Method for operating an internal combustion engine, control device for carrying out a method of this kind, and internal combustion engine comprising a control device of this kind

Abstract

The invention relates to a method for operating an internal combustion engine (1), in which method - at least one combustion parameter for at least one combustion chamber (7) of the internal combustion engine (1) is set according to a rotational-speed value, - at a testing time, a load change which exceeds a first predetermined load-change threshold is identified, - at a prediction time taking place after the testing time at a predetermined testing time interval, an extremal rotational speed predicted for the future on the basis of the prediction time is predicted, and - on the basis of the prediction time, the at least one combustion parameter is set as the rotational-speed value according to the predicted extremal rotational speed.

Description

[0001]Rolls-Royce Solutions GmbH DESCRIPTION Method for operating an internal combustion engine, computer program and control device for carrying out such a method, and internal combustion engine with such a control device. The invention relates to a method for operating an internal combustion engine, a computer program and a control device for carrying out such a method, and an internal combustion engine with such a control device. Internal combustion engines generally face the problem of being able to react to load jumps in a timely manner so that the internal combustion engine does not stall in the event of a sudden load increase, or the engine speed does not overshoot in the event of a sudden load drop. This applies in particular to gas engines with external mixture formation,where the power control by means of a central throttle valve is comparatively slow due to the comparatively long gas transit times from a gas mixer or the throttle valve to the combustion chambers. An additional problem is that, for example, an enrichment of the mixture, i.e. a reduction in the combustion air ratio, carried out in response to a load step, can lead to knocking. Conversely, a leaning of the mixture, i.e. an increase in the combustion air ratio, can lead to misfiring. To date, it has proven difficult, on the one hand, to react sufficiently quickly to a detected load step in this conflicting field of requirements with the control or regulation of the internal combustion engine, and, on the other hand, to detect the load step sufficiently quickly to be able to initiate a suitable reaction at an early stage. The invention is therefore based on the object ofto provide a method for operating an internal combustion engine, a computer program and a control device for carrying out such a method, and an internal combustion engine with such a control device, wherein the aforementioned disadvantages are at least reduced, preferably not occurring. The object is achieved by providing the present technical teaching, in particular the teaching of the independent claims and the preferred embodiments disclosed in the dependent claims and the description. The object is achieved in particular by providing a method for operating an internal combustion engine, wherein at least one combustion parameter for at least one combustion chamber of the internal combustion engine is set as a function of a speed value, wherein at a test time, a load change is detected that exceeds a first predetermined load change threshold,wherein, at a prediction time a predetermined test time interval after the test time, an extreme speed predicted for the future based on the load change from the prediction time is predicted, and wherein, starting from the prediction time, the at least one combustion parameter is set as a function of the predicted extreme speed as the speed value. In particular, by predicting the extreme speed at the prediction time and then, starting from the prediction time, the at least one combustion parameter is no longer set as a function of the actual instantaneous speed value but rather as a function of the predicted extreme speed, an effective reaction to the detected load change can advantageously be carried out early on.which, based on the prediction of the extreme speed, also takes into account the extent of the load change and thus causes an adjustment of the combustion parameter at an early point in time that is adapted to this extent. Thus, the combustion parameter is adjusted earlier and more strongly to the actual load change than with a setting of the combustion parameter that depends solely on the current speed, in which case a corresponding adjustment would only be made well after the prediction time, when an actual extreme speed is reached. The fact that at least one combustion parameter is adjusted as a function of the predicted extreme speed as the speed value means, in the context of the present technical teaching, in particular that at least one - preferably quantitatively significant - contribution of the combustion parameter is adjusted as a function of the predicted extreme speed. It is possible,that the combustion parameter is adjusted entirely as a function of the predicted extreme speed; however, it is also possible for the combustion parameter to have at least two contributions, preferably additively linked to one another, of which at least one is then adjusted as a function of the predicted extreme speed, while at least one other contribution can simultaneously be adjusted as a function of the instantaneous speed. In particular, a first contribution can be read from a basic characteristic map as a function of the instantaneous speed, while a second contribution can be read from a differential characteristic map as a function of the predicted extreme speed, as explained in more detail below. In one embodiment, during steady-state operation of the internal combustion engine, only the first contribution read from the basic characteristic map is used to adjust the combustion parameter.while in the case of a detected load change, the second contribution read from the difference characteristic map is also taken into account. In one embodiment, the extreme speed is predicted based on a speed gradient at the prediction time. It is possible for the extreme speed or an extreme speed change, from which the extreme speed is subsequently determined, to be determined based on a characteristic curve or read from a characteristic map depending on the speed gradient. The characteristic curve or the characteristic map are preferably determined in tests, in particular on the test bench. In one embodiment, the extreme speed is explicitly predicted, preferably by predicting an extreme speed change, i.e., an extreme speed difference or synonymously, an extreme difference speed, at the prediction time.from which the extreme speed is calculated, in particular based on a current speed present at the prediction time – or a current speed difference from a target speed. However, it is also possible for the extreme speed to be predicted implicitly, preferably by predicting the extreme speed change and using it as the speed value. This is particularly the case when the at least one combustion parameter is set as a function of a speed difference from the target speed as the speed value, in particular by reading the combustion parameter itself or a contribution to the combustion parameter, in particular the second contribution, from a characteristic map as a function of the speed difference. However, since speeds and speed differences or difference speeds can ultimately be converted into one another at any time,These different embodiments are to be regarded as analogous and equivalent to one another. In one embodiment, an internal combustion engine is operated which has a plurality of combustion chambers. In this case, the at least one combustion parameter is set for all combustion chambers of the internal combustion engine as a function of the speed value, preferably globally – i.e., identically for all combustion chambers. In the context of the present technical teaching, a test time is understood in particular to be a time at which a load change is detected. The test time is therefore not a specific, fixed time, but is then determined as the starting point for the further method, in particular as the starting point for the test time interval, i.e., for calculating the prediction time when the load change is detected. It is possible that, during operation of the internal combustion engine, continuously, at predetermined times or at predetermined time intervals,or on demand, it is checked whether a load change has occurred; if a load change is detected at a detection time during this test, this detection time is set as the test time. The prediction time is calculated by adding the test time interval to the test time; the prediction time is therefore around the test time interval after the test time. This advantageously ensures that the extreme speed can be predicted as reliably as possible. While it is generally possible to detect relatively early – at the test time – whether a load change has occurred at all, it is typically only possible with a certain time delay to predict with sufficient certainty the extent of the load change.on which the extreme speed ultimately depends. For example, speed curves associated with different load changes are relatively similar at the beginning of the load changes and only diverge significantly from one another as time progresses. The test time interval is preferably selected such that it reflects the described time delay between the recognition that a load change is present at all and the possibility of predicting the extent of the load change with sufficient certainty. The test time interval is preferably determined in experiments, in particular on the test bench. In one embodiment, the test time interval is from 20 ms to 100 ms, in particular from 30 ms to 80 ms, in particular from 40 ms to 60 ms, in particular 50 ms. A load change in the context of the present technical teaching is understood to mean, in particular, a change in the speed and / or torque of the internal combustion engine. It is possible,that only one of these two parameters is used to detect the load change. Alternatively, it is possible for both parameters to be used cumulatively to detect the load change. In one embodiment, a change in the power of the internal combustion engine can also be used. In one embodiment, a separate first load change threshold is defined for the speed and the torque. The fact that the load change exceeds the first load change threshold means, in particular, that the load change exceeds the first load change threshold in terms of amount. It is possible that different first load change thresholds are predetermined for a load increase and a load shedding, on the one hand, but it is also possible that the same first load change threshold is predetermined for the load increase and the load shedding. In one embodiment, only the value of the extreme speed,However, the future point in time at which this value is reached is not predicted. For the rapid adaptation of the control or regulation of the internal combustion engine to the load change, it is not necessary to know the point in time at which the extreme speed is expected to be reached; rather, it is sufficient to adapt the operation of the internal combustion engine to the expected extent of the load change at an early stage by adjusting the at least one combustion parameter based on the prediction time as a function of the extreme speed. In the context of the present technical teaching, the fact that the at least one combustion parameter is adjusted based on the prediction time means, in particular, that the at least one combustion parameter can be adjusted optionally - in particular depending on the type of load change and / or depending on the specific combustion parameter - immediately from the prediction time,in particular already at the prediction time, or with a predetermined time delay from the prediction time. In particular, in one embodiment - as explained in detail below - a combustion air ratio is set as the combustion parameter immediately from the prediction time, regardless of the type of load change, wherein an ignition point is set as the combustion parameter with the predetermined time delay from the prediction time in the case of a load increase and immediately from the prediction time in the case of a load shedding. According to a further development of the invention, it is provided that the at least one combustion parameter is selected from a group consisting of a combustion air ratio,an ignition timing and a combination of the combustion air ratio and the ignition timing. Advantageously, by adjusting the combustion air ratio early in relation to the predicted extreme engine speed, a rapid response to the load change can be achieved or at least supported. By adjusting the ignition timing early in relation to the predicted extreme engine speed, knocking combustion can be avoided, or load shedding can be supported. It is particularly advantageous if both the combustion air ratio and the ignition timing are adjusted accordingly as the at least one combustion parameter. In the context of the present technical teaching, a combustion air ratio is understood to mean, in particular, a so-called lambda value.That is, a mass ratio of combustion air to fuel relative to a stoichiometrically ideal ratio for a theoretically complete combustion process. A combustion air ratio of 1 corresponds to a stoichiometric mixture, a combustion air ratio greater than 1 characterizes a lean mixture with excess air, and a combustion air ratio less than 1 characterizes a rich mixture with excess fuel. Accordingly, an increase in the combustion air ratio corresponds to a leaner mixture, and a decrease in the combustion air ratio corresponds to a richer mixture. In the context of the present technical teaching, an ignition point is understood in particular to be a point in time within a combustion chamber working cycle—preferably specified in degrees of crankshaft angle—at which an ignition of the mixture in the combustion chamber occurs.The ignition timing is preferably specified in relation to a top dead center associated with the combustion stroke, or synonymously the power stroke or expansion stroke, of the working cycle, this top dead center also being referred to as ignition TDC. Typically, the ignition timing is temporally before the ignition TDC in the compression stroke. Accordingly, a retarded ignition timing shift means a shift of the ignition timing toward the ignition TDC if the ignition timing is in the compression stroke, or a shift of the ignition timing away from the ignition TDC if the ignition timing is in the power stroke; correspondingly, an advanced ignition timing shift means a shift of the ignition timing away from the ignition TDC if the ignition timing is in the compression stroke, or a shift of the ignition timing toward the ignition TDC if the ignition timing is in the power stroke. According to a further development of the invention,that the first predetermined load change threshold is selected such that it is characteristic of an instantaneous load change. The advantages of the method are realized particularly in the case of instantaneous load changes. In one embodiment, the method described here is only carried out when an instantaneous load change occurs due to the correspondingly selected first predetermined load change threshold. If, on the other hand, there is a gradual or progressive load change, also referred to as a load ramp, the internal combustion engine is operated conventionally, i.e., in the case of a load ramp, the at least one combustion parameter is set, in particular, not as a function of a predicted extreme speed, but exclusively and at any time as a function of the actual instantaneous speed. In the context of the present technical teaching, an instantaneous load change is understood to mean, in particular, a load changein which a change between two stationary load points takes place within a time window of at most 300 ms, preferably of at most 200 ms, preferably of at most 150 ms, preferably of at most 100 ms, preferably of at most 50 ms. According to a further development of the invention, the combustion air ratio and the ignition timing are set as the at least one combustion parameter. In particular, these two combustion parameters, in combination with one another, are particularly suitable for setting the internal combustion engine,particularly in response to a load change. In one embodiment, the ignition timing—particularly during a load increase—is set relative to the setting of the combustion air ratio with the predetermined time delay. Preferably, in one embodiment, the ignition timing is set relative to the setting of the combustion air ratio with the predetermined time delay only during a load increase—and not during a load shedding. In contrast, in this embodiment, the ignition timing is set without the predetermined time delay during a load shedding. In one embodiment, the predetermined time delay corresponds to a gas transit time in an air path of the internal combustion engine from a gas mixer to the at least one combustion chamber. In this way, a dead time is advantageously taken into account when setting the combustion air ratio in the combustion chamber.wherein the combustion air ratio, following a change in the control of the gas mixer in the combustion chamber, only changes accordingly after the gas flow time has elapsed, i.e., only when the correspondingly changed mixture actually arrives in the combustion chamber. In particular, if the ignition timing is adjusted to avoid knocking combustion in the event of enrichment of the mixture, this adjustment is advantageously only carried out when the mixture in the combustion chamber is actually enriched accordingly. Preferably, the predetermined time delay is determined in experiments, in particular on a test bench. In one embodiment, the predetermined time delay is defined for each combustion chamber individually when operating an internal combustion engine with a plurality of combustion chambers, whereby different gas flow times from the gas mixer to the individual combustion chambers can advantageously be taken into account. According to a further development of the invention,that the predetermined time delay is only applied if and / or as long as the load change exceeds a second predetermined load change threshold. This is based on the idea that the composition of the mixture in steady states – or even when a return to a steady state is already apparent – ​​is changed only slowly, i.e., on comparatively long time scales, in which case the gas transit time from the gas mixer to the at least one combustion chamber can be neglected; in fact, the gas transit time is only relevant during a highly transient phase of the load change. This highly transient phase can be characterized by the load change exceeding the second predetermined load change threshold. The load change exceeds the second predetermined load change threshold even during an instantaneous load change for a specific load change time interval,which is essentially determined by the dynamic behavior of the sensors used to detect the load change and the filtering of signals from these sensors. Even if a load gradient describing the load change were hypothetically to correspond exactly to a delta distribution or Dirac function, i.e., were to differ from zero only in an infinitely short time interval, the load gradient detected in a real control device of the internal combustion engine would differ from zero in a finite time interval due to the dynamic behavior of the sensors and the signal filtering, and would also exceed the second load change threshold for a finite time interval. In one embodiment, the second predetermined load change threshold is smaller than the first predetermined load change threshold. In one refinement, the comparison with the second load change threshold is performed with a sign.and the second load change threshold itself is accordingly assigned a positive sign or is equal to zero. In particular, at least in one embodiment, the second load change threshold is used to essentially check whether the load change is a load connection or a load shedding, wherein the predetermined time delay is applied in the case of load connection and not applied in the case of load shedding - wherein, as explained in more detail below, in the case of load shedding the combustion air ratio is preferably only changed slowly, even in the transient state. In principle, it would be sufficient for this purpose if the second load change threshold were selected to be zero; however, in particular with a view to ensuring sufficient discrimination between the different cases, a positive,A value at least slightly different from zero is preferred for the second load change threshold. According to a further development of the invention, it is provided that the at least one combustion parameter is no longer set as a function of the predicted extreme speed after the prediction time, starting from an extreme time at which a current speed passes through an extreme, but as a function of the current speed as the speed value. This advantageously ensures, on the one hand, that the load change is correctly reflected in the result, in particular if the predicted extreme speed deviates from the actual extreme speed, and, on the other hand, it is achieved that the at least one combustion parameter, which is then set as a function of the actual current speed, is returned to a predetermined steady-state behavior after the passage through the extreme.when the speed approaches the target speed again. In the context of the present technical teaching, the extreme time is understood in particular to be the time at which the current speed passes through its actual extreme. In one embodiment, the at least one combustion parameter is no longer set as a function of the predicted extreme speed after the prediction time, starting from the extreme time plus the predetermined time delay, but rather as a function of the current speed as the speed value. This advantageously applies in particular to the ignition time in the case of load application. In this case, the predetermined time delay is preferably also taken into account at the extreme time. The fact that the at least one combustion parameter is no longer set as a function of the predicted extreme speed but rather as a function of the current speed as the speed value,In the context of the present technical teaching, this means, in particular, that each contribution of the combustion parameter – regardless of the number of contributions – is adjusted depending on the current speed. In particular, the second contribution is then also read from the differential characteristic map depending on the current speed. Thus, in a preferred embodiment, three scenarios can be distinguished: Firstly, during steady-state operation of the internal combustion engine, only the first contribution read from the basic characteristic map is used to adjust the combustion parameter; secondly, in the case of a detected load change, starting from the prediction time to the extreme time – optionally plus the predetermined time delay in each case – the second contribution read from the differential characteristic map depending on the predicted extreme speed is taken into account in addition to the first contribution from the basic characteristic map.and thirdly, from the extreme point in time – optionally plus the predetermined time delay in each case – in addition to the first contribution from the basic characteristic map, the second contribution, now also read from the difference characteristic map as a function of the current engine speed, is taken into account. According to a further development of the invention, if the load change is a load increase, a combustion air ratio and an ignition timing are set as the at least one combustion parameter depending on the predicted extreme engine speed as the engine speed value. Advantageously, this allows a particularly efficient and early response to the load increase, while on the other hand, the load change of the internal combustion engine can be optimally supported by appropriate adjustment of both combustion parameters. Alternatively or additionally, if the load change is a load shedding,As the at least one combustion parameter – at least in the end – only the ignition timing is adjusted as a function of the predicted extreme speed as the speed value. It has been found that, during load shedding, it can be advantageous not to adjust the combustion air ratio as a function of the predicted extreme speed, at least in the end, particularly with a view to high combustion stability. Rather, in one embodiment, the combustion air ratio during load shedding is preferably adjusted continuously as a function of the – actual – instantaneous speed as the speed value. The fact that the combustion air ratio is not adjusted as a function of the predicted extreme speed implies that the second contribution is read from the difference map as a function of the predicted extreme speed,However, the difference characteristic map is constantly set to zero for positive deviations from the target speed, i.e., specifically in the case of load shedding, so that the second contribution is consistently zero during load shedding. Thus, only the first contribution, read from the basic characteristic map as a function of the actual speed, contributes to the combustion air ratio, which in turn means that the combustion air ratio is ultimately only adjusted as a function of the actual speed. In another embodiment, it is possible that, in the case of load shedding, only the first contribution from the basic characteristic map is taken into account for the combustion air ratio. In this embodiment, the combustion air ratio is then not only ultimately adjusted as a function of the predicted extreme speed, but also practically not adjusted as a function of the predicted extreme speed. In yet another embodiment, it is possiblethat the second contribution is taken into account, but the difference map for positive deviations from the target speed is specified differently than previously described—with values ​​other than zero—so that the combustion air ratio is actually adjusted depending on the predicted extreme speed even in the case of load shedding. In this case, regardless of the type or direction of the load change, both the combustion air ratio and the ignition timing are adjusted depending on the predicted extreme speed as the speed value. According to a further development of the invention, it is provided that the at least one combustion parameter is additionally adjusted depending on a—momentary—torque. This is based on the idea thatthat the load point of the internal combustion engine is characterized by the two load parameters speed and torque. In one embodiment, the at least one combustion parameter is read from a characteristic map as a function of the torque. This represents a particularly simple and thus advantageous embodiment of the method. According to a further development of the invention, it is provided that the at least one combustion parameter is set by determining a base combustion value, as the above-mentioned first contribution, for the at least one combustion parameter as a function of the speed value - and optionally additionally as a function of the torque. Further, a differential combustion value, as the above-mentioned second contribution, for the at least one combustion parameter is determined as a function of the speed value - and optionally additionally as a function of the torque. The base combustion value is offset against the differential combustion value.to obtain an instantaneous value—in particular a transient value—for the at least one combustion parameter, to which the at least one combustion parameter is set. In one embodiment, the base combustion value is read from the base characteristic map. Alternatively or additionally, the differential combustion value is read from the differential characteristic map. The base combustion value preferably corresponds to a stationary value for the at least one combustion parameter, i.e., a value to be applied at a stationary load point as a function of the engine speed and optionally the torque. Accordingly, the base characteristic map is a stationary characteristic map, i.e., a characteristic map with values ​​to be applied at stationary load points. Alternatively or additionally, the differential combustion value takes into account predetermined deviations of the at least one combustion parameter during transient operation. Accordingly, the differential characteristic map is a characteristic map,which is data-based on values ​​that represent the corresponding deviations. In steady-state operation, as explained above, the base combustion value is preferably used alone as the steady-state instantaneous value, meaning that the second contribution is not read out but set to zero or, equivalently, omitted. In particular, in the case of a detected load change, i.e., transient operation, the differential combustion value is added to the base combustion value to obtain the instantaneous value. If the instantaneous value of the combustion parameter, which is preferably dependent on speed n and torque M, is generally denoted by p(n,M), the transient instantaneous value ptrans(n,M) is given by: ptrans(n,M) = pstat(n,M) + ^p(n,M), (1) with the base combustion value pstat(n,M) as the first contribution and the differential combustion value ^p(n,M) as the second contribution. In stationary operation, the –stationary – instantaneous value p(n,M) is given as p(n,M) = pstat(n,M). (2) The combustion parameter p is in particular the combustion air ratio ^ or the ignition timing Z. According to a further development of the invention, it is provided that during a load increase as the load change starting from a first, lower stationary load to a second, higher stationary load as the at least one combustion parameter, the combustion air ratio and the ignition timing are adjusted, wherein the combustion air ratio is increased starting from a first, lower combustion air ratio at the first stationary load to a second, higher combustion air ratio at the second stationary load. In principle, the mixture is therefore preferably leaned out towards higher loads in stationary operation, in particular to avoid knocking combustion and to keep emissions within legal limits. In particular, the basic characteristic map for the combustion air ratio is data-based such thatthat the combustion air ratio increases with increasing torque – at least globally; local deviations are possible. The dependence on the speed is typically more complex and can increase or decrease with increasing speed, particularly depending on the torque. However, it is also possible that in other embodiments, the basic characteristic map is differently data-based, and thus the combustion air ratio also exhibits a different steady-state behavior depending on speed and / or torque. However, during a load increase, the combustion air ratio is preferably – transiently – temporarily reduced relative to the first combustion air ratio starting from the prediction time based on the setting depending on the predicted extreme speed as the speed value. This means that the mixture is enriched relative to its composition at the corresponding steady-state load point, but preferably also absolutely.whereby both early and rapid adaptation to the load point is advantageously achieved and, in particular, the target speed is reached again as quickly as possible after an initial drop in speed. Therefore, the differential characteristic map for the combustion air ratio is particularly configured such that it includes negative differential combustion values ​​in the event of load application, i.e., in particular, for a drop in speed and / or an increase in torque. The adjustment as a function of the predicted extreme speed results in a particularly early and particularly pronounced reduction in the combustion air ratio. It is also possible that, in other embodiments, the differential characteristic map is configured differently and, thus, the combustion air ratio also exhibits a different transient behavior depending on speed and / or torque. The ignition timing is determined based on a first,later ignition timing at the first steady-state load is advanced to a second, earlier ignition timing at the second steady-state load. Thus, the ignition timing is adapted to the steady-state leaning of the mixture, and the efficiency of the internal combustion engine is high at the steady-state operating points. The basic characteristic map for the ignition timing is thus preferably at least locally datad such that the ignition timing is advanced with increasing torque, with this preferably being the case particularly at higher torques. It is possible that this dependency is less pronounced at medium torques, or that the ignition timing even remains constant, with the relationship preferably even being reversed at low torques, particularly in the idle range.The ignition timing is typically advanced there with decreasing torque. In particular, the ignition timing is typically maximally advanced at idle and, starting from idle, is retarded with increasing torque. However, it is also possible that in other embodiments the basic characteristic map is differently dataset and thus the ignition timing also exhibits a different steady-state behavior depending on the engine speed and / or torque. During load application, the ignition timing is now preferably – transiently – starting from the predicted time – in particular with the predetermined time delay – temporarily retarded further relative to the first later ignition timing due to the setting as a function of the predicted extreme engine speed as the engine speed value. This is an advantageous setting with regard to the – relative or preferably also absolute – enrichment of the mixture,particularly with a view to avoiding knocking combustion. Therefore, the differential map for the ignition timing is specifically configured such that it includes negative differential combustion values ​​in the event of load application, i.e., particularly for a drop in engine speed and / or an increase in torque. The adjustment as a function of the predicted extreme engine speed results in a particularly early and particularly pronounced retardation of the ignition timing. However, it is possible that the differential map is configured such that the ignition timing—particularly as a function of engine speed, and especially in the case of significant negative deviations from the target engine speed—is advanced at least slightly with increasing torque, and optionally, in the idle range, also with decreasing torque. It is also possible,that in other embodiments, the differential map is differently data-based and thus the ignition timing also exhibits a different transient behavior depending on the engine speed and / or torque. Alternatively or additionally, during a load shedding as the load change starting from a third, higher stationary load to a fourth, lower stationary load than the at least one combustion parameter, the combustion air ratio and the ignition timing are adjusted, wherein the combustion air ratio is reduced starting from a third, higher combustion air ratio at the third stationary load to a fourth, lower combustion air ratio at the fourth stationary load. During load shedding, the mixture is thus enriched,in particular to ensure high combustion stability. In a preferred embodiment, the combustion air ratio reaches its globally lowest value at idle. The basic characteristic map for the combustion air ratio is, in particular, data-based such that the combustion air ratio decreases with decreasing torque – at least globally, local deviations are possible; in this case, too, the dependence on the speed is typically more complex. Preferably, the difference characteristic map for the combustion air ratio is data-based with the value zero for load shedding, i.e., in particular, in the case of a positive deviation of the instantaneous speed from the target speed, regardless of the torque. This means, in particular, that for load shedding, the above equations (1) and (2) are identical in their results. It is also possible,that in other embodiments, the base characteristic map and / or the differential characteristic map are differently data-driven, and thus the combustion air ratio exhibits different behavior depending on the engine speed and / or torque. In particular, however, the base characteristic map for the combustion air ratio is identical regardless of the direction of the load change—that is, for load application and load shedding. The ignition timing is preferably retarded, starting from a third, earlier ignition timing at the third steady-state load, to a fourth, later ignition timing at the fourth steady-state load, and thus adapted to the enrichment of the mixture—in particular with a view to achieving the highest possible efficiency. The base characteristic map for the ignition timing is thus preferably at least locally data-driven such that the ignition timing is retarded as the torque decreases.This is preferably the case, in particular, at higher torques. It is possible that this dependency is less pronounced at medium torques, or that the ignition timing even remains constant. The relationship is preferably reversed at low torques, particularly in the idle range, where the ignition timing is typically advanced as the torque decreases. The basic characteristic map for the ignition timing is identical, in particular, regardless of the direction of the load change—that is, for load application and load shedding. However, the ignition timing is preferably retarded even further from the predicted time point based on the setting as a function of the predicted extreme speed as the speed value, preferably even temporarily retarded relative to the fourth ignition timing. This advantageously temporarily reduces the efficiency of the internal combustion engine.so that the load can be shed more quickly and the initially excessive instantaneous speed drops back to the target speed more quickly. Thus, the differential map for the ignition timing is particularly configured such that, in the event of load shedding, i.e., in particular, for an increase in speed and / or a decrease in torque, it also includes negative differential combustion values. The adjustment as a function of the predicted extreme speed results in a particularly early and particularly pronounced retardation of the ignition timing. It is also possible that, in other embodiments, the differential map is configured differently, and thus the ignition timing also exhibits a different transient behavior depending on the speed and / or torque. According to a further development of the invention,that the load change is detected by comparing at least one load parameter gradient with a predetermined gradient threshold as the first load change threshold, wherein the at least one load parameter gradient is selected from a group consisting of a speed gradient, a torque gradient, a power gradient, and a combination of the aforementioned gradients. Alternatively or additionally, a separate gradient threshold is defined for the speed gradient and for the torque gradient. These typically differ due to the different units and value ranges of the various load parameters. It is possible that the value of the respective gradient threshold is selected differently for the case of load shedding than for the case of load application. However, it is also possible for the gradient threshold to have the same value in both cases. In one embodiment,in which both the speed gradient and the torque gradient are used to detect the load change, the load change is preferably only detected when the respective gradient threshold value is exceeded cumulatively for both load parameter gradients. The term "exceeded" preferably refers, regardless of the sign, to an absolute value exceeding the threshold value. This advantageously ensures that the load change can be reliably detected. For example, this prevents a very short-term load fluctuation that does not lead to a relevant speed change from being detected as a load change within the meaning of the method proposed here; this contributes to smooth running of the internal combustion engine, which continues to operate in a stationary manner even with only short-term load fluctuations. According to a further development of the invention, it is provided that the internal combustion engine is a gas engine,in particular with external mixture formation. In this type of internal combustion engine, the advantages of the method are realized in a very special way. Power control in steady-state conditions is preferably achieved by controlling a throttle valve, i.e., in particular, by opening or closing the throttle valve, thus ultimately by varying the mixture mass supplied to the at least one combustion chamber. The combustion air ratio and the ignition timing are preferably adjusted in steady-state conditions by a knock control and / or an emissions control system. The combustion air ratio is adjusted, in particular, by controlling the gas mixer; the ignition timing is adjusted, in particular, by controlling at least one ignition device assigned to the at least one combustion chamber. In transient conditions, the throttle valve is preferably shifted to an extreme position,In particular, the combustion air ratio is preferably adjusted to support the load change during load application and to be fully closed during load shedding. The combustion air ratio is preferably adjusted in such transient conditions to support the load change: During load application, the combustion air ratio is advantageously reduced, thus enriching the mixture, in order to supply additional energy to at least one combustion chamber and to regulate the initially falling engine speed as quickly as possible; towards lower load points, the combustion air ratio is preferably reduced to ensure high combustion stability. The ignition timing is preferably also adjusted to support the load change during transient conditions: During load application, the ignition timing is preferably retarded to prevent knocking combustion while simultaneously enriching the mixture; during load shedding, the ignition timing is preferably also retarded.to specifically reduce the efficiency of the internal combustion engine and thus accelerate load shedding. In one embodiment, the internal combustion engine is drive-connected to an electric machine. In this case, the electric machine is preferably operated as a generator and driven by the internal combustion engine. In this case, the torque or power is advantageously determined based on at least one electrical parameter of the electric machine. This represents a simple and reliable way of determining the instantaneous torque or power of the internal combustion engine. The object is also achieved by creating a computer program that comprises machine-readable instructions, wherein a method according to the invention or a method according to one or more of the previously described embodiments is carried out on the basis of the instructions, if the computer program is run on a computing device,in particular a control device according to the invention or a control device according to one or more of the embodiments described below. In connection with the computer program, in particular those advantages arise which have already been described previously in connection with the method. According to a further aspect, the object is also achieved by creating an electronic storage device on which a computer program according to the invention or a computer program according to one or more of the embodiments described above is stored. In connection with the electronic storage device, in particular those advantages arise which have already been described previously in connection with the method and the computer program. The object is also achieved by creating a control device for an internal combustion engine,which is designed to carry out a method according to the invention or a method according to one or more of the previously described embodiments. In connection with the control device, in particular those advantages arise which have already been described previously in connection with the method, the computer program and the electronic storage device. The object is also achieved by creating an internal combustion engine which has a control device according to the invention or a control device according to one or more of the previously described embodiments. In connection with the internal combustion engine, in particular those advantages arise which have already been described previously in connection with the method, the computer program,the electronic storage device and the control device. In one embodiment, the control device is operatively connected to the throttle valve and the at least one ignition device assigned to the at least one combustion chamber of the internal combustion engine for their respective control. According to a further development of the invention, it is provided that the internal combustion engine is designed as a gas engine, in particular with external mixture formation and / or for operation with a methane- or hydrogen-containing fuel gas as fuel,is formed. In one embodiment, the control device is additionally operatively connected to a gas mixer of the internal combustion engine for its control. In one embodiment, the internal combustion engine is drive-connected to an electric machine. In one embodiment, the control device is additionally operatively connected to the electric machine for its control and / or for detecting at least one electrical parameter of the electric machine. According to a further development of the invention, it is provided that the internal combustion engine is drive-connected to an electric machine operable as a generator for driving the electric machine. In this way, electrical power can be provided with the arrangement of the internal combustion engine and the electric machine,by driving the electric machine by the internal combustion engine. Such an arrangement of an internal combustion engine with an electric machine is also referred to as a genset. In one embodiment, the control device is operatively connected to the electric machine and is configured to detect a torque or power of the internal combustion engine based on at least one electrical parameter of the electric machine—in particular, a time-dependent one. The invention is explained in more detail below with reference to the drawing. Figure 1 shows a schematic representation of an embodiment of an internal combustion engine.which is drive-connected to an electric machine; Figure 2 shows a schematic representation of an embodiment of a method for operating the internal combustion engine; Figure 3 shows a representation of a basic characteristic map for the combustion air ratio as a combustion parameter; Figure 4 shows a representation of a difference characteristic map for the combustion air ratio as a combustion parameter; Figure 5 shows a representation of a basic characteristic map for the ignition timing as a combustion parameter; Figure 6 shows a representation of a difference characteristic map for the ignition timing as a combustion parameter; Figure 7 shows a schematic representation of the embodiment of the method based on various time-dependent parameter curves for a load application,and Figure 8 shows a schematic representation of the exemplary embodiment of the method based on various time-dependent parameter curves for load shedding. Figure 1 shows a schematic representation of an exemplary embodiment of an internal combustion engine 1 that is drive-connected to an electric machine 3, in particular to drive the electric machine 3. The electric machine 3 can be operated as a generator. Such an arrangement comprising the internal combustion engine 1 with the electric machine 3 is also referred to as a genset. The internal combustion engine 1 has an engine block 5 with at least one combustion chamber 7, wherein only four combustion chambers 7 are shown here by way of example and, for the sake of clarity, only one of the combustion chambers 7 is provided with a reference numeral. The internal combustion engine 1 can, in particular, also have more than four combustion chambers 7, for example eight, ten, twelve, sixteen,eighteen or twenty combustion chambers 7. The internal combustion engine 1 has an air path 9 in which a gas mixer 11 is arranged, which is designed to mix combustion air 13 with a fuel 15, in particular a fuel gas containing, for example, methane or hydrogen. The gas mixer 11 is arranged upstream of a branch of the air path 9 to the individual combustion chambers 7. A throttle valve 17 is arranged between the gas mixer 11 and the branch. If the internal combustion engine 1 has a plurality of combustion chamber groups or cylinder banks, each combustion chamber group or cylinder bank is preferably assigned a separate air path 9 with a separate gas mixer 11 and a separate throttle valve 17. Each combustion chamber 7 is assigned an ignition device 19 for igniting the combustible mixture of fuel 15 and combustion air 13 arranged in the combustion chamber 7.For the sake of clarity, only one of the ignition devices 19 is designated by a reference numeral. The internal combustion engine 1 also has a control device 21 configured to carry out a method described below. For this purpose, the control device 21 is preferably operatively connected to the gas mixer 11, the throttle valve 17, and the at least one ignition device 19, in particular to each ignition device 19 of the plurality of ignition devices 19, for controlling them. In particular, the control device 21 is configured to adjust the combustion air ratio by controlling the gas mixer 11 and to adjust the ignition timing by controlling the ignition devices 19. The control device 21 is also operatively connected to a speed sensor 23 for detecting a speed of the internal combustion engine 1, as well as to the electric machine 3.to detect at least one electrical parameter of the electric machine 3. The control device 21 is particularly configured to determine a torque of the internal combustion engine 1 based on the at least one electrical parameter. Fig. 2 shows a schematic representation of an embodiment of a method for operating the internal combustion engine 1. Identical and functionally identical elements are provided with the same reference numerals in all figures, so that reference is made to the preceding description in each case. Within the scope of the method for operating the internal combustion engine 1, at least one combustion parameter p for the at least one combustion chamber 7 is set as a function of a speed value n, wherein at a test time, a load change ^L is detected that exceeds a first predetermined load change threshold ^LS.wherein, at a prediction time lying a predetermined test time interval after the test time, an extreme speed next,pr predicted for the future based on the load change starting from the prediction time is predicted, and wherein, starting from the prediction time, the at least one combustion parameter p is set as a function of the predicted extreme speed next,pr as the speed value. Specifically, in the exemplary embodiment presented here, a first step S1 checks whether the load change ^L exceeds the first predetermined load change threshold ^LS. If this is not the case, steady-state operation is assumed for the internal combustion engine 1 in a second step S2, and the at least one combustion parameter p(n,M), which here depends on the speed n on the one hand and on the torque M on the other, is set by determining a base combustion value pstat(n,M) is read from a basic characteristic map and assigned to the at least one combustion parameter p(n,M) as a stationary instantaneous value according to equation (2). The method then continues in the first step S1. If, however, it is determined in the first step S1 that the load change ^L exceeds the first predetermined load change threshold ^LS, the time at which the exceeding of the first predetermined load change threshold ^LS is determined in the first step S1 is recorded as the test time. Secondly, in a third step S3, transient operation is assumed for the internal combustion engine 1, and the at least one combustion parameter p(n,M) is then adjusted by adding a differential combustion value ^p(n,M) read from a differential characteristic map as a function of the instantaneous speed n and the instantaneous torque M to the base combustion value pstat(n,M) read from the basic characteristic map.wherein the resulting sum value is assigned to the at least one combustion parameter p(n,M) as a transient instantaneous value ptrans(n,M) according to equation (1). Subsequently, in a fourth step S4, it is checked whether, starting from the recorded test time, the test time interval has already elapsed and thus the prediction time has arrived; if this is not the case, the method continues in the third step S3; otherwise, at the prediction time, the method continues in a fifth step S5, in which the extreme speed predicted for the future next,pr is predicted. This is preferably done by determining a current speed gradient at the prediction time, wherein, based on the current speed gradient, an extreme speed change reached in the future starting from the current speed at the prediction time is determined and added to the current speed at the prediction time.from which the extreme speed n predicted for the future, ext,pris obtained.Then, in a sixth step S6, the at least one combustion parameter p(n,M) is set by adding the difference combustion value ^p(next,pr,M) read from the difference characteristic map as a function of the predicted extreme speed next,pr – instead of the instantaneous speed n – and the instantaneous torque M to the base combustion value pstat(n,M) read from the basic characteristic map, wherein the resulting sum value is in turn assigned to the at least one combustion parameter p(n,M) as a transient instantaneous value ptrans(n,M) according to equation (1).In a seventh step S7, it is checked whether the instantaneous speed has reached a turning point, i.e., an actual extreme speed next.If this is not the case, the method continues in the sixth step S6; otherwise, the method continues in an eighth step S8, in which the at least one combustion parameter p(n,M) is again calculated completely based on the actual instantaneous speed n, as in the third step S3. Then, in a ninth step S9, a check is made as to whether a transient state or already steady-state, regulated operation exists. For this purpose, it can be checked, for example, whether the speed has again approached a target speed within a predetermined range, or, equivalently, whether a speed difference from the target speed is smaller than a predetermined limit value; alternatively or additionally, an internally calculated torque can be used.If a transient condition still exists, the method continues in the eighth step S8; otherwise, in a tenth step S10, the differential combustion value ^p(n,M) is set to zero using a filter, and the method continues in the second step S2 after a time delay determined by the filter. In this way, the method preferably runs continuously throughout the operation of the internal combustion engine 1. The at least one combustion parameter p is preferably a combustion air ratio ^or an ignition timing Z, or both the combustion air ratio ^and the ignition timing Z are used as combustion parameters p. In the case of a load application, the ignition timing Z is preferably set as the load change with a predetermined time delay relative to the setting of the combustion air ratio. This is explained in more detail below, particularly with reference to Figures 7 and 8. Fig.Figure 3 shows a representation of an embodiment of a basic characteristic map for the combustion air ratio ^(n,M) as a combustion parameter p(n,M). The basic characteristic map shown here is configured such that the combustion air ratio ^(n,M) increases globally with increasing torque M – with speed-dependent local deviations – whereby it increases or decreases with increasing speed n depending on the torque M. Of course, the basic characteristic map can also be configured differently than shown here. Figure 4 shows a representation of an embodiment of a difference characteristic map for the combustion air ratio ^(n,M) as a combustion parameter. In particular, the combustion air difference value ^^(n,M) configured here – as a difference combustion value – is equal to zero, regardless of the torque M, if there is a positive speed difference ^n (^n = nactual - nsetpoint) between an actual speed nactual and the setpoint speed nsetpoint.For negative speed differences ^n to the target speed nsoll, the difference map includes negative values ​​for the combustion air difference value ^^(n,M). The magnitude of the negative combustion air difference value ^^(n,M) decreases over wide torque ranges with increasing torque M, but has a maximum in the medium torque range, so that it also decreases towards small torque ranges, especially at idle (M = 0). Furthermore, the magnitude of the negative combustion air difference value ^^ increases, at least substantially, with the increasing magnitude of the negative speed difference ^n.The differential map is spanned here over the speed n – just as in Figure 6 for the ignition timing Z – with the target speed nsoll indicated. In both cases, however, it would also be possible, completely analogously, to span the respective differential map over the speed difference ^n and also read it out accordingly. In practice, this would have the advantage that no adjustment to the specific target speed nsoll would be necessary, for example, with different grid frequencies for the electrical power feed-in depending on the application. The differential map can also be data-based differently than shown here. Figure 5 shows an embodiment of a basic map for the ignition timing Z(n,M) as a combustion parameter p(n,M). The basic map shown here for the ignition timing Z(n,M) is data-based, at least locally, such that the ignition timing Z(n,M) is advanced at higher torques with increasing torque M.This dependence is less pronounced in the medium torque range, or rather the ignition timing Z(n,M) even remains constant there, although in the low torque range - particularly in the idle range - it is advanced as the torque M decreases. In particular, the ignition timing Z(n,M) is typically maximally advanced at idle and, starting from idle, is retarded as the torque M increases. According to the basic characteristic map, the ignition timing Z(n,M) is, in itself, at least almost independent of the speed n. The basic characteristic map can also be datad differently than shown here. Fig. 6 shows an embodiment of a difference characteristic map for the ignition timing Z(n,M) as a combustion parameter p(n,M).The difference map for the ignition timing Z shown here is data-driven in such a way that it includes negative ignition timing difference values ​​^Z(n,M) – as differential combustion values ​​– for both negative and positive speed differences ^n from the target speed n, the magnitude of which increases with increasing speed difference ^n, regardless of its sign. The ignition timing difference values ​​^Z(n,M) are largely independent of torque M; only with larger negative speed differences ^n, i.e., for larger load applications, does a more pronounced dependence on torque M result. The difference map can also be data-driven differently than shown here. Fig. 7 shows a schematic representation of the exemplary embodiment of the method using various time-dependent parameter curves for a load application.Figure 7a) shows the setting of the combustion air ratio ^, and Figure 7b) the setting of the ignition timing Z – each for the same load step. The first curve from the top shows that, in an idealized view, the torque M jumps instantaneously from a first, lower, steady-state torque value M1 to a second, higher, steady-state torque value M2 at a first time t1. Accordingly, the speed n, which at a time t0.1 preceding the first time t1 and still has the value of the target speed nsoll at the first time t1, collapses starting from the first time t1.The check in the first step S1 of the previously presented flowchart according to Figure 2, whether the load change ^L exceeds the first predetermined load change threshold ^LS, includes two check aspects in the exemplary embodiment considered here in Figure 7: first, whether a temporal speed gradient dn / dt exceeds a predetermined speed gradient threshold—in particular in terms of magnitude—and second, whether a temporal torque gradient dM / dt exceeds a predetermined torque gradient threshold—preferably in terms of magnitude. It is only concluded that the load change ^L exceeds the first predetermined load change threshold ^LS if both conditions are cumulatively met. At least one of these thresholds is defined such that exceeding the first predetermined load change threshold ^LS is only detected if an instantaneous load change—as opposed to a gradual load ramp—exists.At the first time t1 and a time t1.1 following the first time t1, the temporal torque gradient dM / dt exceeds the predetermined torque gradient threshold, but the temporal speed gradient dn / dt does not yet exceed the predetermined speed gradient threshold. At a later, second time t2, however, the temporal speed gradient dn / dt also exceeds the predetermined speed gradient threshold, so that both conditions are cumulatively met and the load change is detected. At this time, a load change signal S is activated; preferably, its binary value is set from 0 to 1. This switches the internal combustion engine 1 from steady-state operation to transient operation.The third curve from the bottom in the diagram of Figure 7a) shows the base combustion value for the combustion air ratio ^ read from the basic characteristic map for the combustion air ratio ^, which is also referred to as the steady-state combustion air ratio ^stat. This steady-state combustion air ratio ^stat is increased from a first, lower value ^1,s at the first, lower torque M1 to a second, higher value ^2,s at the second, higher torque M2; thus, the mixture is leaned out overall in a steady-state manner during load application. However, the actual combustion air ratio ^ is set identically to the steady-state combustion air ratio ^stat only up to the second time t2, as long as the steady-state operation of internal combustion engine 1 applies.As explained above, during steady-state operation, the power of the internal combustion engine 1 is controlled by controlling the throttle valve 17, and the combustion air ratio ^ and the ignition timing Z are set based on the basic characteristic maps provided for this purpose, in particular by an emission control and a knock control. The lowest curve in the diagram of Figure 7a) shows the actually set combustion air ratio ^, which is equal to the steady-state combustion air ratio ^stat up to the second time t2 and, during steady-state operation, results from the following equation derived from equation (2) for the combustion air ratio ^: ^(n,M) = ^stat(n,M).(3) In transient operation, the lowest curve for the actually set combustion air ratio ^ is calculated as the sum of the curve of the steady-state combustion air ratio ^stat and the second curve from the bottom, which represents the differential combustion value read from the differential map for the combustion air ratio ^, also referred to here as the differential combustion air ratio ^^, thus according to the following equation derived from equation (1) for the combustion air ratio ^: ^(n,M) = ^stat(n,M) + ^^(n,M). (4) Any deviations in the curves specifically shown in Figure 7a) are due to graphic inaccuracies and the extremely schematic representation. The second time t2 is set as the test time, and the predetermined test time interval begins, ending at a third time t3 – the prediction time.At the third time t3, the predicted extreme speed next,pr is now predicted by determining or calculating a predicted speed change ^npr starting from the third time t3 based on a characteristic curve, as a function of the instantaneous temporal speed gradient dn / dt at the third time t3, and adding it to a momentary actual speed difference ^nist from the target speed nsoll at the third time t3 – that is, as a result, to the instantaneous speed nist(t3) = nsoll + ^nist(t3): next,pr = nsoll + ^nist(t3) + ^npr. (5) Accordingly, the fourth curve from the top in Figure 7a) shows both the actual speed difference ^nist and a predicted extreme speed change ^npr,max: ^npr,max = ^nist(t3) + ^npr.(6) Starting at the third time t3, the differential combustion air ratio ^^ is no longer read from the differential map as a function of the actual instantaneous speed n, but as a function of the extreme speed next,pr, so that the combustion air ratio ^ is calculated as follows: ^(n,M) = ^stat(n,M) + ^^(next,pr,M). (7) In this exemplary embodiment, the differential map for the combustion air ratio ^ is configured such that the steady-state leaning is temporarily overcompensated and the mixture is effectively enriched—preferably also absolutely, i.e., by reducing the combustion air ratio ^ to values ​​less than 1.This leads – in addition to the complete opening of the throttle valve 17 in response to the load application, which is not specifically shown here and is usual in itself – to an early and strong enrichment of the mixture, so that the internal combustion engine 1 can accept the applied load reliably and quickly. In this exemplary embodiment, the differential combustion air ratio ^^ increases slightly from the third time t3 despite the currently constant extreme speed next,pr due to the additional dependence on the torque M. The combustion air ratio ^ increases on the one hand for this reason, but on the other hand also because the first contribution ^stat(n,M) from the stationary basic characteristic map increases as a function of the current speed n and the torque M.In other embodiments, the base characteristic map, and in particular also the difference characteristic map, can be differently datad, so that, for example, the combustion air ratio ^ starting from the third time t3 can remain constant, increase more sharply, decrease further, or even initially decrease and later increase again. At a fourth time t4 – the extreme time – it is determined that the current speed is passing through an actual extreme, here a minimum speed nmin. From this time on, the difference combustion air ratio ^^ is then no longer read from the difference characteristic map as a function of the extreme speed next,pr, but again as a function of the actual current speed n, so that the combustion air ratio is again calculated according to equation (4). The enrichment is thus gradually reduced. At a fifth time t5, it is determined that a transient condition no longer exists.In response to this, the load change signal S is deactivated; preferably, its binary value is set from 1 to 0, and thus the internal combustion engine 1 switches to steady-state operation. Subsequently, as shown in the tenth step S10 of Figure 2, the differential combustion ratio ^^ is driven to zero via a filter, and from a sixth time t6 determined by the time constant of the filter, the combustion ratio ^ is again calculated according to equation (3). Figure 7b) shows the setting of the ignition point Z for the same load step. The first four curves from the top therefore correspond to the representation in Figure 7a), so reference is made to the explanations there.However, in the case of load application, the ignition timing Z is now set relative to the setting of the combustion air ratio ^ at least temporarily – with regard to the second time t2, the third time t3 and the fourth time t4 – with the predetermined time delay ^t, where: tx.x = tx + ^t, with (8) x ∈ {2, 3, 4}. (9) The third curve from the bottom in the diagram of Figure 7b) shows the base combustion value for the ignition timing Z read from the basic map for the ignition timing Z, which is also referred to as the stationary ignition timing Zstat. This stationary ignition point Zstat is adjusted starting from a first, later value Z1,s at the first, lower torque M1 to a second, earlier value Z2,s at the second, higher torque M2. However, the actual ignition point Z is only adjusted up to a point in time t2 following the second point in time t2 by the predetermined time delay ^t.2 is set identically to the stationary ignition time Z. The lowest curve in the diagram of Figure 7b) shows the actually set ignition time Z, which is equal to the stationary ignition time Z up to the time t2.2 following the second time t2 and results in stationary operation from the following equation derived from equation (2) for the ignition time Z: Z(n,M) = Zstat(n,M). (10)In transient operation, the lowest curve for the actually set ignition timing Z is calculated as the sum of the curve of the stationary ignition timing Z and the second curve from the bottom, which represents the difference combustion value read from the difference map for the ignition timing Z, here also referred to as the difference ignition timing ^Z, thus according to the following equation derived from equation (1) for the ignition timing Z: Z(n,M) = Zstat(n,M) + ^Z(n,M).(11) Any deviations in the curves specifically shown in Figure 7b) are due to graphic inaccuracies and the extremely schematic representation. Starting at a time t3.3 following the third time t3 by the predetermined time delay ^t, the differential ignition timing ^Z is no longer read from the differential map as a function of the actual instantaneous speed n, but as a function of the extreme speed next,pr, so that the ignition timing Z is calculated as follows: Z(n,M) = Zstat(n,M) + ^Z(next,pr,M). (12) In this exemplary embodiment, the differential ignition timing ^Z increases slightly from time t3.3 due to the additional dependence on the torque M, despite the time-constant extreme speed next,pr. The ignition timing Z here even increases significantly due to the stationary advance of Zstat(n,M).In other embodiments, the base characteristic map, but in particular also the difference characteristic map, can be differently dataset so that, for example, the ignition timing Z can remain constant starting from time t3.3, be advanced even more, advanced less, retarded further, or even initially retarded and later advanced again. Starting at a time t4.4 following the fourth time t4 by the predetermined time delay ^t, the difference ignition timing ^Z is then no longer read from the difference characteristic map as a function of the extreme speed next,pr, but rather as a function of the actual instantaneous speed n, so that the ignition timing Z is again calculated according to equation (11). The difference characteristic map for the ignition timing Z is dataset such that the stationary advance is temporarily overcompensated, so that the ignition timing is effectively temporarily retarded accordingly.This advantageously prevents knocking combustion during the temporary enrichment of the mixture. The retardation is gradually reduced starting at time t4.4 following the fourth time t4. In the meantime, it is preferably checked periodically whether the load change ^L still exceeds a second predetermined load change threshold. This is the case for a finite time because, firstly, the load change in reality does not correspond to a delta distribution or Dirac pulse—as shown here in a simplified manner—but has a finite duration. Secondly, the sensor system used exhibits a finite time response. Furthermore, the torque signal and / or the torque gradient are filtered in the control device 21 so that the torque gradient dM / dt exceeds the predetermined torque gradient threshold for a finite time period, which here, purely by way of example, also ends at the fifth time t5.From this fifth time t5 onwards, the predetermined time delay ^t is therefore no longer taken into account. Furthermore, since the transient operation also ends at the fifth time t5, the differential ignition timing ^Z is reduced to zero via the filter, as shown in the tenth step S10 of Figure 2, with the ignition timing Z being calculated again according to equation (10) from the sixth time t6 onwards. The undershooting of the second predetermined load change threshold and the cessation of transient operation do not necessarily have to coincide in time and, as a rule, will not even do so; merely for the sake of simplicity, both are shown together here at the fifth time t5. Figure 8 shows a schematic representation of the exemplary embodiment of the method based on various time-dependent parameter curves for a load shedding.Figure 8a) shows the setting of the combustion air ratio ^ and Figure 8b) the setting of the ignition timing Z – each for the same load step. In this respect, the representation in Figure 8 is analogous to the representation in Figure 7, so reference is made to the explanations for this figure. In the following, therefore, only the differences resulting from the load shedding, which are not immediately obvious from the figures themselves, are explained. Since in this case the torque M – in an idealized view – drops instantaneously from the second, higher steady-state torque value M2 to the first, lower steady-state torque value M1 at the first time t1, the speed n increases starting from the first time t1 and passes through its extreme at a maximum speed nmax at the fourth time t4. Accordingly, the speed differences ^n to the target speed nsoll are also positive.For positive speed differences, the difference map for the combustion air ratio ^ is consistently set to zero, so that in this case the combustion air ratio ^ is adjusted at all times, even during transient operation, as per equation (3) – i.e., as in the steady-state case. This involves a gradual reduction starting from the second, higher value ^2,s to the first, lower value ^1,s, i.e., an enrichment of the mixture. The steady-state ignition timing Zstat is adjusted starting from the second, earlier value Z2,s at the second, higher torque M2 to the first, later value Z1,s at the first, lower torque M1.In the case of load shedding, the predetermined time delay ^t for the ignition timing Z is not taken into account – advantageous due to the only slow change in the combustion air ratio ^ – since the negative load change in this case is always smaller than the second predetermined load change threshold, which, as explained above, is selected to be zero or slightly positive. The difference characteristic map results in a – here additional – retardation for the ignition timing Z, even for positive speed differences, so that the already provided, stationary retardation starting from the second, earlier value Z2,s to the first, later value Z1,s is further amplified and the ignition timing is temporarily retarded further. In this way, the efficiency of the internal combustion engine 1 is advantageously greatly reduced and thus load shedding is accelerated. The additional retardation is gradually reduced starting from the fourth time t4.

Claims

CLAIMS 1. Method for operating an internal combustion engine (1), wherein at least one combustion parameter for at least one combustion chamber (7) of the internal combustion engine (1) is set as a function of a rotational speed value, wherein at a test time, a load change is detected which exceeds a first predetermined load change threshold, wherein at a prediction time lying after the test time by a predetermined test time interval, an extreme rotational speed predicted for the future based on the prediction time is predicted, and wherein, based on the prediction time, the at least one combustion parameter is set as a function of the predicted extreme rotational speed as the rotational speed value. 2.The method of claim 1, wherein the at least one combustion parameter is selected from a group consisting of a combustion air ratio, an ignition timing, and a combination of the combustion air ratio and the ignition timing.

3. The method of any one of the preceding claims, wherein the first predetermined load change threshold is selected to be characteristic of an instantaneous load change. 4.Method according to one of the preceding claims, wherein a combustion air ratio and an ignition timing are set as the at least one combustion parameter, wherein the ignition timing - in particular during load application - is set with a predetermined time delay relative to the setting of the combustion air ratio, wherein the predetermined time delay preferably corresponds to a gas transit time in an air path (9) of the internal combustion engine (1) from a gas mixer (11) to the at least one combustion chamber (7).

5. Method according to claim 4, wherein the predetermined time delay is only applied if and / or as long as the load change exceeds a second predetermined load change threshold.

6. Method according to one of the preceding claims, wherein, after the prediction time, the at least one combustion parameter, starting from an extreme time at which a current speed passes through an extreme, is no longer set as a function of the predicted extreme speed, but rather as a function of the current speed as the speed value – optionally plus the predetermined time delay.

7. Method according to one of the preceding claims, wherein, as the at least one combustion parameter, a combustion air ratio and an ignition timing are set as a function of the predicted extreme speed as the speed value during a load increase as the load change, and / or only the ignition timing is set as a function of the predicted extreme speed as the speed value during a load shedding as the load change. 8.Method according to one of the preceding claims, wherein the at least one combustion parameter is additionally adjusted as a function of a torque, in particular read from a characteristic map.

9. Method according to one of the preceding claims, wherein the at least one combustion parameter is adjusted by determining a base combustion value as a function of the speed value and optionally additionally as a function of the torque, in particular reading it from a base characteristic map, wherein a differential combustion value is further determined as a function of the speed value and optionally additionally as a function of the torque, in particular reading it from a differential characteristic map, wherein the base combustion value is offset against the differential combustion value to obtain an instantaneous value for the combustion parameter, to which the combustion parameter is adjusted. 10.Method according to one of the preceding claims, wherein ˗in the case of a load application, as the load change, starting from a first, lower stationary load to a second, higher stationary load, as the at least one combustion parameter, a combustion air ratio and an ignition timing. be adjusted, wherein the combustion air ratio is increased starting from a first, lower combustion air ratio at the first stationary load to a second, higher combustion air ratio at the second stationary load, but starting from the prediction time due to the setting as a function of the predicted extreme speed as the speed value is temporarily reduced relative to the first combustion air ratio, wherein the ignition timing is advanced starting from a first, later ignition timing at the first stationary load to a second, earlier ignition timing at the second stationary load, but starting from the prediction time due to the setting as a function of the predicted extreme speed as the speed value is temporarily retarded further relative to the first later ignition timing, and / or wherein ˗in the case of a load shedding as the load change, starting from a third,higher stationary load to a fourth, lower stationary load, as the at least one combustion parameter, a combustion air ratio and an ignition timing are set, wherein the combustion air ratio is reduced from a third, higher combustion air ratio at the third stationary load to a fourth, lower combustion air ratio at the fourth stationary load, wherein the ignition timing is retarded from a third, earlier ignition timing at the third stationary load to a fourth, later ignition timing at the fourth stationary load, and, starting from the prediction time, is temporarily retarded further, in particular relative to the fourth ignition timing, based on the setting as a function of the predicted extreme speed as the speed value.

11. Method according to one of the preceding claims,wherein the load change is detected by comparing at least one load parameter gradient with a predetermined gradient threshold value as the first load change threshold, wherein the at least one load parameter gradient is selected from a group consisting of a speed gradient, a torque gradient, a power gradient, and a combination of said gradients.

12. Method according to one of the preceding claims, wherein a gas engine, in particular with external mixture formation, is operated as the internal combustion engine (1), wherein the, Internal combustion engine (1) is preferably drivingly connected to an electric machine (3).

13. Control device (21) configured to carry out a method according to one of claims 1 to 12.

14. Internal combustion engine (1) with a control device (21) according to claim 13, wherein the internal combustion engine (1) is preferably designed as a gas engine, in particular with external mixture formation, wherein the internal combustion engine (1) is alternatively or additionally drivingly connected to an electric machine (3).

15. Internal combustion engine (1) according to claim 14, wherein the internal combustion engine (1) is drivingly connected to an electric machine (3) operable as a generator for driving the electric machine (3), wherein the control device (21) is operatively connected to the electric machine (3) and configured to detect a torque of the internal combustion engine (1) over time based on at least one electrical operating parameter of the electric machine (3).

Images