FR2859763A1 - Internal combustion engine management process for use in motor vehicle, involves correcting fuel injection duration for injection in cylinder, before end of injection operation and during operation, based on fuel pressure in container - Google Patents

Abstract

The process involves determining fuel pressure in a pressure container before triggering fuel injection operation. Fuel injection duration is calculated for a cylinder to trigger the injection operation. The calculated injection duration is corrected for the fuel injection in the cylinder, before an end of the injection operation and during the operation, based on the determined fuel pressure. An independent claim is also included for an injection device for an internal combustion engine having N cylinders.

Description

Field of the invention

  The present invention relates to a method of managing an internal combustion engine, in particular a motor equipping a motor vehicle comprising N cylinders in which the fuel is injected directly, the method comprising the following steps: capturing the fuel pressure in a fuel accumulator (ramp) before the injection operation to be triggered, preliminary calculation of the injection time ti for a cylinder (n) with n = 1 ... N for the injection operation to be triggered based on the detected boom pressure and any other quantities.

  The invention also relates to the implementation of this method an injection device for an internal combustion engine with N cylinders, comprising: - a fuel accumulator (ramp) for accumulating fuel 15 and supply by injectors to combustion chambers of the internal combustion engine, - a high pressure pump for pumping fuel under high pressure into the fuel accumulator, - a pressure sensor for capturing the actual fuel pressure in the fuel accumulator and - a control apparatus for regulating the actual pressure sensed by the pressure sensor in the fuel accumulator to a preset ramp pressure value by the actuator control of the high pressure pump and for the pre-calculation of ble of an injection time for a cylinder n (n = 1 ... N) of the internal combustion engine.

  STATE OF THE ART According to DE 196 45 715 A1 a control device of a direct injection internal combustion engine is known.

  The control device described in this document calculates a basic injection time for a cylinder of the engine. This basic injection time is calculated to result in a mass or dose of fuel that it is desired to inject for a presumed fixed pressure ratio in the fuel accumulator (ramp) and the combustion chamber of the cylinder. But in reality the reports or pressure conditions in particular in the combustion chamber of the cylinder are not stationary.

  This is why DE 196 45 715 A1 proposes to correct the basic injection time, calculated beforehand, taking into account the pressure curve calculated beforehand in the combustion chamber of the cylinder for the calculated position. the crankshaft angle by correcting the injection and setting a clock to emit the injection time based on the previously corrected injection time. This correction calculation, proposed for the duration of injection, however assumes that the pressure of the ramp is stationary, that is to say constant.

  In reality, not only the pressure in the combustion chamber but also that in the ramp is exposed to variations in time. This is why any calculation based on fixed pressure ratios is always tainted by error. The pressure curve in the ramp can only be predicted in a very complicated and imprecise way because it is dynamically modified by many influence quantities.

  Knowing this situation, the document DE 197 26 756 C2 proposes another method and another management system of an internal combustion engine. This method proposes for this purpose to enter the pressure of the ramp during injection operation and from the ramp pressure thus captured it calculates continuously and from the start of the injection, the amount of fuel which has already been injected.

  The injection will then be interrupted when the fuel mass injected during the current injection operation begins to exceed a preset setpoint.

  The teaching described in the document cited last has the disadvantage that the calculation of the fuel mass injected (or more precisely ejected) is very complicated and the delay time of the injectors during the current injection must be calculated separately for their opening and closing. In addition, the calculation must be done in a very high time frame, that is to say much more frequently than in a frame of lms because otherwise, for the small injection times (for example 0.5 ms), there would be considerable errors in the injected fuel dose or the process could be totally faulty. In addition, this method does not take into account the variation of the pressure in the combustion chamber.

  OBJECT OF THE INVENTION Based on this state of the art, the object of the present invention is to develop a method and a device for managing a direct injection internal combustion engine so as to produce an accurate injection of a dose of Setpoint fuel regardless of the variations of the boom pressure taking into account the variable pressure of the combustion chamber.

  The invention also proposes to develop a computer program for the execution of this method or the management of the mo-tor.

  DESCRIPTION AND ADVANTAGES OF THE INVENTION To this end, the invention relates to a method of the type defined above, characterized in that relying on at least one ramp pressure value detected just before or / and during injection operation and on the other possible variables, the value calculated beforehand of the injection duration is corrected before the end thereof at least once in a corrected injection time ti_K for the injection in progress in the cylinder n.

  This method thus makes it possible to grasp the pressure of the ramp and to correct the calculation thus made of the injection time for the cylinder (n) before the start of the injection operation in this cylinder, and then to correct this calculated value. beforehand for the injection duration before it is completed by relying on at least one ramp pressure value entered just before or / and during the injection operation and, if appropriate, the other quantities, for obtain at least once a corrected injection time for the cylinder (n).

  Advantageously, in the method according to the invention, the seizure of the pressure of the fuel in the ramp is not only just before but additionally during the duration of the injection duration calculated beforehand, always or as often as possible following a narrow time frame. From these values renewed in pei Inanence for the pressure of the fuel in the ramp can be grasped in a very precise way variations of the pressure of the ramp in time and the resulting variations for the mass or dose of bus - injected fuel; this is taken into account for the calculation of the correction of the current injection time and finally there will be a very precise determination of the quantity of fuel globally injected into a cylinder.

  The method according to the invention allows a precise metering of the fuel whatever the modulations or gradients of the pressure of the ramp. After starting, the pressure of the boom can be reduced as quickly as possible according to the maximum flow rate of the high-pressure pump, which allows a better preparation of the mixture and an improvement in the emission of the pollutants. In the case of sudden load variations in the positive or negative direction, during the operation of the internal combustion engine combined with the necessary rapid increases or decreases in the pressure of the boom, the deviation of the value of the lambda coefficient is minimized. - s haited or adjusted, because according to the invention is corrected the injection time to always inject the target fuel dose, as calculated. Pump flow tolerances or other flow influences are eliminated by entering the boom pressure during the current injection.

  It is advantageous that the first correction of the pre-calculated injection time not be done only during the current injection operation but already just before its start to guarantee corrections even for inferior injection times to the weft time of the ramp pressure input.

  The entry of the ramp pressure before and during the injection operation is preferably done according to a fixed time frame, for example every millisecond. Directly after such an entry, however, it is recommended to make a correction calculation of the duration of the injection.

  The pre-calculation of the injection duration is preferably done at a triggering time which is based not on the time frame indicated but more on a fixed crankshaft angle frame. This trigger mark may precede the start of the injection differently depending on the position of the crankshaft angle.

  The pre-calculation of the injection duration is preferably based on the ramp pressure value entered according to the time ramp, the last moment of tripping which precedes the synchronism angle and with the prior calculation of a pressure value in the combustion chamber for the predictable angular position of the crankshaft for injection.

  Preferably, the ramp pressure signal used for the calculation and / or the correction of the injection duration is a ramp pressure signal filtered by a low-pass filter.

  The calculation of the correction of the injection time can be specified by taking into account not only the ramp pressure but also the pressure in the combustion chamber of the cylinder in which the injection is made precisely at the same time. moment of the entry of the ramp pressure in the calculation of the correction of the injection duration.

For example:

  the corrected value ti_K of the injection time is calculated from the ramp pressure value minus a pressure value as another quantity representing the pressure in the combustion chamber for an internal combustion engine operating in a combustion chamber. homogeneous mode it is assumed that the pressure in the combustion chamber is constant and corresponds to the atmospheric pressure, for types of injections for which it is injected during the compression phase, the pressure in the combustion chamber is measured according to the the same time frame as for the ramp pressure, either calibrated or evaluated from the position of the crankshaft and the compression ratio, is measured using a pressure sensor of the combustion chamber.

  Advantageously according to the invention there is a simple method or a simple mathematical algorithm allowing a simple and fast calculation of the corrected injection duration. The algorithm represents a possible additional calculation for a normal calculation of an injection timing.

  By way of examples, the corrected injection duration ti_K is calculated by applying the following formula: k tiK = tiRejAp_RBR Ek / EIAp_RB (j) l = 1 or k ti_K = tiR Ep_RBR Ek / E-OpRB (j ) + (p_R (j) - p_R (j -1)) / 2 in which p_R (j) represents the value of the pressure of the ramp at the instant of detection j, ti_K corrected injection duration, ti_R duration of injection calculated beforehand at a moment of tripping in angular synchronism, Ap_RB_R pressure difference between the pressure in the ramp and the pressure in the combustion chamber at the moment of release, k current number of detections during an operation of injection, instant synchronous detection time being located directly before and / or during the injection in progress, 4p_RB (j) pressure difference between the pressure in the ramp and the pressure in the combustion chamber at the instant of detection j.

  At high rotational speeds and in case of travel time problem of the internal combustion engine can be removed or simplify this calculation. If the last injection time calculated and corrected in the most current manner is completed, the injection operation is stopped. This is advantageously done by updating a transmission clock in the time frame of the calculation with the injection time each time corrected in the current manner.

  In order to calculate the injection duration, the effective pressure curve in the ramp, which can be subjected to high-frequency oscillations, is preferably not used, but a value of the damped ramp pressure, in particular filtered by a pass filter. -low.

  The problem mentioned above is furthermore dealt with by an injection device, characterized in that the control device corrects the value calculated beforehand of the injection duration before the end of the latter by relying on a minus a ramp pressure value entered just before and / or during the injection phase and any other quantities, at least once in a corrected injection time ti_K for the injection in progress in the cylinder n.

  Advantageously, a filter fitted to the input circuit of the control unit uses the signals of the pressure sensor, in particular a low-pass filter in the form of an element (RC) for damping the curve of the ramp pressure.

  In a variant, the fuel accumulator comprises a hydraulic low-pass filter in the form of a throttling element and an intermediate fuel accumulator for damping the high-frequency variations of the ramp pressure, and the pressure is connected to the intermediate fuel accumulator to detect a damped ramp pressure curve.

  Drawings The present invention will be described below in more detail with the aid of embodiments shown in the accompanying drawings in which: - Figure 1 shows the basic structure of an injection device according to the invention, FIG. 2 shows a first timing diagram explaining the operation of the injection device; FIG. 3 shows a second diagram explaining the operation of the injection device during the starting phase of the internal combustion engine, FIG. shows a third chronogram explaining the process according to the invention.

  Description of embodiments

  Figure 1 shows the structure of an injection device 100 according to the invention applied to a four-cylinder internal combustion engine (these cylinders are not shown). The device comprises a fuel accumulator 110 also called rail or common rail for accumulating fuel and supply it to the combustion chambers of the cylinders of the internal combustion engine via four injectors 120. The injection device 100 also comprises a high pressure pump 130 for pumping fuel under high pressure in the ramp 110. The high pressure pump 130 is supplied with fuel by an electric fuel pump 105. The fuel supplied to the high pressure pump 130 by the fuel pump Electrical fuel 105 is first supplied to a damping member 134. From there, the fuel arrives through a first check valve 136 in the transfer member 138 when the flow control valve 132 is connected. , that is, blocked. This transfer element 138 comprises a piston 139 controlled by the camshaft 160 of the internal combustion engine. In the situation shown in Figure 1, the piston 139 is in its lowest position to form in the transfer member 138 a chamber 142 for receiving or accumulating the fuel arriving through the check valve 138. If later the piston 139 of Figure 2 rises due to the corresponding rotation of the camshaft 160, the volume of the chamber 142 decreases and the fuel accumulated in the chamber is compressed. The compressed fuel is then supplied via a second check valve 137 to the ramp 110 in which the fuel is stored under high pressure.

  The storage or accumulation pressure of the fuel in the ramp, that is to say the pressure of the ramp, is regulated on a predetermined set pressure by the injection device 100. For this, the device of injection 100 comprises a control apparatus 150. This control apparatus receives the current pressure of the ramp provided by the ramp pressure sensors 410 or 140 ', i.e. the actual value. Depending on the regulation deviation thus noted between the preset setpoint pressure and the actual detected pressure, the control device acts on the flow control valve 132 of the high pressure pump 130 and thus performs the regulation of the pressure of the ramps to the desired set pressure. More precisely, this regulation consists in cutting the flow control valve 132, that is to say the connecting pipe 133 shown in FIG. 1, connecting the transfer element 138 and the damper 134, if the pressure the ramp is too important and it must be reduced. In this case, the connecting line 133, open, allows the fuel to leave the chamber 142 in the damping element 134 and thus achieve the desired pressure decrease. In the reverse case, if it is necessary to increase the pressure of the ramp, the flow control valve 132 is correspondingly connected, that is to say that the connecting pipe 133 is closed.

  The opening of the injectors 120 can trigger high frequency pressure pulsations in the ramp 110 which combine with the slow changes in pressure. When these pressure pulsations are detected by the pressure sensors 140, 140 'and are provided as components of the actual pressure sensed at the control unit 150, this can distort the measurement value input and be translated thus to an undesired regulation of the pressure of the ramp. The time constant of these pulsations, however, is significantly less than that of the increase in pressure caused by the transfer stroke of the piston 139 in the transfer member 138 of the high-pressure pump 130 or the pressure decrease. because of the mass of fuel injected, so that appropriate filters can eliminate such pulsations. A first possibility is to use a low pass filter 142 in the input circuit of the ramp pressure sensor 140 in the control apparatus 150; it may be for example an RC element having an appropriate time constant which eliminates by low-pass filtering, the high frequency components.

  The frequency of the troublesome pressure pulsations at high frequency is defined by the resonance frequency of the ramp and thus, inter alia, by its geometry and its volume. This volume is generally designed large enough to achieve pressure accumulation effects, resulting in a lower resonance frequency. The time constant of the troublesome pressure pulsations is thus increased and is brought closer to the time constant of the pressure variations generated by the transfer strokes of the high-pressure pump 130 and is offset with respect to the time constants due to injections. Alternatively or in addition to the first possibility already mentioned can be used as a second possibility low-pass filtering to eliminate high frequency pulsations of pressure, inconvenient, by performing a hydraulic low-pass filter. This second possibility is shown in broken lines in FIG. 1. This possibility makes it possible to reduce the actual volume of the ramp to as low a level as possible and to connect this volume by a throttling point 144 to a complementary volume 146 or to communicate with it. In this case, the ramp pressure sensor 140 'is not connected directly to the ramp 110 but to the complementary volume 146. On this complementary volume, it will be possible to advantageously detect a heavily damped ramp pressure curve as the actual size of the control apparatus 150. The described arrangement can be considered as a hydraulic low-pass filter.

  The control method or management of the internal combustion engine using the injection device 100 described above with reference to Figure 1 will be explained below with reference to Figures 2, 3 and 4.

  Figure 2 shows different operations that occur in parallel with time. References ti_1 ... 4 on the abscissa represent the chronology of the injection into an internal combustion engine, not shown, which is supposed to be four cylinders. The timing of the injection is shown with reference to the trigger marks (tr marks) indicated above and which are also associated with each of the cylinders 1 ... 4; from these markers incremental signals and reference marks are formed by a crankshaft angle sensor. The timing of the injection is shown with reference to the BM markings of the crankshaft with an angular position of for example 60 crankshaft angle (60 KW) before the top dead center OT of the cylinders 1 and 3. The landmarks tripping give a calculation frame for each cylinder used for calculation according to the invention of the duration of injection as will be explained later.

  FIG. 2 also shows, above the BM marks, the curve of the stroke (stroke-HDP) of the high-pressure pump 130 according to FIG. 1 when the piston 139 of the transfer element 138 is driven by the crankshaft 160 of the internal combustion engine. Above the stroke (stroke-HDP) is shown the MSV actuation signal output from the controller 150 to control the flow control valve 132; a high level corresponds to the connection of the flow control valve and thus at the same time this represents the closing of the connecting pipe 133. Conversely, a low level of this signal corresponds to the cut-off of the flow control valve and thus at the opening of the connecting pipe 133. The opening and closing of the flow control valve 132 are made on command of the control device responding to the detected ramp pressure p_R so as to regulate this pressure on a preset setpoint p_R_cons.

  Above the MSV actuation signal of the flow control valve 132, Fig. 2 shows the ramp pressure p_R for the internal combustion engine running at a high speed and for a large fuel debit. From the curve of the ramp pressure thus represented in combination with the actuation signal MSV of the flow control valve appears the operation of the regulation.

  It is thus noted that during the stroke of the piston of the high pressure pump, when the flow control valve is connected, that is to say when the connecting pipe 133 is closed, there is a pressure increase to a maximum of instant at which the preset setpoint of the ramp pressure is reached, i.e. p_R_cons. Then the flow control valve 132 is shut off by the control apparatus 150, i.e., the connecting line 133 is opened and thus the unconsumed fuel is returned from the chamber 142 into the low pressure circuit. , in particular in the damping element 134. The pressure curves are shown theoretically in FIGS. 2-4, that is to say without taking into account the damping effect produced by the ramp. In reality, these pressure curves do not include a fold (sudden angle variation) but are corrugated in a sinusoidal shape.

  If at a high speed according to Figure 2 there is a long injection by the control of the flow valve 132, that is to say when the connecting line is closed, in the piston stroke phase such as the injection time ti of the cylinder 1, ti_1, then the pressure of the ramp p_R increases more slowly for 3) than for no injection 1) or for low injection times. If from a certain position of the piston stroke the flow control valve 132 is cut off because the set pressure of the ramp p_R_cons is reached, then the pressure of the ramp decreases from this moment during that the injection continues ti_1 and also in the direct-ment following injection ti_2 of the cylinder 2 according to the curve 4 given in Figure 1, because in this phase the piston 139 sucks and does not transfer fuel into the ramp 110.

  The basic calculation of the injection times ti is done each time at the instant of a trigger mark tr. This is indicated in FIG. 2 by vertical lines in broken lines at the trigger marks and by the arrows corresponding to these lines and directed downwards. The calculation of the injection times is then done each time according to the most current available value of the measured pressure of the ramp, that is to say the last measured pressure in the time frame. The ramp pressure is entered each time in a renewed manner, preferably every millisecond, in particular for an internal combustion engine running at a speed of 6000 rpm. In practice, for the curve of the ramp pressure shown in FIG. 2, this means, for example, that at the trigger mark tr_4 the injection time ti_2 for the cylinder 2 is calculated using the pressure of the ramp, the last available, that is to say that the value of the ramp pressure p_R_R2 is calculated. Based on this, the injection time for the cylinder 3 is calculated based on the detected ramp pressure p_R_R3. In FIG. 2 it appears that the ramp pressure values used for the calculation do not correspond to the average pressure values p_R_Mi during the injection times ti_i in the cylinders i.

  FIG. 3 shows in principle the evolution over time of the same physical quantities as in FIG. 2 but for the starting phase of the internal combustion engine and not for the stabilized state of the internal combustion engine running at steady state. high and for a high fuel flow. The same physical magnitudes carry the same references here. For the basic description of the timing diagrams, see the description of FIG. 2.

  In addition to the already known curves, the lower part of FIG. 3 shows the engine speed nMot of the internal combustion engine during the starting phase, reported synchronously with the angular plane of the other physical quantities. At the beginning of the start-up phase, there is first a relatively low rotational speed; this rotational speed corresponds to a ramp pressure which is low p_ND and corresponds to the low pressure system as shown by the comparison of the diagram in the upper part and that of the lower part of figure 3. In the start-up phase , especially for still low ramp pressures, during the pressure rise, after start-up, considerable errors can be made in the calculation of the injection time. This is represented as follows by way of example in FIG. 3. The average ramp pressure p_R_M3 during the injection in the cylinder 3 corresponds relatively well to the detected ramp pressure p_R_R3 while for the following injection which occurs at the end of a piston stroke, the pressure p_R_M4 is significantly higher than the calculated pressure p_R_R4.

  As shown in FIGS. 2 and 3, in general, during common injections, especially in the particular situations represented by way of example in FIG. 3, there are large variations in the pressure of the ramp and this is why It is necessary to correct the injection time as a function of the effective pressure of the boom. Only in this way can the amount of fuel to be injected into a cylinder be accurately metered. According to the invention, two methods are used for this purpose. However, beforehand, we will discuss a mathematical presentation.

  The quantity or dose of reference fuel qK_cons at the instant of calculation ti- for the reference mark t_R-reference is obtained as follows in the case where at the instant of calculation there is the following pressure difference between the ramp and the combustion chamber 3.p_RB_R: qK_cons = ti_R ÉK_EV É-JAp_RB_R. (1) By inverting this formula, the injection time is obtained: ## EQU1 ## in this formula: K_EV constant of the injector ti_R injection duration calculated at the time of calculation of the duration ti for the t_R-landmark.

  using: Ap_RB_R = p_R_R p_B_R, (3) in which p_R_R represents the ramp pressure at the instant of calculation of t_P_B_R represents the pressure of the combustion chamber at the instant of calculating ti-.

  For the injection in the suction phase can be taken as pressure of the combustion chamber, by approximation, the atmospheric pressure and consider it as constant. In addition, for this calculation of ti it is assumed for the mark t_R-mark that the fuel pressure is equal to the pressure detected last and that it is constant for the entire injection time.

  Until the moment of the injection and while the injection occurs, the pressure in the ramp changes as described above especially with reference to FIGS. 2 and 3. Therefore, it is proposed to detect this modified pressure in the ramp, preferably every millisecond and use the actual detection value thus obtained to calculate a corrected injection duration. On the basis of this current value of the ramp pressure it is necessary not only to carry out a single correction calculation of the injection duration just before the start of the injection but also to carry out a permanent correction during injection with the injection time, calculated last, until the injection time again last updated has elapsed. For this, the invention provides a simple correction formula.

  To determine this correction formula it is necessary to know what dose of fuel qK is ejected from the ramp under a variable pressure. This fuel dose is calculated as follows: qK = K _ EV EJ -JAp _ RB (t) E dt,, 14 (4) to in this formula K_EV represents the constant of the injector Ap_RB (t) represents the difference pressure between the ramp and the combustion chamber at time t.

  By digitizing the sum formula, for a detection frame At of, for example, 1 ms, the following formula is obtained: k qK = K_EV É1VAp_RB (j) EAt (5) j = 4 If we consider the duration of injection , by approximation, as a multiple of At, we obtain for ti = k E At or by transforming, for At = ti / k: k qK = K_EV Eti / kE> -Jap_RB (j), (6) l = in this formula: k represents the number of detections during an injection. For the target fuel dose qK_cons and for the corrected value ti_K we obtain: k qK cons = K EVÉti_K / kÉÉ.JAp_RB (j). (7) j = 1 If we eliminate qK_cons from equations (1) and (7), we obtain the corrected injection duration ti_K: k tiK = ti_RÉ JAp_RB_R Ék / E-JOp_RB (j) i = l = ti _R KF (8) In the formula 8, the correction coefficient KF represents the ratio between the injection time ti_K corrected, according to a current calculation and the injection time ti_R, initially calculated for the triggering mark R. Figure 4 explicit this correction of the injection duration. It appears in particular that the injection time ti_R1 initially calculated for the cylinder 1 on the trigger mark tr_3 based on the pressure difference determined just before the pressure of the ramp and the pressure of the combustion chamber Op_RB_R1 for the cylinder 1 will be corrected again afterwards. This injection time initially calculated is shown in FIG. 4 as the second horizontal bar from the top. This correction finally leads to a reduction of the injection time ti_K1 as indicated by the horizontal beam placed above.

  FIG. 4 further shows that the taking of the pressure according to the time ramp of 1 millisecond, hypothesized, is not done in a synchronous manner with the instants of triggering since these instants and the corresponding start times for the injections are connected to fixed landmarks of crankshaft angle. To be able to make a correction calculation also for very short injection times corresponding for example only to 0.5 ms, it is not enough to perform a correction calculation only after the start of the injection. The injection could indeed be completed before the correction calculation is made. This is why it is necessary to take an initial value and already make a first correction of this initial value of the duration of injection before the very beginning of the injection.

  The calculation of the correction formula (8) is advantageously done by determining the expression in the counter without the coefficient k already in the frame tR for the calculation of ti_R. Moreover, the sum of the numerator is formed as a distinct quantity, and each time the new value of the pressure of which one takes the root is added. Then, the counter is divided by the new sum and the result is multiplied by the index obtained k. In this way, at each instant of the frame at 1-ms we will have with less calculation to perform, a current value corrected ti_K.

  The correction of a transmission clock ti to control the injectors 120 in Figure 3 according to the calculated injection times can be done either in each frame of 1-ms with each time a recalculated injection time ti_K up to in that way, the corrected current injection time is so mined (see the upper line V1 in FIG. 4). Alternatively, it is also possible to correct the transmission clock only in the respective 1-ms frame in which the currently valid injection time t i_A is to add only one increment of time 8t_TK to the corrected injection duration ti_K , the increment of time being necessary to correct the transmission clock (see V2).

  Thus, during the current injection ti_R1 in the cylinder 1 and according to the stepped curve shown in FIG. 4, the integral of the roots of the detected pressure values is formed along the pressure curve Ap_RB and corrected. the injection as a function of the true value ti_K1. In Figure 4 this results in a reduction in the duration ti because of the increase in pressure. In the opposite case, if the pressure decreases, we would have an extension of the duration ti.

  If we want to reduce the error by the detection of the pressure, each time we form the root for the calculation of the numerator in the formula 8, we can use according to the stepped curve in the right part of the figure 4, instead of the new pressure value, detected, this value increased by half the difference between the new and the old detection value (for example for j = 5: (Ap5- 4) / 2 = (p (j = 5) - p (j = 4)) / 2 Thus, the detection curve curve passes better on average by the effective evolution of the pressure and one can choose possibly a slower calculation frame by For example, a 2ms frame, we arrive at the following formula: k ti_K = ti_RÉ. \ /Ap_RB_R.klEiMp_RB (j) + (p_R (j) -p_R (j-1)) l2 i = 1 The main purpose of the calculation The correction procedure according to equation (8) is to allow a rise in pressure as fast as possible during start-up, since starting and phase after start-up usually homogeneous njection, the pressure in the combustion chamber is constant and for Ap_RB one can take the pressure of the ramp reduced by the atmospheric pressure.

  However, the correction calculation can also be done after all the pressure in the fuel has been established in the stratified modes and homogeneous mode operating modes. In this case, instead of a constant pressure in the combustion chamber, the pressure in the combustion chamber can be used at the time of injection. Or for that, one can use the pressure in the combustion chamber (9) for the injection angle used for the calculation of ti_R and corresponding to the time t_R-mark.

  Alternatively, according to the same time frame of lms as for the input of the pressure of the ramp, one can either calculate an instantaneous pressure in the combustion chamber using the compression ratio and the position of the crankshaft, or measure this pressure by a pressure sensor associated with the combustion chamber. Thus, by using additional calculation means, the pressure curve in the combustion chamber will be taken into account even more accurately than for the prior calculation of the injection duration.

  If the operating time reserves of the microcomputer used are limited, it is possible to cut the correction calculation described above when a certain speed threshold is exceeded.

Claims (16)

  1) A method of managing an internal combustion engine, in particular a motor equipping a motor vehicle comprising N cylinders in which the fuel is directly injected, the method comprising the following steps: seizing the fuel pressure in a fuel accumulator (ramp) before the injection operation to be triggered, preliminary calculation of the injection time ti for a cylinder (n) with n = 1 ... N for the injection operation to be triggered on the basis of the detected boom pressure and any other quantities, characterized in that relying on at least one ramp pressure value detected just before and / or during the injection operation and on any other quantities, the predetermined value of the injection time is corrected before the end of the injection time at least once in a corrected injection time ti_K for the injection in progress in the cylinder n.   2) Method according to claim 1, characterized in that one takes the ramp pressure before and / or during the injection operation in a fixed time frame, for example every 1 ms.   3) Process according to claim 1, characterized in that the injection time ti of the cylinder n is pre-calculated each time at the tripping instants tr predefined by a fixed frame of crankshaft angle and lying far before the injection into the cylinder n according to the angular position of the injection, and the calculation is carried out based on the ramp pressure value, 30 entered, preferably the last value entered before the triggering time.   4) Method according to claim 1, characterized in that the calculation of correction of the injection time occurring just before the start and if necessary during the injection in progress, each time directly after having seized the pressure ramp according to the time frame.   5) Method according to claim 1, characterized in that the ramp pressure signal used for calculating and / or the correction of the injection duration is a ramp pressure signal filtered by a low-pass filter.   6) Process according to claim 1, characterized in that the corrected value ti_K of the injection time is calculated from the ramp pressure value minus a pressure value as another quantity representing the pressure. in the combustion chamber.   7) Process according to claim 6, characterized in that for an internal combustion engine operating in homogeneous mode it is assumed that the pressure in the combustion chamber is constant and corresponds to atmospheric pressure.   8) Method according to claim 6, characterized in that for types of injections for which is injected during the compression phase, the pressure in the combustion chamber is measured according to the same time frame as for the ramp pressure, calculated or evaluated from the crankshaft position and the compression ratio, or measured using a combustion chamber pressure sensor.   9) A method according to claim 1, characterized in that the corrected injection time ti_K is calculated by applying the following formula: k tiK = tiR .. \ / OpRBRÉk / iJ4pRB (j) j = 1 in which tiK duration d corrected injection, ti_R injection time calculated beforehand at a moment of tripping in angular synchronism, Ap_RB_R pressure difference between the pressure in the ramp and the pressure in the combustion chamber at the moment of tripping, k number current detections during an injection operation, i synchronous time detection instant located directly before or / and during the injection in progress, ap_RB (j) pressure difference between the pressure in the ramp and the pressure in the combustion chamber at the instant of detection j.   10) Method according to claim 1, characterized in that the corrected injection time ti_K is calculated by applying the following formula: k ti_K = ti_RÉ. \ IAp_RB_R Ékl1 - \ / Op_RB (j) + (p_R (j ) - p_R (j -1)) / 2 j = 1 in which p_R (j) represents the value of the pressure of the ramp at the detection time j, ti_K corrected injection time, ti_R calculated injection time beforehand at a moment of tripping in angular synchronism, Ap_RB_R pressure difference between the pressure in the ramp and the pressure in the combustion chamber at the moment of tripping, k current number of detections during an injection operation, time synchronous detection instant located directly before or / and during the injection in progress, Op_RB (j) pressure difference between the pressure in the ramp and the pressure in the combustion chamber at the instant of detection j .   11) Method according to claim 1, characterized in that after each correction calculation of the injection duration is updated a transmission clock ti with the last value of the injection duration until one has reached the end of the injection duration continuously updated by the repeated corrections.   12) A method according to claim 1, characterized in that one updates a transmission clock ti only if the injection duration 15 continuously updated by the correction calculation, less a duration for the correction of the clock emission, is complete.   13) Method according to claim 1, characterized in that the correction calculation is no longer performed above a rotation speed threshold or is switched from a calculation of multiple corrections to a single correction calculation.   14) Injection device (100) for an internal combustion engine having N 25 cylinders, comprising: - a fuel accumulator (ramp) (110) for accumulating fuel and supply it by injectors (120) to the combustion chambers of the internal combustion engine, a high pressure pump (130) for pumping fuel under high pressure into the fuel accumulator, a pressure sensor (140, 140 ') for capturing the actual fuel pressure in the accumulator of fuel (110) and a control apparatus (150) for regulating the actual pressure sensed by the pressure sensor (140, 140 ') in the fuel accumulator (110) to a predefined ramp pressure value by the actuator control (132) of the high-pressure pump (130) and for the preliminary calculation of an injection time for a cylinder n (n = 1 ... N) of the internal combustion engine , characterized in that the control device (150) corrects the calculated value in the field alable of the injection duration before the end of it by relying on at least one ramp pressure value entered just before or / and during the injection phase and any other quantities, at least once in a corrected injection time ti_K for the injection in progress in the cylinder n.   15) Injection device (100) according to claim 14, characterized in that a filter equipping the input circuit of the control device uses the signals of the pressure sensor (140), including a low-pass filter (142) as an element (RC) for damping the curve of the ramp pressure.   16) Injection device (100) according to claim 14, characterized in that the fuel accumulator (110) comprises a hydraulic low-pass filter in the form of a throttle (144) and a intermediate fuel accumulator (146) for damping the high frequency variations of the ramp pressure, and the pressure sensor (140 ') is connected to the intermediate fuel accumulator (146) for detecting a pressure curve of ramp, damped.