METHOD FOR MONITORING ADAPTATION VALUES IN AN ENGINE CONTROL UNIT

Description

[0001] The present invention claims priority from French application No. 2105495 ​​filed on 05 / 27 / 2021.

[0002] The invention relates to a method, implemented in a thermal engine engine control, for monitoring several adaptives. This thermal engine is advantageously but not limited to a spark-ignition engine, in particular a gasoline or gasoline-mixture engine, and the monitoring is carried out advantageously but not limited to monitoring the impact of the adaptives on the richness of fuel injected into the thermal engine.

[0003] This non-limiting application will be taken as an example to illustrate the monitoring method but the method according to the invention can be implemented to monitor adaptives of an operating parameter other than the fuel richness in the heat engine.

[0004] Due to manufacturing variations, wear, fouling and the quality of representation of the engine actuator models, the physical behavior of these actuators may differ from the behavior models integrated into the engine control. This mismatch in the actuator models, particularly those relating to the air intake branch and the fuel injection branch, can lead to richness drifts, thus to overconsumption or an increase in pollutant emissions and can also have an impact on the driving pleasure experienced by the driver. The richness of a fuel mixture indicates the proportion value between air and fuel in the mixture admitted into the engine combustion chamber. The quality of combustion mainly depends on this dosage.

[0005] The engine control system will therefore have to correct these variations in richness throughout the vehicle's life. This correction is carried out by a known richness regulation function which constantly corrects the injector control time based on the richness measurement provided by the richness sensor present in the exhaust.

[0006] To optimize this correction, it is known to use a learning-type engine control strategy which will either memorize (via adaptives) the richness correction necessary for each zone or operating point of the engine and restore it when the engine returns to a given learned zone, or identify the sources of richness deviation and correct these sources (the actuator models for example) by adaptives. In either case, the richness is thus automatically well centered. Such an engine control strategy by learning, according to the first variant mentioned above, is for example described in patent document FR 3 085 721 B1.

[0007] The regulations relating to on-board diagnostics (OBD) require that these richness learning systems be monitored, in particular to alert the driver in the event of a fault occurring which could lead to the tolerable thresholds of pollutant emissions being exceeded. This monitoring is generally carried out by diagnosing the values ​​taken by the richness adaptors in comparison with limit values.

[0008] It is known to perform the diagnosis of the richness adaptors independently for each adaptor, by comparing the values ​​taken by the adaptor at the time of its update with respect to a limit value.

[0009] This limit value is generally set in order to monitor an excess of what the adaptive is supposed to correct: manufacturing dispersions, wear, fouling, or even the quality of representation of the engine actuator models. If this limit is exceeded, the system considers that there is an abnormality and that a failure may have occurred, causing an exceptional richness deviation that has been learned in part by the adaptive in question. In this case, there is then a potential risk of exceeding the tolerable thresholds of pollutant emissions.

[0010] The diagnosis of the adaptives by such a method is thus carried out independently for each adaptive based on a limit value that the adaptive considered is not supposed to reach. However, in the case where the engine control strategy provides several adaptives of different nature to correct the behavior models of the actuators (for example, injector model, cylinder air filling model, camshaft position model, etc.), this method has the following drawbacks: the diagnosis is carried out, for each adaptive, on the value reached by the adaptive from limits based essentially on experience and the classic values ​​that the adaptive should take (typically correction of manufacturing dispersion, wear, or even fouling). However, it is the risk of exceeding the level of polluting emissions (or at least its equivalence in richness) which must be monitored precisely according to the regulatory texts; the diagnosis is carried out, for each adaptive, on a value reached by the adaptive but without considering its real impact in richness (and therefore in polluting emissions). However, in the case where this adaptive is a constant (independent of the engine operating point therefore), its impact in richness can vary depending on the engine operating point.It is possible that an unusual value of the adaptive, beyond the limits, only has an impact on the richness in very precise zones of the engine operating field that the driver of the diagnosed vehicle does not travel; the interaction between the different adaptives is not taken into account by the method: an adaptive can thus reach limits considered as having an impact on the richness but, due to the values ​​of the other adaptives applied to other behavior models of the engine control actuators, the effects on the overall richness (and therefore on the polluting emissions) of these adaptives can be significantly reduced.

[0011] Document DE 10 2008 012607 A1 discloses a method for controlling the injection of an internal combustion engine. Lambda control is used to correct the air-fuel ratio. To do this, a deviation from the predefined target value of the air-fuel ratio is determined. The determined deviation is determined and learned as an adaptation value depending on at least one operating parameter and the adaptation value is analyzed with regard to the cause and assigned to the air path and / or the fuel path.

[0012] Document DE 102 44 539 A1 discloses a global adaptive correction of injected quantities and / or air mass measurement errors in internal combustion engines, which involves a polynomial-based recursive learning technique using information about current engine operation.

[0013] The aim of the invention is to overcome the drawbacks of the prior art by proposing a method, implemented in an engine control, for monitoring several adaptives of at least one parameter, which is more exhaustive and more precise, and which makes it possible to take into account the possible interactions between adaptives while reducing the number of false detections.

[0014] To do this, the invention thus relates, in its broadest sense, to a method, implemented in an engine control of a vehicle, for monitoring several adaptives of at least one parameter, the method comprising the following steps: a calculation, for each adaptive, of an individual impact value of said adaptive on said at least one parameter for a current operating point of the engine; a calculation of an overall impact value of all the adaptives on said at least one parameter for the current operating point of the engine; a comparison of the calculated overall impact value with at least one predefined overall impact threshold value; and an emission, depending on the result of the comparison, of an alert signal to a device of the vehicle.

[0015] The method according to the invention is therefore an engine control function, which assists a known learning function by monitoring it and alerting the driver in the event of a failure occurring which risks causing the tolerable pollutant emission thresholds to be exceeded. The present invention can of course be adapted to engine control learning functions other than a richness correction learning function. The method according to the invention makes it possible to monitor the overall impact of all the adaptives on the parameter concerned, in order to refocus the parameter in question. The monitoring permitted by the method according to the invention is more exhaustive and more precise, because it takes into account the possible interactions between the adaptives and because the estimation made of the impact of all the adaptives on the parameter varies according to the current operating point of the engine.The method according to the invention in fact combines the impact of all the adaptives (which may be of a different nature), thus taking into account the influence that the adaptives have on each other; and detects an excessive correction of the parameter by the adaptives (such an excessive correction being directly correlated to pollutant over-emissions when the parameter is the fuel richness) and not an excessive value of a given adaptive which is not easily convertible into effects on the pollutant emissions. This makes it possible to reduce the number of false detections. The monitoring permitted by the method according to the invention also contributes to better compliance with the regulations in force, in particular the regulations relating to OBD on-board diagnostics in terms of monitoring pollutant emission levels when the parameter is the fuel richness.The process in fact detects an excessive correction of the parameter by the combination of adaptives, which is easily transposed into a probable impact on polluting emissions.

[0016] The predefined overall impact threshold value(s) is (are) preferably configurable by a user or by a vehicle manufacturer. This makes it possible to calibrate this (these) threshold value(s) as close as possible to what the regulatory texts require, particularly in terms of pollutant emissions when the parameter is fuel richness.

[0017] Preferably, said at least one parameter is a fuel richness.

[0018] According to a particular technical characteristic of the invention, the step of calculating, for each adaptive, an individual impact value of said adaptive on said at least one parameter consists of multiplying a current value of said adaptive by a predetermined transfer function between said at least one parameter and said adaptive, thus providing the individual impact value. Such a transfer function represents the sensitivity of the adaptive to the parameter. This step of calculating an individual impact value of each adaptive makes it possible to obtain the impact of each adaptive on the parameter for the current operating point of the engine.

[0019] According to another particular technical characteristic of the invention, the step of calculating an overall impact value of all the adaptives on said at least one parameter consists of adding the individual impact values ​​calculated for all the adaptives, thus providing the overall impact value. This step of calculating an overall impact value makes it possible to take into account the influence that the adaptives have on each other. For example, two adaptives of different nature can, depending on the operating point of the engine, compensate each other in terms of richness correction if they are of opposite signs.

[0020] According to another particular technical characteristic of the invention, the step of comparing the calculated overall impact value to at least one predefined overall impact threshold value comprises a first phase consisting of comparing the calculated overall impact value to a minimum overall impact threshold value, and a second phase consisting of comparing the calculated overall impact value to a maximum overall impact threshold value, and the alert signal is emitted to a device of the vehicle if the calculated overall impact value is lower than the minimum overall impact threshold value or higher than the maximum overall impact threshold value.

[0021] Advantageously, the steps of calculating individual impact values, calculating a global impact value, comparing and issuing an alert signal are re-performed for each current operating point of the engine. This makes it possible to make the diagnosis of the adaptors dependent on the current operating point of the engine. The accuracy in detecting the operating point of the engine during which a failure occurs is therefore greatly improved, since the diagnosis depends on the current operating point and is no longer limited to monitoring a fixed value varying only during updates of the parameter learning function within the engine control, as is the case in certain methods of the prior art.

[0022] Embodiments of the present invention will be described below, by way of non-limiting examples, with reference to the single appended figure [ Fig.1 ] which is a flowchart representing a method of monitoring several adaptives of at least one parameter according to the present invention.

[0023] Referring to the figure 1The present invention relates to a method, implemented in an engine control of a thermal engine, for monitoring several adaptives of at least one parameter. Such monitoring of the adaptives makes it possible to carry out a diagnosis of these same adaptives, in particular a diagnosis relating to compliance with the regulations in force for the parameter. The parameter may be a richness of injected fuel but this is not limiting in the context of the present invention. In this case, each adaptive is a fuel richness adaptive. The different adaptives are preferably applied directly to the sources of errors in the fuel richness, that is to say to the models of the different elements of the thermal engine. These adaptives are applied for example: on the position models of the camshaft phase shifters; and / or on the model for estimating the quantity of air sucked in by the cylinders; and / or on the behavior model of at least one injector by correcting the modeling of physical parameters such as the static gain of the injector, its control dead time.

[0024] For example, without this being limiting, a first adaptive may be an adaptive on the position of an intake camshaft phase shifter, a second adaptive may be an adaptive on the position of an exhaust camshaft phase shifter, a third adaptive may be an adaptive on the modeling of the static gain of a fuel injector in the heat engine, and a fourth adaptive may be an adaptive on the modeling of the injector control dead time. Alternatively or additionally, one of the adaptives may also be an adaptive relating to an opening duration of at least one fuel injector in the heat engine.

[0025] The method comprises a first step 10 during which the engine control calculates, for each of the adaptives, an individual impact value of the adaptive on the parameter for a current operating point of the engine. Preferably, this calculation step 10 consists of multiplying, for each adaptive, a current value of the adaptive by a predetermined transfer function between the parameter and the adaptive. This multiplication then provides the individual impact value of the adaptive concerned, on the current operating point of the engine. This predetermined transfer function (and stored for example in memory means of the engine control) represents the sensitivity of the adaptive to the parameter.The transfer function can be determined in advance by any known method, for example by mathematical calculations of derivatives of equations of the system parameter with respect to the adaptive considered, or by calculation of the local gradient of variation of the parameter for a variation of the adaptive.

[0026] For example, for the first adaptive of the example previously described (adaptive on the position of an intake camshaft phase shifter), if we call A the first adaptive, R the parameter which is here the fuel richness, and IA→R the individual impact of the first adaptive on the fuel richness; then this individual impact IA→R is expressed according to the following equation (1): I A → R = K A , R ∗ A with KA,R the transfer function between the fuel richness and the first adaptive, which is expressed in % / °CK; the notation °CK designating crankshaft degrees.

[0027] The individual impact on fuel richness of each of the second, third and fourth adaptives of the previously described example is expressed according to an equation analogous to that of equation (1), with a corresponding individual transfer function.

[0028] During a following step 12, the engine control calculates an overall impact value of all the adaptives on the parameter for the current operating point of the engine. Preferably, this calculation step 12 consists of adding the individual impact values ​​calculated for all the adaptives during the previous step 10. This addition then provides the overall impact value of the adaptives, on the current operating point of the engine.

[0029] During a following step 14, the engine control compares the overall impact value calculated during the previous step 12 to at least one predefined overall impact threshold value. The comparison step 14 comprises, for example, a first phase consisting of comparing the overall impact value calculated to a minimum overall impact threshold value, and a second phase consisting of comparing the overall impact value calculated to a maximum overall impact threshold value. The first phase may be carried out before the second phase, or vice versa. Alternatively, the first and second phases are carried out simultaneously. In the case where the parameter is the fuel richness, the minimum and maximum overall impact threshold values ​​are regulatory minimum and maximum thresholds on the richness deviation, which correspond to the pollutant emission limits authorized by the regulations.

[0030] During a following step 16, the engine control emits, depending on the result of the comparison carried out during the previous step 14, an alert signal to a device of the vehicle. The device may in particular be a display device such as a screen for example, making it possible to visually inform the user of the vehicle upon receipt of the alert signal. The device may alternatively be a sound reproduction device, making it possible to emit an auditory signal intended for the user, upon receipt of this alert signal. The device may more generally be any device making it possible to alert a user of the vehicle.When the previous comparison step 14 comprises the two aforementioned phases, the alert signal is transmitted to the device if the overall impact value calculated during step 12 is lower than the minimum overall impact threshold value or higher than the maximum overall impact threshold value.

[0031] Steps 10, 12, 14 and 16 described above are repeated for each current operating point of the engine.

[0032] The method according to the invention allows more exhaustive and more precise monitoring of adaptives, and makes it possible to take into account possible interactions between adaptives while reducing the number of false detections.

Claims

1. Method, implemented in an engine control of a vehicle, for monitoring several adaptives of at least one parameter , characterized in that the method comprises the following steps: - a calculation (10), for each adaptive, of an individual impact value of said adaptive on said at least one parameter for a current operating point of the engine; - a calculation (12) of an overall impact value of all the adaptives on said at least one parameter for the current operating point of the engine; - a comparison (14) of the calculated overall impact value to at least one predefined overall impact threshold value; and - an emission (16), depending on the result of the comparison (14), of an alert signal to a device of the vehicle.

2. Method according to claim 1, characterized in that said at least one parameter is a fuel richness.

3. Method according to claim 1 or 2, characterized in that the step (10) of calculating, for each adaptive, an individual impact value of said adaptive on said at least one parameter consists in multiplying a current value of said adaptive by a predetermined transfer function between said at least one parameter and said adaptive, thus providing the individual impact value.

4. Method according to any one of claims 1 to 3, characterized in that the step (12) of calculating an overall impact value of all the adaptives on said at least one parameter consists in adding the individual impact values calculated for all the adaptives, thus providing the overall impact value.

5. Method according to any one of claims 1 to 4, characterized in that the step (14) of comparing the calculated overall impact value to at least one predefined overall impact threshold value comprises a first phase consisting of comparing the calculated overall impact value to a minimum overall impact threshold value, and a second phase consisting of comparing the calculated overall impact value to a maximum overall impact threshold value, and in that the alert signal is emitted to a device of the vehicle if the calculated overall impact value is lower than the minimum overall impact threshold value or higher than the maximum overall impact threshold value.

6. Method according to any one of claims 1 to 5, characterized in that the steps of calculating (10) individual impact values, calculating (12) a global impact value, comparing (14) and emitting (16) a warning signal are repeated for each current operating point of the engine.