Method for operating a hydrogen internal combustion engine, and hydrogen internal combustion engine

Abstract

The present invention relates to a method for operating a hydrogen internal combustion engine, preferably as a lean concept, for adapting a hydrogen / air mixture which is supplied to a combustion chamber and is determined on the basis of a lambda value of an exhaust gas of the hydrogen internal combustion engine, comprising the following steps: determining an adaptation for an amount of hydrogen supplied to the combustion chamber on the basis of a first adaptation method A1, which ascertains a first adaptation value Z1, and on the basis of a second adaptation method A2, which ascertains a second adaptation value Z2, coordinating the first adaptation value Z1 and the second adaptation value Z2 in order to determine an overall adaptation, and, on the basis of the adaptation, correcting a hydrogen supply in order to maintain a specified threshold value for a lambda value of the exhaust gas, the first adaptation method A1 being based on a least square method and the second adaptation method A2 being based on an operating-point-dependent method.

Description

[0001] Description

[0002] title

[0003] Method for operating a hydrogen combustion engine and hydrogen combustion engine

[0004] State of the art

[0005] The present invention relates to a method for operating a hydrogen internal combustion engine for adapting a hydrogen-air mixture supplied to a combustion chamber. Furthermore, the present invention relates to a hydrogen internal combustion engine configured for carrying out a method according to the invention.

[0006] Hydrogen combustion engines are known in various configurations from the prior art. One operating concept for hydrogen combustion engines is the so-called lean-burn concept, in which there is more combustion air than hydrogen to be burned. This results in an excess of air during combustion, so that, as long as a certain lambda threshold is not undershot, low NOx emissions can be achieved. However, due to component tolerances, aging effects, and dynamic operating conditions, deviations in the hydrogen-air mixture can occur. A control unit is intended to ensure that a critical lambda threshold is not undershot. Lambda control in hydrogen combustion engines operating in lean-burn mode, however, presents a significant challenge.Firstly, the control of the fuel path is not moment-neutral, and secondly, the measurement of the lambda value using sensors is always subject to a larger measurement tolerance, which can lead to inaccurate control.

[0007] Disclosure of the Invention: The inventive method for operating a hydrogen internal combustion engine for adapting a hydrogen-air mixture supplied to a combustion chamber, with the features of claim 1, has the advantage that improved control is possible in the event of deviations from a predetermined lambda value of the hydrogen internal combustion engine. The hydrogen internal combustion engine is operated in lean mode, and the control is performed based on a lambda value of an exhaust gas from the hydrogen internal combustion engine. In particular, the inventive method adjusts a hydrogen feedforward control system to maintain the desired lambda value.

[0008] This is achieved according to the invention by the method first determining an adjustment for the amount of hydrogen supplied to the combustion chamber based on a first and a second adaptation method. The first adaptation method determines a first adaptation value Z1, and the second adaptation method determines a second adaptation value Z2. Subsequently, the first and second adaptation values ​​are coordinated to determine the adjustment. Based on this adjustment, the hydrogen supply is then corrected to achieve the specified lambda value and / or to avoid exceeding the critical lambda threshold. The first adaptation method is based on a least-squares method, and the second adaptation method is based on an operating-point-dependent method (map adaptation).

[0009] Thus, an adaptation of a hydrogen-air mixture can be achieved based on two different adaptation principles.

[0010] The dependent claims describe preferred embodiments of the invention.

[0011] Preferably, the coordination is performed using a handshake method. The handshake method is designed to ensure interference-free determination of the hydrogen-air mixture adaptation, since, according to the invention, two different adaptation methods are used. This prevents faulty control of the hydrogen-air mixture. Preferably, either the first or the second adaptation value Z1, Z2 is used as the adaptation value. That is, preferably only an intermediate value is used to determine the adaptation, particularly for short-term control. Alternatively, however, both adaptation values ​​can also be used.

[0012] One of the adaptation values ​​Z1 or Z2 is only preferred for adjustment if the adaptation value Z1 is stable. Stability is defined as the need to learn a specific number of newly adapted measurement points within a predetermined cycle, and the resulting newly calculated deviation of the hydrogen-air mixture value must remain within defined limits.

[0013] Preferably, the handshake procedure is carried out using priorities to determine which of the two adaptation procedures may be continued to perform the next mixture adaptation.

[0014] Furthermore, it is preferable that the quantity of hydrogen supplied to the combustion chamber in the first and second adaptation procedures is determined from historical, and in particular already adapted, values. This allows for adjustments to the hydrogen-air mixture, if necessary, even during the adaptation procedures.

[0015] Preferably, the least-squares method uses a straight line equation where the slope of the line is used as a multiplicative adaptation value and the intersection of the line with the y-axis is used as an additive adaptation value. Thus, the least-squares method determines the mixture deviation based on a straight line equation. For this purpose, measurement points are determined within defined areas, and a straight line is derived from them. A measurement point is preferably defined by the calculated air mass and the product of the relative calculated fuel mass and the lambda sensor value. This results in a ratio between the calculated value and the actual measured lambda value, which describes the deviation between an actual and a target value.

[0016] Preferably, the adaptation values ​​are used continuously for hydrogen feedforward control. More preferably, at very high lambda values, especially lambda values ​​> 2, the adaptation is weakened or terminated. In the operating-point-dependent method, a PI control, particularly of the supplied hydrogen quantity, is also preferably implemented, whereby an adaptation value at a respective map point defines an additive adaptation value. Thus, the operating-point-dependent adaptation is a map-based adaptation. A modulated lambda setpoint is preferably set using a PI controller. The modulated lambda setpoint is preferably calculated such that map-based learning of the mixture error with active as well as passive mixture correction can be performed. A control variable of the controller is preferably adjusted using a defined learning algorithm.An adaptation value at the respective map point describes an additive correction value for the hydrogen feedforward control. This is preferably used in a defined lambda range with upper and lower limits.

[0017] Preferably, based on the determined adaptation value, the injection length of hydrogen and / or the injection pressure of the hydrogen is changed in order to achieve a predetermined lambda value or at least not to violate the critical lambda threshold.

[0018] Preferably, the first and second adaptation methods are only carried out within a predetermined lambda range between a lower and an upper limit. This is particularly advantageous when a lambda value is very large, especially > 2, since mixture adaptation can then preferably be omitted because NOx emissions are minimal or do not occur.

[0019] Furthermore, the present invention relates to a hydrogen combustion engine, equipped for carrying out the method according to the invention.

[0020] Brief description of the drawings

[0021] The present invention is described in detail below with reference to the accompanying drawing. The drawing shows:

[0022] Figure 1 is a schematic representation of a process of the method according to the invention and Figure 2 is a diagram to illustrate the Least Squares method.

[0023] Preferred embodiments of the invention

[0024] The following describes, with reference to Figures 1 and 2, a method for operating a hydrogen internal combustion engine in a lean-burn concept and such an internal combustion engine in detail.

[0025] The lean-burn concept offers the significant advantage of low NOx emissions, as long as a certain threshold of the lambda value of the exhaust gas of the hydrogen combustion engine is not undercut.

[0026] According to the present invention, the supply of hydrogen can be optimized, enabling the adaptation of a hydrogen-air mixture supplied to a combustion chamber. A control variable is preferably a lambda value of the exhaust gas from the hydrogen combustion engine.

[0027] The method according to the invention is particularly effective when the lambda value of the exhaust gas is below a predetermined threshold, especially close to the lambda threshold. Then, according to the invention, an adjustment is made for a quantity of hydrogen supplied to the combustion chamber based on a first adaptation method A1 and a second adaptation method A2.

[0028] The first adaptation method A1 determines a first adaptation value Z1 and the second adaptation method A2 determines a second adaptation value Z2.

[0029] In the next step, the first and second adaptation values ​​Z1 and Z2 are coordinated to determine the adjustment. Based on this adjustment, the hydrogen supply is corrected to bring the lambda value back above the specified threshold or to prevent the lambda threshold from being exceeded.

[0030] The first adaptation method, A1, is based on a least-squares method, while the second adaptation method, A2, differs from A1 and is based on an operating-point-dependent method (map adaptation). In the first adaptation method, A1, which is based on the least-squares method, an average mixture deviation of the hydrogen combustion engine is adapted. Here, an additive adaptation value and a multiplicative adaptation value are determined. This adaptation continuously corrects the feedforward control of the hydrogen supply. However, due to increasing sensor tolerance with increasing lambda value, it is necessary to restrict a lambda range for adaptation. To a certain extent, adaptation values ​​can also be extrapolated for lambda ranges that were not used for the adaptation itself.

[0031] As can be seen in Figure 2, the least-squares method determines a mixture deviation based on a linear equation. For this purpose, a large number of measurement points are collected within a defined lambda range, and a straight line G is derived from them (see Figure 2). The measurement points are schematically plotted in the diagram of Figure 2. The diagram of Figure 2 thus shows a linear correlation between current measurements on the y-axis and predicted measurements on the x-axis.

[0032] The slope of line G is the multiplicative adaptation value, and the y-intercept of line G is the additive adaptation value. A measurement point is described by the calculated air mass and the product of the relative calculated hydrogen mass and the lambda sensor value. This yields a ratio between the calculated value and the actual measured value, thus describing the deviation of the actual value from the target value. The adaptation values ​​are continuously used to control the hydrogen supply to the combustion chamber. At very high lambda values, it is also possible that the adaptation is weakened or completely suspended, since at very high lambda values ​​there is practically no NOx in the exhaust gas.

[0033] Figure 1, on the left, schematically illustrates the determination of the first adaptation value Z1 by the first adaptation procedure A1 based on the least-squares method. In a first step S1, it is checked whether the lambda value has fallen below a predefined threshold or, more specifically, whether it is close to, i.e., within ±10% of, the lambda threshold. If yes (J), the procedure proceeds to step S2; if no (N), a loop returns to a processing unit 3.

[0034] If yes (J), a large number of measurement points are collected in step S2.

[0035] In step S3, it is checked whether the number of collected measurement points has reached a predetermined number. If yes (J), the procedure continues to step S4; if no (N), the procedure returns to step S1 and collects further measurement points if necessary, provided the lambda value remains below or close to the predetermined lambda threshold.

[0036] In step S4, the line G is calculated, as schematically shown in Figure 2. The result is an initial adaptation value Z1 to the computer unit 3.

[0037] According to the inventive method, the second adaptation method A2 (shown on the right side of the diagram in Figure 1) is also carried out, wherein the second adaptation method is based on the operating point-dependent method.

[0038] The first adaptation procedure A1 and the second adaptation procedure A2 can be carried out simultaneously, or the first adaptation procedure can be carried out before the second adaptation procedure, or there can be an overlap where the first adaptation procedure is started and the second adaptation procedure is started before the first adaptation procedure is finished.

[0039] The second adaptation method, A2, uses a map to compensate for operating-point-dependent deviations in the hydrogen-air mixture. Due to the lambda sensor's tolerance during lean-burn operation of the hydrogen combustion engine, it is necessary to restrict the lambda range for the adaptation process. Since extrapolating a map cannot be reliably performed, the adaptation values ​​are only applied in ranges that have already been successfully adapted. Therefore, the second adaptation method is preferably used to correct and / or limit the advance control of the hydrogen supply near the lambda threshold. The advantage lies in the fact that good adaptation is possible precisely in this emissions-relevant operating range of the hydrogen combustion engine, and thus, in particular, a good correction of the exhaust gas lambda value is also possible.

[0040] In the second adaptation method, A2, a lambda value is preferably adjusted to a modulated lambda setpoint using a PI controller. The lambda setpoint is determined in such a way that a mixture error can be learned using a map with both active and passive mixture correction. A control variable for the controller is adapted, for example, in a fuel-speed map using a defined learning algorithm. The adaptation value at the respective map point then describes an additive correction value for the feedforward control of the hydrogen supply, which can then be used in the defined lambda range.

[0041] According to the invention, the first and second adaptation methods are combined. It is necessary that the adaptation be coordinated in such a way as to avoid, firstly, double adaptation of the hydrogen-air mixture and, secondly, to ensure reliable compliance with the NOx emission limit at all times and in every operating point of the hydrogen combustion engine. This is achieved by means of the computer unit 3, which is configured to perform a handshake procedure.

[0042] The handshake procedure is preferably only started once the first adaptation method A1, based on the least-squares method, has been stably adapted. Stable adaptation is defined as the learning of a specific number of newly adapted measurement points within a driving cycle, and the resulting newly calculated mixture error remaining within defined limits. Only then does the computer unit 3 release the second adaptation method A2, based on the operating-point-dependent method.

[0043] At the same time, it is always preferably checked whether an intervention through adjustment is necessary to comply with the NOx emission limit. If the NOx emission limit is met, the process is preferably terminated immediately.

[0044] As already explained, the handshake procedure is carried out in such a way that the first adaptation method is stable before the second adaptation method is executed. Computing unit 3 can determine, via defined priority criteria, whether and which of the first and second adaptation methods may continue adapting. Furthermore, the adaptation speed of the two adaptation methods is preferably adjusted to prevent cross-contamination.

[0045] It should also be noted that, in parallel with the adaptation procedures, previously stored adapted values ​​(historical values) are used for mixture correction, especially if the historical values ​​have previously been assessed as stable and reliable by the computer unit 3.

[0046] This allows for the adjustment of the hydrogen-air mixture in hydrogen combustion engines, ensuring compliance with prescribed NOx emission limits, particularly during lean-burn operation. In particular, this also compensates for the measurement tolerance of a lambda sensor, which exhibits increasingly larger measurement tolerances with rising lambda values, potentially leading to inaccurate mixture control.

Claims

Claims 1. Method for operating a hydrogen internal combustion engine, preferably as a lean-burn concept, for adapting a hydrogen-air mixture supplied to a combustion chamber, which is determined based on a lambda value of an exhaust gas of the hydrogen internal combustion engine, comprising the steps: Determining an adaptation for a quantity of hydrogen supplied to the combustion chamber based on a first adaptation method A1, which determines a first adaptation value Z1, and a second adaptation method A2, which determines a second adaptation value Z2; coordinating the first adaptation value Z1 and the second adaptation value Z2 to determine an overall adaptation; and, based on the adaptation, correcting a hydrogen supply to comply with a predetermined threshold for an exhaust gas lambda value and / or to correct the lambda value, wherein the first adaptation method A1 is based on a least-squares method and wherein the second adaptation method A2 is based on an operating point-dependent method.

2. The method of claim 1, wherein the coordination is carried out by means of a handshake method.

3. Method according to one of the preceding claims, wherein the first adaptation value Z1 or the second adaptation value Z2 or a combination of the first and second adaptation values ​​Z1 , Z2 is used as the adaptation.

4. Method according to claim 3, wherein one of the adaptation values ​​Z1 , Z2 is used for the adaptation only if the intermediate value is stable.

5. Method according to claim 3 or 4, wherein the adjustment value used for adjustment is further adjusted depending on a lambda value of the exhaust gas.

6. Method according to one of the preceding claims, wherein, in carrying out the first adaptation method A1 and the second adaptation method A2, the quantity of hydrogen supplied to the combustion chamber is determined from historical values, in particular already adapted values.

7. Method according to one of the preceding claims, wherein the least-squares method determines a straight line equation of a straight line G, where a slope is used as a multiplicative adaptation value and an intersection point of the straight line G with a y-axis is used as an additive adaptation value.

8. Method according to one of the preceding claims, wherein the operating point-dependent method performs PI control, wherein an adaptation value at a respective characteristic map point defines an additive adaptation value.

9. Method according to one of the preceding claims, wherein, based on the specified adjustment, the injection length of the hydrogen and / or the injection pressure of the hydrogen is changed in order to reach and / or exceed a predetermined lambda threshold.

10. Method according to one of the preceding claims, wherein the first adaptation method A1 and the second adaptation method A2 are carried out only in a predetermined lambda range between a lower limit and an upper limit.

11. Hydrogen internal combustion engine configured to carry out a method according to any one of claims 1 to 10.

Images