combustion engine or a hybrid drive;

2 a schematic representation of a conventional engine simulation environment U as shown 1.4 of the 1 ;

3 a schematic representation of a vehicle cycle of an internal combustion engine with different operating modes and environmental conditions in a detailed representation;

4 a schematic representation of an extended engine simulation environment according to the invention U ';

5 an engine control unit and a schematically represented characteristic map structure which comprises basic maps and correction maps of a correction structure;

6 a schematically illustrated exemplary map structure as an exemplary embodiment, which comprises a basic map and further associated correction maps;

7 a further schematic representation of the extended engine simulation environment according to the invention U 'according to 4 with a controller functional architecture according to the invention;

8th a schematic representation of an optimization process according to the invention.

In 1 1, a conventional sequence process of map calibration for a vehicle cycle F of an internal combustion engine or a hybrid drive is shown in various figures.

The illustration 1.1 symbolizes a vehicle roller test bench P on which a measurement data record is created.

From the measurement data generated on the vehicle dynamometer P, as shown in the figure 1.2 created a multi-dimensional engine simulation model S.

The preparation of the engine simulation model S on the basis of the measurement data generated on the vehicle roller test bench P is part of the calibration and optimization process K.

A preferred procedure for creating the engine simulation model S on the basis of selected data to create a reduced measurement data set according to .2 is explained in another patent application of the Applicant.

Based on the driving cycle F (speed [km / h] over time t [s]) according to the diagram in figure 1.3 is now using a so-called engine simulation environment U as shown 1.4 a driving cycle optimization performed.

So far, disadvantageously, the calculation of an optimized engine control unit data set for an engine control unit MSG ( 5 ) calculated per engine parameter only for a basic map G, which symbolically in the figure 1.5 is shown.

In case of a component change or a function adaptation to an internal combustion engine or to a hybrid drive, the illustration shows 1.6 In addition, disadvantageously, the necessity that the described optimizing calibration and optimization process K must always be performed again.

In the 2 will the in 1 under the picture 1.4 illustrated conventional engine simulation environment U shown and explained in more detail.

In the picture 2.1 is analogous to the picture 1.3 the driving cycle F shown.

Based on the driving cycle F, in the currently applied engine simulation environment U on the basis of the engine parameters, target values Z, as shown in FIG 2.2 for legal requirements, such as emission values and / or consumption values calculated on the basis of emission and consumption models.

In addition, customer-specific target values Z have already been determined on the basis of the engine parameters, as shown in Figure 2.2 , such as dynamic response and / or engine acoustics.

The illustration 2.3 symbolizes an optimizer by means of an optimization algorithm according to Figure 2.4 an optimization of a target value Z or more target values Z is performed.

For example, the target value Z - fuel consumption - is optimized in compliance with another statutory target value Z, in particular as a function of a legal emission value and thus minimized.

Upon reaching the desired target value Z as part of the optimization process, according to the preceding description and the figures 2.2 . 2.3 and 2.4 Finally, taking into account the predetermined target values Z, an optimized Engine control unit data set as shown 2.5 created for all considered engine parameters.

A conventional optimized engine control unit dataset as shown 2.5 then consists of the basic map G determined as part of the optimization process, as shown in the illustration 1.5 and takes into account one motor parameter each. This conventional engine control unit data record is stored in the engine control unit MSG.

For the fulfillment of the legal regulations, which are steadily tightened, and the target values Z derived therefrom with regard to the emissions and the fuel consumption as well as the mentioned customer specific target values, technical extensions on the component level side of the internal combustion engines and hybrid drives and the associated functional structures are required.

A holistic optimization is for EU-VI powertrains with always in many details in a short time technically modified internal combustion engines and hybrid drives with the previous procedure according to the description of the 1 and 2 only possible with considerable effort. To ensure ever-shorter development times, it is essential to make the calibration and optimization process more efficient and to further reduce vehicle roll test bench times.

Currently, only a gradual optimization over the vehicle dynamometer is possible using the simple parameterization software. In addition, a holistic consideration for all target values Z with the previously described calibration and optimization process K or the described optimization process with the addition of the conventional parameterization software is very difficult.

• – das Aufladesystem Turbolader, ein- oder mehrstufig
• – das Einspritzsystem, Einspritzventile, Raildruckstrecke
• – das Abgasrückführsystem, Hoch- und Niederdruck-Abgasrückführung
• – das Klappensystem, Drosselklappe, Abgasklappe
• – das Hybridsystem, elektromotorische Unterstützung, Leistung und Einkopplung in den Antriebsstrang

möglich sein. Konstruktive Neuauslegung und/oder Bauteiländerungen und/oder Funktionsanpassungen der Verbrennungsmotore oder Hybridantriebe soll/sollen möglichst einfach und effizient innerhalb des Kalibrierungs- und Optimierungsprozesses K berücksichtigt werden können.The constructive component design, if applicable, a subsequent component change as shown in the illustration 1.6 and operating coordination of the control, if appropriate, a subsequent functional adaptation of the component actuator and factors also as shown in the figure 1.6 , which belongs to an internal combustion engine or hybrid drive components should advantageously for

• - the turbocharger supercharging system, single or multi-stage
• - The injection system, injectors, rail pressure line
• - the exhaust gas recirculation system, high and low pressure exhaust gas recirculation
• - The flap system, throttle, exhaust flap
• - The hybrid system, electromotive assistance, power and coupling in the powertrain

to be possible. Constructive redesign and / or component changes and / or functional adjustments of the internal combustion engines or hybrid drives should / should be considered as simply and efficiently as possible within the calibration and optimization process K.

For each component controller, that is to say each engine parameter and each operating strategy, the engine control unit MSG already contains parameterizable map structures. The values of the maps are currently determined by measurements or simulation calculations. Usually, the map values are determined, for example, at operating engine conditions in the standard operating mode. For deviating environmental conditions, such as, for example, cold coolant, low intake air temperatures, or variable air pressure values, only manually generated correction maps and correction characteristics exist disadvantageously in the engine control unit. For reasons of capacity, complete maps are not created and stored manually in a detrimental manner in the engine control unit MSG for each correction, but the corrections are converted via memory-space-optimized characteristic curves and operating mode control maps. Data compression represents a huge optimization task for more complex part-specific engine control unit functions due to the minimization of interpolation errors and iterative adaptation.

Based on 3 the problem of the existing optimization task of the existing calibration and optimization process K is explained in detail by means of a concrete exemplary embodiment.

The 3 shows the driving cycle F analogous to the less detailed figures 1.3 . 2.1 . 4.1 and 7.1 now in detail.

The graph (speed [km / h] versus time t [s]) shows a velocity profile with velocity plotted on the y-axis and time on the x-axis.

The solid line I shows the engine temperature T [° C], which is also plotted on a y-axis.

The dashed line II shows different Injection pattern 1 . 2 . 3 . 4 , which are plotted on a separate y-axis. To the respective injection pattern 1 . 2 . 3 . 4 includes a different number of pre and post injections. To the injection pattern 1 heard a pre-injection PI1. To the injection pattern 2 include two pilot injections PI1, PI2. To the injection pattern 3 include two pilot injections PI1, PI2 and a post-injection POI3. To the injection pattern 4 include a pilot injection PI1 and a post-injection POI3.

The speed [km / h] of the vehicle is represented by the dotted line III.

Within the described conventional calibration and optimization process K, there is currently only the possibility for a velocity profile at a constant engine temperature Tmot, for example 90 °, and for one of the injection patterns 1 . 2 . 3 . 4 to make an optimization.

Like from the 3 can be seen, but the engine temperature Tmot in reality according to the line I is not constant, but it increases in the selected embodiment over the time [t] of an output temperature in the embodiment 30 ° C according to the characteristic I, for example, to the engine temperature Tmot 90 ° , In addition, in the driving cycle F for attaining the legal target values Z and the customer-specific requirements, that is, the customized target values Z, switching from one of the injection patterns becomes 1 . 2 . 3 . 4 to another of the injection pattern 1 . 2 . 3 . 4 necessary.

That is, by an additional consideration of the non-constant engine temperature and / or the switching of the injection pattern 1 . 2 . 3 . 4 Within a considered driving cycle F, the conventional calibration and optimization process K can be improved, that is optimized.

In the previous calibration and optimization process K, this switching of the injection pattern 1 . 2 . 3 . 4 or the non-constant engine temperature Tmot in the cycle optimization within the existing engine simulation environment U is not adversely affected.

It is pointed out again that the engine temperature Tmot and the consideration of the injection pattern 1 . 2 . 3 . 4 Depending on the speed v of the vehicle in the selected embodiment, only three of very many parameters to be considered simultaneously within the calibration and optimization process K to be optimized.

The task is thus in analogy to the task formulated in more general terms to ensure that the calibration and optimization process can be optimally designed, in particular engine temperatures considering the large and small cooling circuit, the pressure and temperature environment conditions and operating strategies such the respective injection patterns and the respective exhaust gas recirculation strategy of the respective internal combustion engines or hybrid drives are calibrated (parameterized) more optimally within a calibration and optimization process.

The enhanced engine simulation environment U ', modified to solve the problem, will be described in 4 by further illustration 4.1 to 4.9 explained in detail.

First, as previously for a drive cycle F, as shown in Figure 4.1 a measurement data record on a test stand P detected.

Subsequently, the multi-dimensional engine simulation model S is created on the basis of the measured data set determined on the test stand P, on the basis of which a parameter data set for the subsequent simulation in the extended engine simulation environment U 'is again created.

In this case, in the parameter data set for the driving cycle to be optimized F, as shown in FIG 4.1 , beyond the actual engine parameters, other parameters and influences in the over the prior art extended engine simulation environment U 'considered.

From the driving cycle F as shown 4.1 For the first time, the engine parameters and an associated engine map will be used to create actuator setpoints. The setpoint values are indicated by the figure 4.2 symbolizes.

The actuator setpoints represent input variables for a dynamic model of the motor control variables, which is shown in Figure 4.3 is integrated into the extended engine simulation environment U 'according to the invention.

The dynamic model 4.3 is characterized in that the control deviation of actuator actual values simulated in the dynamic model compared to the actuator setpoint values measured on the vehicle roller test stand P or, if appropriate, also controller actuator values calculated from existing characteristic diagrams, as shown in FIG 4.4 modeled over time t, that is, the control deviation is determined.

The illustration 4.4 shows for clarity the time course measured on the vehicle roller dynamometer P or in maps (upper Line) already present controller setpoints in comparison to the actual controller values (lower line) over the time t.

An example: According to the invention, for example, a dynamic time profile of the engine filling setpoint values of the respective injection pattern is obtained 1 . 2 . 3 . 4 in comparison to the time course of the engine fill actual values by determining the control deviation 4.4 modeled.

The on the basis of the dynamic model 4.3 modeled actuator setpoints and actuator actual values and their determined system deviation, are now used according to the invention as input variables for further models, as shown in the figure 4.5 which serve in a next step to determine the respective target values Z.

By considering the dynamics within the dynamic model 4.3 , which are the input variables for the other models 4.5 provides for target value determination, it is achieved that within the calibration and optimization process K more realistic target values Z can be determined by the modified - compared to the prior art extended - engine simulation environment U '.

A model integrated within the enhanced and thus improved engine simulation environment U 'which conforms to the dynamic model as shown in the figure 4.3 is, for example, an emission model as shown 4.5 , which engine parameters, such as the non-constant engine temperature Tmot, as shown in FIG 4.6 analogous to the line I of 3 and the ambient conditions, optionally additionally taking into account the ambient pressure and / or optionally the engine intake temperature.

In the exemplified temperature-dependent emission model 4.5 , which is described in more detail below with the aid of the figure 5 In addition, the operation modes such as the operation modes of the various injection patterns are considered 1 . 2 . 3 . 4 related to 3 already described and clarified.

The models in illustration 4.5 For example, the emission model or a consumption model form the extended engine simulation environment U 'that takes into account dynamic effects and environmental conditions.

Based on the driving cycle 4.1 and the method of the invention further described below then arise as output variables of the various models 4.3 and 4.5 , the dynamic model 4.3 and the exemplified emission model 4.5 Time-resolved target values, in particular emission target values Z, over the time-resolved course of the driving cycle 4.1 be integrated, so as shown in Figure 4.7 Cumulative cycle values Z, according to the embodiment in particular time-resolved cumulated emission target values Z or just time-resolved cumulative consumption target values Z and so on.

Based on the time-resolved cumulated emission target values Z, it is now advantageously possible to check already via the engine simulation whether the limits of the legally required emission target values or consumption target values are met.

The explained engine simulation according to the invention by means of the extended engine simulation environment U 'ensures that the time-resolved cumulated emission target values Z according to FIG 4.7 are so close to reality that the time-resolved cumulative emission target values Z determined by the engine simulation environment U 'according to the invention can be detected on a real vehicle roller test bench P as shown in FIG 1.1 calculated cumulative cycle values.

This means, in principle, that the cumulative cycle values determined by the engine simulation environment U 'according to the invention, which are determined and optimized for compliance with the legally required target values Z, can almost completely replace the complex work on a vehicle roller test bench P, with the exception of a few control measurements.

Finally, how will 4 further shows an optimization algorithm using an optimizer as shown 4.8 the entire map structure 4.9 bedatet, that is the map structure is as shown 4.9 provided with optimized parameter data, so that the map structure 4.9 now with all optimized engine parameters and optimized environmental conditions as well as optimized operating modes, as shown in the picture 4.3 to 4.6 described, is.

Optimization via an optimization algorithm is as illustrated 2.3 and 2.4 explained, already known, however, the optimization under inventive consideration of all engine parameters and environmental conditions and operating modes by means of the procedure described already leads to the optimized parameter data set, without making any changes in the field of optimization.

According to the invention, in addition to the described improved approach the Finally, optimization is done via a modified optimization process that is still related to 8th is explained.

The 5 Figure 11 shows a controller map structure including basic maps G and associated correction maps and correction maps.

As already explained in the prior art, so far within the conventional calibration and optimization process K per engine parameter only one basic map G has been shown in FIG 1.5 of the 1 determined.

Several of these basic maps G per engine parameter are symbolic in 5 in the picture 5.1 shown. So far, the several basic maps G as shown in Figure 5.1 exclusively with associated manually generated correction formulas or correction characteristics or correction characteristics corresponding to that in 5 filed engine control unit MSG stored.

The 5 clarified in figure 5.2 in that, starting from the engine simulation according to the invention - according to the invention 4 described procedure maps G now automatically generated correction maps and correction curves for each desired engine parameters (and not only as previously for a motor parameter) with additional consideration of the environmental conditions and / or the operating modes of the parameter data set calculated and optimized.

The peculiarity of the invention is that the correction maps and correction curves are now optimized simultaneously to the basic maps G via correction values cor0, cor1, cor2, cor3, this determination within the optimization process for calculating the cumulative cycle values according to FIG 4.7 . 4.8 and 4.9 on the basis of the extended engine simulation environment U '.

The correction maps and correction curves that are within the optimization process 4.7 . 4.8 and 4.9 optimized, that is to be bedatet, result in a so-called correction structure 5.2 , in the 5 through the picture 5.2 pictorially illustrated. Thus, it is no longer necessary to manually rework the correction maps and correction curves of the correction structure 5.2 within the new simulation process in an advantageous manner, automatically fill with optimized data of the optimized parameter data set. The special feature according to the invention is thus, in other words, that for the correction structure 5.2 an optimized parameter data set automates during the described optimization process 4.7 . 4.8 . 4.9 created by the calculated cumulative cycle values 4.7 emanates. The optimized data of the parameter data set, the correction structure provided with the data 5.2 , are thus advantageously based on and take into account the engine parameters within the engine simulation environment U 'within the dynamic model 4.3 as well as the environmental conditions and operating modes according to the to the picture 4.3 to 4.6 , described procedure for determining the cumulative cycle values 4.7 ,

A determination of the optimized parameter data set and the automatic definition of the correction structure 5.2 taking into account the mentioned parameters is not possible in this way so far.

For further explanation an example:
In 6 For example, nine maps are shown. So far, for example, only one basic map G has been optimized per engine parameter. For example, so far, a basic map G, as shown in Figure 6.1 at a constant engine temperature Tmot of 90 ° C and one of the operating modes - in the exemplary embodiment with two pilot injections PI1 and PI2 and a post-injection POI3 - optimized. Correction maps and correction curves have hitherto been created manually on the basis of the current engine simulation environment U and stored in the engine control unit MSG.

The 6 Now shows an engine map according to the invention starting from a basic map G as shown 5.1 with the appropriate correction structure 5.2 which has automatically been filled with optimized parameter data within the calibration and optimization process K according to the invention and has thus been - been - been.

It becomes clear that now at different engine temperatures Tmot, for example 90 ° C, 60 ° C and 30 ° C etc. (Fig 6.1 . 6.2 . 6.3 ) also several modes of operation 3 . 4 . 2 be taken into account. Thus, at the different engine temperatures Tmot also for the operating mode 3 - Two pilot injections PI1, PI2 and a post-injection POI3 and the operating mode 4 - a pilot injection PI1 and a post-injection POI3 - and for the operating mode 2 - Two pilot injections PI1, PI2 - in addition to the previously only basic map G eight additional maps automatically optimized, that is automatically provided with optimized parameter data.

Starting from a previous basic map G at an engine temperature Tmot of for example 90 ° C at two pilot injections PI1 and PI2 and a post-injection POI3 (operating mode = injection pattern 3 ) put the According to the invention additionally eight maps, the corresponding correction maps of the correction structure 5.2 in response to a changing engine temperature Tmot (60 °, 30 °) or as a function of a changed operating mode (injection pattern 3 . 4 . 2 ).

By means of 7 the difference according to the invention over the prior art compared to 2 further clarified.

Again starting from a driving cycle 7.1 is in picture 7.2 symbolically represented a controller functional architecture that the basic characteristics G as shown in Figure 5.1 as well as the correction structure 5.2 in the form of automatically create correction maps analogous to the description of the 6 or automatically generate correction maps based on the enhanced engine simulation environment U '.

The procedure for determining the time-resolved cumulated target values Z within the extended engine simulation environment U 'according to the invention with subsequent optimization according to the illustration 7.3 was related to the 4 already explained in detail.

The control unit functional architecture now automatically created according to the invention in the new calibration and optimization process K 7.2 is thus based on a realistic simulation and optimization of all target values Z as shown 7.5 , wherein the simulation according to the invention compared to previous simulations of reality is much closer.

The optimized data of the parameter data set is stored in an optimized engine control unit data set as shown in Figure 7.6 in 7 transferred. In this case, interpolation errors are minimized so that the global optimum is found holistically for all parameter data.

As a result, this approach has the effect of optimizing the engine control unit data set optimized in this way 7.6 very close to the previously achieved on a real vehicle dynamometer P laboriously optimized data approaches.

As a result, resources can be saved by the described procedure with the aid of the described calibration and optimization process. As a result, fewer engine test bench and real vehicle roller test bench times P are required.

8th shows in a schematic sequence finally a special feature within the optimization process, which according to the figure 4.8 . 4.9 of the 4 and as shown 7.4 and 7.5 of the 7 has already been shown and explained.

8th illustrates that in the context of the calibration and optimization process K according to the invention at least two optimization algorithms are used during the optimization process, wherein an optimization algorithm can process several 10,000 structure parameters within a practical computing time.

The optimizer used so far 2.3 within the optimization process as shown in the figure 2.4 of the 2 can not optimize more complex optimization structures due to the previously necessary to process data sets for the optimization of a basic map per engine parameters, or perform the optimization in a reasonable time.

For the calibration and optimization process K according to the invention, a special iterative optimization method is carried out, which must be able to process very large amounts of data and, thus, the large number of variations in the data determination of the map structure 4.9 from the basic maps 5.1 and their correction structures 5.2 taking into account the target values Z and abort criteria can be performed.

Thus, only those optimization algorithms come into question that can process several 10,000 structural parameters in a short time.

Since the necessary computing time in the inventive calibration and optimization process K with the conventional optimization algorithm Figure 2.3 of the 2 would increase exponentially, it is proposed according to the invention, within the optimization process for input parameterization an additional optimizer as shown 8.1 with high local convergence.

After determining a sufficiently high number of optimized parameter data sets as shown in the figure 8.2 by means of the optimizer 8.1 With local convergence, the optimization process ends with another "global" optimizer 8.3 a global optimization run is started, which determines the desired global optimum of the parameters of the parameter record.

In the case of local optimization, the user performs a parameterization with few parameterization passes, so that with high probability only a local optimum is found. In the global optimization, the user performs a parameterization with many parameterization runs, so that slowly with high Probability a global optimum is found. However, finding the global optimum with very many parameterization runs is very time-consuming and thus not very practical.

Therefore, the procedure according to the invention is proposed, which consists in providing a globally optimally configured optimizer 8.3 locally "pre-optimized" starting values by means of the optimizer 8.1 be provided with local convergence, thereby advantageously by the combined application of the two optimizers 8.1 . 8.3 the global optimum of the data to be parameterized is found faster. Advantageously, with this procedure, the probability increases with many locally pre-optimized data to generate the desired globally optimized parameter data set in a time-optimized manner.

The parameter data set determined in this way finally leads to an optimized engine control unit data record 8.4 on the basis of the optimization process according to the invention by means of the two optimization algorithms according to the illustration 8.1 and 8.3 ,

The calibration and optimization process K according to the invention is thus changed at different locations compared with the conventional calibration and optimization process K, as a result of which an optimized engine control unit data record is obtained stepwise from the conventional optimized engine control unit data record 2.5 about the improved optimized engine control unit data record Figure 7.6 , to the still further improved ECU dataset 8.4 can be created.

The step-by-step optimized engine control unit dataset 8.4 , which represents the global optimum of the data of the parameter data set, eventually becomes the engine controller functional architecture 7.2 according to 7 stored in the engine control unit MSG.

LIST OF REFERENCE NUMBERS

K

Kalibrierungs- und Optimierungsprozess

Abbildung 1.1

Fahrzeugrollenprüfstand P

Abbildung 1.2

dreidimensionales Motorsimulationsmodell S

Abbildung 1.3

Fahrzyklus F

Abbildung 1.4

Motorsimulationsumgebung U

Abbildung 1.5

Grundkennfeld G

Abbildung 1.6

Bauteiländerung und Funktionsanpassung

Fig. 2

Abbildung 2.1

Fahrzyklus F

Abbildung 2.2

Zielwerte Z

Abbildung 2.3

Optimierer

Abbildung 2.4

optimierte Zielwerte

Abbildung 2.5

optimierter Motorsteuergeräte-Datensatz

Fig. 3

F

Fahrzyklus F

1, 2, 3, 4, 5

Einspritzmuster

PI1

erste Voreinspritzung

PI2

zweite Voreinspritzung

POI3

Nacheinspritzung

t

Zeit

T

Temperatur

Tmot

Motortemperatur

v

Geschwindigkeit

I

Kennlinie (durchgehend) Motortemperatur Tmot

II

Kennlinie (gestrichelt) Einspritzmuster 1, 2, 3, 4, 5

III

Kennlinie (gepunktet) Geschwindigkeit v

Fig. 4

U'

Motorsimulationsumgebung

Abbildung 4.1

Fahrzyklus F

Abbildung 4.2

Motorkennfeld/Steller-Sollwerte

Abbildung 4.3

dynamisches Modell der Steller-Sollwerte des Motors

Abbildung 4.4

Modellierung der Regelabweichung

Abbildung 4.5

temperturabhängiges Emissionsmodell/Verbrauchsmodell

Abbildung 4.6

Modellierung des Temperatureinflusses/Motortemperatur Tmot

Abbildung 4.7

kumulierte Zykluswerte

Abbildung 4.8

Optimierungsprozess des kumulierten Zielwertes Z mittels eines Optimierers

Abbildung 4.9

Kennfeldstruktur

Fig. 5

MSG

Motorsteuergerät

Abbildung 5.1

Grundkennfelder G

Abbildung 5.2

Korrekturstruktur (Korrekturfelder, Korrekturkennlinien)

Cor0

erster Korrekturwert

Cor1

zweiter Korrekturwert

Cor2

dritter Korrekturwert

Cor3

vierter Korrekturwert

Fig. 6

Abbildung 6.1

Grundkennfelder G in Abhängigkeit der Motortemperatur Tmot während des Betriebsmodus PI1, PI2, POI3

Abbildung 6.2

Grundkennfelder G in Abhängigkeit der Motortemperatur Tmot während des Betriebsmodus PI1, POI3

Abbildung 6.3

Grundkennfelder G in Abhängigkeit der Motortemperatur Tmot während des Betriebsmodus PI1, PI2

Fig. 7

Abbildung 7.1

Fahrzyklus F

Abbildung 7.2

Motorsteuergeräte-Funktionsarchitektur

Abbildung 7.3

Motorsimulationsumgebung U' gemäß 4

Abbildung 7.4

Optimierungsprozess eines Zielwertes Z mittels eines Optimierers

Abbildung 7.5

optimierte Zielwerte

Abbildung 7.6

optimierter Motorsteuergeräte-Datensatz auf Basis der Motorsimulationsumgebung U'

Fig. 8

Abbildung 8.1

„lokaler” Optimierer für lokalen Datensatz

Abbildung 8.2

Optimierte Datensätze

Abbildung 8.3

„globaler” Optimierer für den globalen Datensatz

Abbildung 8.4

optimierter Motorsteuergeräte-Datensatz auf der Basis der Motorsimulationsumgebung U' und nach einem Optimierungsvorgang mittels lokalem und globalem Optimierer 8.1 und 8.3

Fig. 1

K

Calibration and optimization process

Figure 1.1

Vehicle roller test bench P

Figure 1.2

three-dimensional engine simulation model S

Figure 1.3

Driving cycle F

Figure 1.4

Engine simulation environment U

Figure 1.5

Basic map G

Figure 1.6

Component change and function adjustment

Fig. 2

Figure 2.1

Driving cycle F

Figure 2.2

Target values Z

Figure 2.3

optimizer

Figure 2.4

optimized target values

Figure 2.5

optimized engine control unit data record

Fig. 3

F

Driving cycle F

1, 2, 3, 4, 5

Injection pattern

PI1

first pilot injection

PI2

second pilot injection

PoI3

injection

t

Time

T

temperature

tmot

engine temperature

v

speed

I

Characteristic (continuous) Motor temperature Tmot

II

Characteristic (dashed) Injection pattern 1 . 2 . 3 . 4 . 5

III

Characteristic (dotted) Speed v

Fig. 4

U '

Engine Simulation Environment

Figure 4.1

Driving cycle F

Figure 4.2

Engine map / controller setpoints

Figure 4.3

dynamic model of the actuator setpoints of the motor

Figure 4.4

Modeling the control deviation

Figure 4.5

temperature-dependent emission model / consumption model

Figure 4.6

Modeling of the temperature influence / engine temperature Tmot

Figure 4.7

cumulative cycle values

Figure 4.8

Optimization process of the cumulative target value Z by means of an optimizer

Figure 4.9

Map structure

Fig. 5

MSG

Engine control unit

Figure 5.1

Basic characteristics G

Figure 5.2

Correction structure (correction fields, correction characteristics)

COR0

first correction value

cor1

second correction value

Cor2

third correction value

cor3

fourth correction value

Fig. 6

Figure 6.1

Basic maps G as a function of the engine temperature Tmot during the operating mode PI1, PI2, POI3

Figure 6.2

Basic maps G as a function of the engine temperature Tmot during the operating mode PI1, POI3

Figure 6.3

Basic maps G as a function of the engine temperature Tmot during the operating mode PI1, PI2

Fig. 7

Figure 7.1

Driving cycle F

Figure 7.2

Engine control unit function architecture

Figure 7.3

Motor simulation environment U 'according to 4

Figure 7.4

Optimization process of a target value Z by means of an optimizer

Figure 7.5

optimized target values

Figure 7.6

optimized engine control unit data record based on the engine simulation environment U '

Fig. 8

Figure 8.1

"Local" optimizer for local record

Figure 8.2

Optimized records

Figure 8.3

"Global" optimizer for the global record

Figure 8.4

optimized engine control unit data record on the basis of the engine simulation environment U 'and after an optimization process using a local and global optimizer 8.1 and 8.3

Claims (13)

Method for the optimal parameterization of complex components and associated complex functional structures, in particular of internal combustion engines and hybrid drives and their associated functional structures, in which at least one target value (Z) within a calibration and optimization process (K) is parameterized step by step by: a) for a drive cycle (F) a measured data record is recorded on a test stand (P), and b) a multi-dimensional engine simulation model (S) is created on the basis of the test data record determined on the test bench (P) and from this a parameter data record, characterized in that c) the parameter data set of the driving cycle (F) is optimized by means of an extended engine simulation environment (U ') and at least one optimum target value (Z) is determined, after which • d) the optimized parameter data record automatically takes into account an optimized engine control unit taking into account the optimum target value (Z) -Record ( 7.6 ) for an engine control unit (MSG) of the internal combustion engine and hybrid drive is transferred. A method according to claim 1, characterized in that in the extended engine simulation environment (U ') in determining the at least one optimal target value (Z) for the optimized in c) parameter data set of the driving cycle (F), • at least one motor parameter and / or • at least an environmental parameter and / or at least one operating strategy parameter ( 1 . 2 . 3 . 4 . 5 ) in at least one dynamic model ( 4.3 ) and subsequently in an emission model ( 4.5 ) and / or a consumption model is taken into account and / or the at least one optimal target value (Z) by means of at least one optimization algorithm ( 8.1 ; 8.3 ) is optimized and determined. A method according to claim 2, characterized in that within the extended engine simulation environment (U ') at least one engine parameter as a simulation parameter of the parameter data set, in particular a turbocharger charging system and / or an injection system and / or a rail pressure line and / or an exhaust gas recirculation system and / or a high - And low-pressure Abgasrückführungssytems and / or a flap system and / or a hybrid system is / are considered. A method according to claim 2, characterized in that within the engine simulation environment (U ') at least one environmental parameter as a simulation parameter of the parameter data set, in particular an ambient pressure and / or an intake temperature is taken into account / are. Method according to claim 2 and 3, characterized in that within the extended engine simulation environment (U ') at least one operating strategy parameter as simulation parameter of the parameter data set, in particular different injection patterns ( 1 . 2 . 3 . 4 . 5 ) of the injection system and / or exhaust gas recirculation modes of the exhaust gas recirculation system are taken into account. Method according to at least one of the preceding claims 2 to 5, characterized in that for the simulation parameters in one associated engine map ( 4.2 ) Setpoint values which are used as input variables for the dynamic model ( 4.3 ), where within the dynamic model ( 4.3 ) a control deviation ( 4.4 ) is modeled over time (t) relative to a controller actual value, the simulation parameters of the parameter data set thus modeled being used as input variables for the models for determining the target value of the optimum target value (Z), in particular for the emission model ( 4.5 ) and / or the consumption model. Method according to claim 6, characterized in that within the dynamic model ( 4.3 ) modeled simulation parameters into the emission model ( 4.5 ) and / or consume the consumption model as input variables, according to which at least one integrated time-resolved accumulated optimum target value (Z), in particular an emission target value and / or a consumption target value, is formed from the modeled simulation parameters of the parameter data record via the drive cycle (F) , Method according to at least one of claims 2 to 7, characterized in that within the extended engine simulation environment (U ') in the optimization of the driving cycle (F) automatically for the respective simulation parameters of the parameter data set according to at least one of claims 3 to 5, by means of at least one Optimization algorithm ( 8.1 . 8.3 ) a map structure ( 4.9 ) is provided with the optimized parameter data set. A method according to claim 8, characterized in that for optimizing the parameter data set of the Fahrzyklusses (F) and for determining the at least one target value (Z) an optimization algorithm ( 8.1 ) with local convergence and a global optimization algorithm ( 8.3 ) are used in combination. Method according to one of the preceding claims, characterized in that the method is carried out in a data processing system. Data processing system, configured for acquiring measured data and for calculating and optimizing parameter data, in particular engine parameters, environmental parameters and operating strategy parameters of an internal combustion engine and hybrid drive, characterized in that the data processing system is adapted to carry out a method according to one of claims 1 to 10. A computer program product, comprising program code means, on a computer-readable medium for carrying out a method according to any one of claims 1 to 10, when the program is executed on a data processing system, in particular a computer. Engine control unit (MSG) with a map structure ( 4.9 ) with an optimized parameter data set ( 7.6 . 8.4 ), which is optimized in a method according to at least one of claims 1 to 10.

Images