FR2880921A1 - METHOD FOR MANAGING AN INTERNAL COMBUSTION ENGINE AND DEVICE FOR IMPLEMENTING THE METHOD - Google Patents

Abstract

A method of managing an internal combustion engine (10) whose exhaust zone (13) is equipped with at least one catalyst (15) and a lambda probe (16) downstream or in a portion thereof catalyst (15), diagnosing the catalyst based on a measurement of the oxygen storage capacity of or in the catalyst (15), wherein it is assumed that an oxygen store of the catalyst (15) is at least substantially empty or filled and the reference value of the lambda coefficient of the internal combustion engine (10) is changed to a value greater or less than 1. The lambda probe (16) is broadband, a first variation is detected (60) of the lambda signal (lam-nK, lam-nKF) provided by the wide-band lambda probe (16), and the accumulated / evacuated oxygen after the first variation (50) is determined and ending this determination when a second variation (62) of the lambda signal (lam-nK, lam-nKF) or if a predefined quantity of oxygen has been accumulated / evacuated.

Description

Field of the invention

  The present invention relates to a method for managing an internal combustion engine, the exhaust gas zone of which is equipped with at least one catalyst and a downstream probe A or in part of the catalyst, a process that carries out a diagnosis. catalyst based on a measurement of the oxygen storage capacity of the catalyst or in the catalyst, wherein it is assumed that an oxygen store of the catalyst is at least substantially empty or filled and the setpoint of the coefficient A of the internal combustion engine at an air ratio of greater than or less than 1.

  The invention also relates to a device for implementing such a method.

  STATE OF THE ART A catalyst diagnostic capability utilizes the determination of the oxygen storage capacity of the catalyst. Reduction of oxygen storage capacity is exploited as a measure of the decrease in catalyst conversion capacity. The oxygen storage capacity of a catalyst can be determined by subjecting the exhaust air coefficient upstream of the catalyst to a cyclic variation and passing this coefficient around the value of 1.0. and comparing the signal measured downstream of the catalyst with the values of the pre-defined upstream coefficient. For a high oxygen storage capacity, the oxygen accumulated in the catalyst for values of predefined coefficients inférieures less than 1 makes it possible to oxidize in a very large measure the reducing components contained in the exhaust gases. ; in the case of values of the coefficient A, predefined, greater than 1, it is possible to accumulate excess oxygen to a very large extent. The measurable variations of the coefficient ΔC downstream of the catalyst are relatively low if the catalyst has a large capacity to accumulate oxygen, that is to say, whether the catalyst is working properly.

  Document DE 41 12 478 A1 describes for example a written translation of the method mentioned above. This document provides for the installation of an abrupt variation probe (or abrupt variation sensor) upstream of the catalyst and an abrupt variation probe A downstream of the catalyst. It is first of all examined whether for a regulation oscillation of the coefficient A upstream of the catalyst, passing from the rich mixture to the lean mixture or vice versa, the value of the coefficient A downstream of the catalyst has a corresponding transition. If this is the case, it is assumed that the oxygen accumulator of the catalyst is either completely filled or completely empty, which corresponds to a defined initial state. Then, the mass flow rate of the gases passing through the catalyst is determined. Then, the integral as a function of time of the product of the mass flow rate of gas and a term containing the coefficient A upstream of the catalyst as well as the integral as a function of time and the product of the mass flow rate of the gases and a term having a coefficient A downstream of the catalyst. As a measure of the aging state of the catalyst, either the difference between the two integrals, the quotient of the two integrals or the quotient of the difference and one of the two integrals is used.

  DE 102 57 059 A1 discloses a method and a device corresponding to the type defined above which provides a diagnosis of the catalysts installed in several exhaust gas pipes (number N) of the internal combustion engine. Downstream, after the meeting of the N exhaust gas pipes, there is a probe A whose signal is compared with the values of the coefficient Δ obtained upstream of the catalysts. The diagnosis is based on the exploitation of the oxygen storage capacity in the catalysts. As probe A, for example, broadband probes are used.

  OBJECTS OF THE INVENTION The present invention aims to develop a method of managing an internal combustion engine and a device for implementing this method for reliable diagnosis of the catalyst or catalysts.

  To this end, the invention relates to a method of the type defined above, characterized in that the probe A is a broadband probe A, a first variation of the signal A is detected by the broadband probe A, and the accumulated / evacuated oxygen is determined after the first variation and the determination of the accumulated / evacuated oxygen is completed when a second change in the signal λ is recorded or if a predefined quantity of oxygen has been accumulated / discharged.

  The invention also relates to a device for carrying out this method, this device being characterized in that the control device comprises a diagnostic control which modifies the setpoint value X, a modification determination means for processing the A signal and an integrator to determine a measure of accumulated / evacuated oxygen.

  According to the method of the invention, in the zone of the exhaust gas of the internal combustion engine, there is at least one catalyst and downstream of the catalyst or part of the catalyst there is a sensor or a probe. The catalyst is diagnosed by taking at least a measure of the oxygen storage capacity by the catalyst or in the catalyst. It is assumed that the oxygen store of the catalyst is at least substantially empty or substantially charged. Then the reference value of the coefficient of the internal combustion engine is modified so as to set this reference value to an air coefficient λ greater than 1 or less than 1. According to the invention, a first variation is firstly noted. A signal that provides a probe to installed downstream of the catalyst or downstream of a portion of the catalyst. This probe is a wide band probe. A measurement of the amount of oxygen accumulated or evacuated after the first variation of the signal λ is determined. The determination of the quantity of oxygen accumulated or evacuated ends when a second variation of the signal A is noted or if the measurement of the quantity of oxygen exceeds a threshold.

  The use of a wideband probe allows the process of the invention to have a relatively high accuracy in determining at least one measure of the oxygen storage capacity of the catalyst which can serve as a measure of the aging state of the catalyst. A high oxygen storage capacity corresponds to a catalyst in good condition.

  The process according to the invention can use the amount of oxygen accumulated in the catalyst or the amount of oxygen removed from the catalyst. The accumulation of a reducing agent, such as, for example, hydrogen and / or carbon monoxide, resulting from the operation of the internal combustion engine with an air coefficient Δ less than 1 is synonymous with evacuation. oxygen. Since the accumulation / evacuation of oxygen in / from the catalyst is an equilibrium reaction, in a transient zone there will be no accumulation / evacuation of all the available oxygen. Due to the kinetics of accumulation / evacuation and diffusion operations, the evolution of the oxygen concentration, downstream of the catalyst, is smoothed over time. We arrive at a transient zone or beach.

  An excess of oxygen / lack of oxygen occurs downstream of the catalyst before the maximum oxygen storage capacity is exhausted / or the entire amount of oxygen is accumulated.

  The determination of at least one measurement of the oxygen accumulated / evacuated and which begins after having noted a first variation of the signal λ, will be completed when a second variation of the signal λ is noted or if the measurement of the oxygen accumulated / evacuated exceeds a threshold. The diagnosis of the catalyst can thus terminate prematurely without waiting for the second variation if one already has an oxygen storage capacity sufficiently good for the catalyst. If, as a second criterion, the second variation of the signal A is used to complete the determination of the measurement of the quantity of oxygen accumulated / evacuated, it will be possible to obtain a relatively accurate measurement of the residual oxygen storage capacity. . By comparison with a predefined capacity threshold, it can be decided to replace, if necessary, the catalyst.

  The method according to the invention for managing an internal combustion engine makes it possible to precisely perform the catalyst diagnosis by appropriately influencing the fuel / air ratio of the air / fuel mixture supplied to the internal combustion engine to predefine the engine. initial condition and the consequent variation of the air coefficient.

  The method according to the invention makes it possible to perform the catalyst diagnosis also in the context of the normal operation of the internal combustion engine insofar as, in the context of this normal operation, it is in the initial condition, it is that of a catalyst oxygen accumulator which is completely empty or completely filled and then, under normal operation, the air coefficient λ is changed to a value greater than 1 or less than 1.

  According to a development, the variation of the signal λ is determined from the gradient of the signal λ. The gradient of the signal of which we have an approximate value preferably in the form of a difference quotient can for example be determined continuously at predefined time intervals. The detection of the variation may, for example, be based on the fact that the gradient must exceed a predefined gradient threshold. Preferably, a minimum duration is also provided during which the threshold of the gradient must be exceeded. An additional detection or variation detection alternative uses the maximum of the gradient. In addition, or alternatively, the variation can be determined by the gradient which first has a maximum then decreases and falls below a gradient threshold.

  A development provides filtering of the signal A by a low-pass filter to eliminate the disturbing signals and the rapid variations of the signal A caused by dynamic processes in the determination of the variation.

  A development is to determine the oxygen content from the interval as a function of time depending on the coefficient de of the combustion and the air signal supplied by an air gripping means which detects the combustion air supplying the engine. internal combustion. Taking into account the combustion air and integrating, the mass of oxygen accumulated or evacuated is determined. Another development provides for intermediate recording of the air signal and, where appropriate, the coefficient de of the combustion for a delay time which corresponds at least substantially to the transit time of the gases in the catalyst until reaching the probe. To broadband. By this means, the non-stationary operating state of the internal combustion engine during the diagnosis of the catalyst has only a slight effect on the result of the diagnosis.

  A development provides conditioning of the catalyst before its diagnosis; for this, the combustion coefficient pour for the internal combustion engine is fixed in a predetermined manner to a value greater than 1 / less than 1 so that the oxygen storage tank of the catalyst is at least approximately completely filled or completely emptied. The predefinement of the coefficient de of the combustion to a value greater than 1 can be suppressed if the diagnosis of the catalyst is made after a phase of thrust failure of the internal combustion engine which lasts at least as long as the oxygen accumulator is at least substantially filled.

  A development provides the exploitation of the accumulated / evacuated oxygen as a function of the catalyst temperature and / or the mass flow rate of the exhaust gases. This means makes it possible to set the threshold for the comparison of the measurement of the catalyst accumulation capacity according to the operating conditions of the catalyst.

  The device for implementing the method of the invention comprises a control apparatus.

  The control apparatus includes a diagnostic control which modifies the set value of the coefficient λ, a modification determination means for processing the signal A supplied by the broadband probe A and an integrator for determining the quantity measurement. of accumulated / evacuated oxygen. The control apparatus preferably comprises at least one electrical memory in which the process steps are recorded in the form of a computer program.

  Drawings The present invention will be described hereinafter in more detail with the aid of embodiments shown in the accompanying drawings in which: - Figure 1 shows the technical environment in which the process according to the invention takes place; Figures 2a-2b show concentration versus location curves, and Figure 3 shows a curve or A signal as a function of time downstream of a catalyst.

  Description of the embodiments

  FIG. 1 shows an internal combustion engine 10 whose intake zone 11 comprises an air gripping means 12 and whose exhaust gas zone 13 comprises an exhaust gas temperature sensor 30, a catalyst 15 and downstream, after the catalyst 15, an A 16 probe. The catalyst 15 has a cat-inlet catalyst inlet and a cat-exit catalyst outlet.

  The air gripping means 12 provides an air signal msL to a control apparatus 20; the internal combustion engine 10 furnishes a speed or rotation speed signal n, the exhaust gas temperature sensor 14 provides an exhaust gas temperature signal Tabg and the probe A 16 provides a signal X , lam-nK. The controller transmits a fuel signal mK to a fuel metering device 21. In the exhaust zone 13, there is a mass flow rate of exhaust gas msabg.

  The air signal msL is further transmitted to a thrust cut-off control 30, a delay 31 and a threshold fixing means 32. The thrust cut-off control means 30 provides a thrust cut-off signal 34 The delay means 31 provides a timed air signal msL-d to an integrator 35. The threshold setting means 32 provides a comparator 36 with a first and a second value. threshold 37, 38. The rotation speed n is supplied to the thrust cut-off control 30 and the threshold fastening means 32.

  The diagnostic control 33 provides the threshold setting means 32 with a first diagnostic control signal 39 and a change determination means 40 with a second diagnostic signal 41; it supplies a diagnostic X-ray signal D1 to a coefficient pre-setting means A 42. The diagnostic X-signal lam-D is furthermore applied to the delay means 31 which emits a delayed X signal lam-d destined for the integrator 35. The diagnostic control 33 receives a diagnostic stop signal 41 provided by a comparator 36.

  The presetting means of the coefficient λ 42 receives not only the diagnostic X signal lam-D but also a signal X of nominal operation lam-N. The coefficient pre-setting means λ 42 provides a set value X, λ-S to a regulator λ 50 which provides the fuel signal mK. The signal X, lam-nK supplied by the sensor A 16 is applied not only to the coefficient regulator 50 50 but also to a low-pass filter 51. The coefficient regulator 50 50 further receives a signal X, λl-vK which represents the air coefficient Δ in the exhaust gas upstream of the catalyst 15.

  The filtered lam-nKF signal X supplied by the low-pass filter 51 arrives at the change determination means 40. The integrator 35 supplies the result 52 of the integration to the comparator 36 which emits a fault signal F. FIGS. -2e show the curves of the signals according to the location x. These curves represent the% O 2 oxygen concentration as well as the% Rea reagent concentration.

  Figure 2a shows the situation at a first location x1 after a jump between rich and lean mode, after which there is a high concentration of oxygen% O 2 and a low concentration of reagent% Rea.

  Figure 2b shows the situation at a later time for concentration variations at a second location 02.

  FIG. 2c shows the case where the concentrations of% O 2 and% Rea downstream of the cat-entry catalyst inlet have changed: there is a first transient range x 10 in which the oxygen concentration% O 2 changes only slightly.

  Figure 2d shows the situation at a later time when the% O 2 and% Rea concentrations in the catalyst have changed. There is a second transient zone x 20 which has a greater extension with respect to the first transient zone x10.

  Figure 2e shows the case in which the concentrations% 02,% Rea, have largely changed after the exit of cat-exit catalyst. There is a third transient zone x 30 which again has a greater extension with respect to the second transient zone x 20.

  Figure 3 shows the signal X, lam-nK as a function of time t. Until a first moment t1, the coefficient λ must be equal to 0.97. Between the first and the second instant t1, t2, there is a first variation 60 of the signal X, lam-nK. At a third instant t3 the signal X, lam-nK has at least substantially a constant level or plateau 61 for which the coefficient λ is at least substantially equal to 1. At a fourth instant t4, begins a second variation 62 for the signal X , lam-nK which ends at a fifth instant t5. After the fifth time t5, the coefficient λ is 1.03.

  The method according to the invention operates as follows: The sensor at 16 is in the form of a wideband sensor which can capture a wide range of air coefficient λ. This range is for example between 0.7 and 4.0. The sensor A 16 is provided downstream of at least one catalyst 15. For large catalysts, the sensor or probe A 16 can also be installed downstream. a partial volume of the catalyst. The signal X, lam-nK supplied by the probe A 16 can not only be used for diagnosis but also for the regulation of the combustion coefficient λ. The signal X, lam-nK is, for this reason, supplied not only to the low-pass filter 51 but also to the regulator λ 50 which can influence the fuel signal mK as a function of the signal X, lam-nK. If necessary, it is possible to take into account a signal X, lam-vK measured by a non-represented probe A installed upstream of the catalyst 15. This latter signal is supplied to the coefficient regulator λ 50.

  The coefficient regulator 50 50 serves to control the combustion coefficient qui which corresponds to the set point value X, lam-S predefined by the coefficient presetting means λ 42. The set value X, lam- S corresponds to the normal operation of the internal combustion engine 10 at nominal mode X, lam-N. Instead of the coefficient regulator A 50, it is also possible to provide a command. In addition, the signal X, lam-vK which represents the air coefficient Δ in the exhaust gas upstream of the catalyst 16 is not necessary. Instead, the signal X, lam-nK captured downstream of the catalyst 15 can be used.

  The diagnostic control 33 can prescribe the diagnostic X signal lamD for the diagnosis of the catalyst that the means of predefining the coefficient At 42 will provide as setpoint X, lam-S. The diagnosis is prepared by practically filling or emptying the oxygen accumulator of the catalyst 15 completely. In the case of the conditions represented by the curves of FIGS. 2a-2e, one starts from a sudden transition from the rich mixture to the mixture. poor at the beginning of the diagnosis when the oxygen concentration% 02 goes from a low level to a high level. The variation of the coefficient A of the internal combustion engine from a rich mode value to a lean mode, correspondingly changes the reagent concentration% Rea which goes from a high level to a low level. In order to carry out the diagnosis, the catalyst must first receive a low concentration of oxygen% O 2 until the oxygen accumulator is at least substantially completely empty.

  Alternatively, one can start from an oxygen accumulator substantially completely filled with the catalyst 15 and in this case the catalyst 15 is first solicited by a high oxygen concentration% 02. This mode of operation occurs when the engine internal combustion has been running long enough in 2c mode.

  According to an advantageous development, it is provided that the thrust cut-off control 30 which determines the throttle failure of the internal combustion engine 10, for example from the air signal msL and the speed n, signals to the diagnostic control 33 by the thrust cutoff signal 34 that the diagnosis can be started. The thrust cutoff signal 34 is provided when the thrust breaking phase has lasted long enough.

  The change from the reference value X, lam-S to the diagnostic X signal lam-D occurs according to FIG. 2a upstream of the catalyst 15 in the form of a sudden change in the oxygen concentration% O 2 and as a sudden change in the reagent concentration% Rea, we have a first point xl. At a later time, the concentration changes to a second point x 2 for which the concentration transitions are still relatively abrupt. The abrupt concentration ranges at the origin are already slightly attenuated by the diffusion and turbulence in the exhaust zone 13.

  The lean exhaust gases with a high concentration of oxygen% 02, thus exhaust the rich exhaust gases containing a higher concentration of% Rea reagents, which have only a very low concentration of oxygen% 02. In the case represented by the graph of Figure 2c, the con-centration variations correspond to the downstream location of the cattened catalyst inlet. After the high oxygen concentration% O 2 has reached the catalyst 15, the excess oxygen stores in the catalyst 15. This corresponds to the first transient zone x 10 in which the oxygen concentration% O 2 represents the first variation 60. Then, there is a bearing 61 at least relatively short followed by the second variation 62.

  The concentration of reagent% Rea drops during the first variation 60 to lower values. Since the catalyst can not store all the available oxygen, a relatively high concentration of oxygen remains in the first transient zone x 10, such as in the rich exhaust gas.

  The graph in Figure 2d shows the situation for concentration changes at a later time. The second transient zone x 20 is longer than the first transient zone x 10.

  Fig. 2e shows the situation at an even later time when the concentration changes occur partly already downstream of the cat-exit catalyst outlet or after a path portion in the catalyst 15. Only this moment that one can measure the variations of concentration with the probe À 16.

  First, the signal X, lam-nK of the probe at 16 will indicate a rich coefficient λ for the exhaust gas set, for example, at 0.97. The evolution over time is shown in Figure 3 in which up to time t1, there must be a lack of oxygen. The variation of the oxygen concentration% 02 must start again at the first moment t 1 by the first variation 60. The exploitation of the signal X, lam-nK is done in the variation determination means 40 after the diagnostic command 33 has transmitted the second diagnostic control signal 41 by means of variation determination 40.

  It is preferably provided a low-pass filter 51 which eliminates the signal X, lam-nK, on the one hand the high frequency interference signals and, on the other hand, the rapid variations related to dynamic operations that do not concern the first variation 60 so as to avoid measurement errors. The variation determining means 40 can then determine the gradient of the signal X, lam-nK or that of the filtered X signal lam-nKF.

  To recognize the first variation 60, the variation determining means 40 can determine, for example, the gradient of the signal X, lam-nK or the filtered X signal lam-nKF.

  The gradient can be determined continuously with rapid succession. For example, an approximate value can be obtained by forming difference quotients. According to a first embodiment, the gradient is compared with a first predetermined gradient threshold. If this gradient threshold is exceeded, the integrator release signal 53 is emitted. In another embodiment, the gradient must exceed the gradient threshold for a predefined period of time before the integrator release signal 53 not be issued. According to another embodiment or a complementary embodiment, it is possible to determine the point of inversion of the signal X, lam-nK or the filtered X signal lam-nKF and to use it to supply the integrator release signal 53. or alternatively, one can first determine the gradient maximum and then check whether the gradient exceeds a threshold before providing the integrator release signal 53.

  As soon as the first variation 60 is detected, it is possible to begin the diagnosis of the catalyst. The oxygen stored in the catalyst 15 is determined. It can be the oxygen mass or the amount of oxygen. The determination is made for example by the integrator 35 which multiplies the value (1-1 / λ) to the air signal msL and with a constant corresponding to the percentage of the oxygen contained in the air and then to integrate according to the time. For a relative evaluation, the constant can be set to 1. In order to take into account the transit time of the gases through the catalyst 15, the air signal msL supplied by the air gripping means can be tempered. 12 with delay means 31 and supply the integrator 35 with the timed air signal msL-d. In addition, it is possible to delay the coefficient λ used for the integration by means of the delaying means 31. The combustion coefficient λ during the diagnosis corresponds to the diagnostic X signal lam-D transmitted as delayed X signal lam-d to the integrator 35. The delay of the coefficient λ can be suppressed because during the diagnosis, the diagnostic X signal lam-D is normally kept constant.

  The delay time to be preset depends, preferably, on the msL air signal. In addition, the delay time may be dependent on the load of the internal combustion engine 10. The load is given for example by the fuel signal mK, possibly in combination with the speed n, or by the torque of the engine internal combustion experienced by the control apparatus 20.

  Oxygen according to FIG. 3 is determined in the region of the plateau 61 of the signal X, lam-nK or the filtered X signal lam-nKF. The bearing 61 is more or less accentuated. During step 61, the coefficient λ can vary from a value that is just below 1 to a value that is just over 1. Experimentally, it has been found that the values of the coefficient λ vary for a jump between the rich and the the lean diet, in a range between 0.99-1.01 and in the case of a leap between the lean diet and the rich diet, in a range between 0.998 and 1.002.

  According to a first embodiment of the method of the invention, the integration is completed when the second variation 62 which occurs at the fourth instant t4 is found. The second variation 62 can be determined by proceeding analogously to the observation of the first variation 60. When the second variation 62 occurs, the integration release signal 53 is neutralized, which ends the integration. The integration result 52 which represents a measure of oxygen storage / oxygen removal or storage of reagents is compared with the first threshold 37 provided by a threshold setting means 32. If the threshold is exceeded Translate a bad catalyst, the comparator 36 provides the fault signal F registered for example in a fault memory or which will be displayed.

  According to a second embodiment of the method of the invention, the integration can be completed before reaching the second variation 62. In this case, the result of the integration 52 is compared with the second threshold 38 provided by the threshold fixation 32; this second threshold has been set at a value corresponding to a good catalyst 15. Insofar as the integration result 52 already corresponds to a good catalyst 15, it is possible to complete the catalyst diagnosis before having noted the second variation 62 .

  Since the first or second threshold 37, 38 has been reached or exceeded, the comparator 36 provides the diagnostic stop signal 43 so that the diagnostic control 33 stops the diagnosis. For this purpose, the first and the second control and diagnostic signals 39, 41 are neutralized. The variation determination means 40 picks up the release signal from the integrator 53 and thus restores the integrator 35 to its position of rest to be then available for a new determination of the oxygen concentration.

  The threshold determining means 32 may set the first and / or second threshold 37, 38 as a function of the air signal msL, the speed n and / or the temperature of the catalyst. In a simple embodiment, the exhaust gas temperature signal Tabg, upstream of the catalyst 15, can be used for the measurement of the catalyst temperature. The msL air signal is a measurement of the mass exhaust gas flow rate signal msabg which influences the oxygen storage capacity of the catalyst and the catalyst / temperature temperature of the exhaust gases Tabg.

  The exhaust gas temperature signal Tabg can be measured by the temperature sensor 14. According to an advantageous development, it is possible to calculate the temperature or the exhaust gas temperature signal Tabg with the air signal msl. and, for example, the fuel signal mK as a measure of the load of the internal combustion engine giving an at least approximate signal.

Claims (15)

  1) A method of managing an internal combustion engine (10) whose exhaust gas zone (13) is equipped with at least one catalyst (15) and a probe A (16) downstream of the catalyst (15) or in a portion of the catalyst (15), a method of diagnosing the catalyst based on a measurement of the oxygen storage capacity of the catalyst (15) or in the catalyst (15), which assumes that an oxygen storage tank of the catalyst (15) is at least substantially empty or filled and the reference value of the coefficient of the internal combustion engine (10) is changed to an air ratio Δ greater than 1 or less than 1 , characterized in that the probe A (16) is a wide-band probe A, a first variation (60) of the signal A (lam-nK, signal a, filtered lam-nKF) is detected provided by the probe A at band (16), and the oxygen accumulated / evacuated after the first variation (50) is determined, and the determination of the ox is completed. ygene accumulated / evacuated when a second variation (62) of the signal A (lam-nK, lam-nKF) or a predefined quantity of oxygen has been accumulated / evacuated.   2) A method of managing an internal combustion engine according to claim 1, characterized in that the variation (60, 62) of the signal λ (lam-nK) is determined from the gradient of the signal λ (lam-nK). , lam-nKF).   3) A method of managing an internal combustion engine according to claim 2, characterized in that one detects the variation (60, 62) by exceeding a gradient threshold predefined by the gradient for a given duration.   4) A method of managing an internal combustion engine according to claim 2, characterized in that the variation (60, 62) is detected if the gradient passes through a maximum.   5) A method of managing an internal combustion engine according to claim 2, characterized in that the variation (60, 62) is detected if the gradient has exceeded a maximum and then it passes under a gradient threshold.   6) A method of managing an internal combustion engine according to claim 1, characterized in that the signal A (lam-nK) supplied by the probe A (16) is filtered by a low-pass filter.   7) A method of managing an internal combustion engine according to claim 1, characterized in that the oxygen is determined from the integral as a function of the time dependent of a signal X diagnostic lam-D predefined during the diagnosis and a signal lm provided by an air gripping means (12) which captures the combustion air supplying the internal combustion engine (10).   8) A method of managing an internal combustion engine according to claim 7, characterized in that the msL air signal is recorded intermediate, for a delay time corresponding to the gas travel time, until reaching the probe A (16) broadband.   9) A method of managing an internal combustion engine according to claim 8, characterized in that the delay time depends on the msL air signal and / or a load (mK, n) of the internal combustion engine ( 10).   10) A method of managing an internal combustion engine according to claim 1, characterized in that the oxygen storage tank is filled with or completely empty of the catalyst (15) or part of the catalyst (15) before diagnosis of the catalyst, by predefining a combustion coefficient supérieur greater than 1 / less than 1.   11) A method of managing an internal combustion engine according to claim 10, characterized in that the combustion coefficient greater than 1 / less than 1 is predefined in the normal operation of the internal combustion engine ( 10).   12) A method of managing an internal combustion engine according to claim 1, characterized in that the catalyst is diagnosed after a thrust breaking phase of the internal combustion engine (10) by which the accumulator of Oxygen catalyst (15) has been at least substantially filled.   13) A method of managing an internal combustion engine according to claim 1, characterized in that the accumulated / exhausted oxygen is used as a function of the catalyst temperature / exhaust gas temperature (Tabg) and / or exhaust gas mass flow rate (Msabg).   14) Device for managing an internal combustion engine, characterized in that it comprises a control apparatus (20) for carrying out the method according to one of claims 1 to 13.   15) Device for managing an internal combustion engine according to claim 14, characterized in that the control device (20) comprises a diagnostic control (33) which modifies the setpoint X, (lam-S). , change determination means (40) for processing the signal λ (lam-nK, lam-nKF) and an integrator (35) for determining an accumulated / evacuated oxygen measurement. io