FR2917460A1 - Indicated average torque correcting method for e.g. petrol engine, involves controlling quantity of fuel to be injected based on comparison result between setting torque value and indicated average torque - Google Patents

Abstract

The method involves determining a setting torque value, and estimating an indicated average torque in an engine cylinder. A comparison between the setting torque value and the indicated average torque is carried out, and a value of the corrected setting torque is calculated. Quantity of fuel to be injected is determined by a mapping, and the quantity of the fuel to be injected is controlled based on the comparison result between the setting torque value and the indicated average torque.

Description

The invention relates to engines for rotary combustion vehicles

  whether they are gasoline or diesel and direct or indirect fuel injection. The number of cylinders is indifferent. It is known that the amount of fuel injected determines the characteristics of the combustion. It has a direct influence on torque performance, the level of pollutant emissions and the engine thermodynamics. The main benefits of the engine (pollution, performance, consumption and approval) are therefore directly related to controlling the amount of fuel injected. Whatever the torque demand, it is difficult to control the achievement of this torque by the engine. Indeed, the injectors all have different characteristics. In addition, the characteristics of the injectors evolve during their aging, thus generating drifts. In addition, each cylinder has dispersions, especially in its dimensions, which affect how to achieve torque. In addition, the thermodynamic, aerodynamic and chemical conditions are different in each cylinder. In the prior art, there are known devices such as that of document US Pat. No. 6,398,692, which provides a torque regulation for the purpose of limiting the maximum quantity to be injected. His goal is not to control the couple. In addition, the regulation is performed by means of an open loop circuit which is sensitive to drifts and dispersions of the engine. A device of this type does not make it possible to correct in real time the indicated torque of the motor. An object of the invention is to improve the control of the torque provided by the engine, by allowing the estimation and correction in real time of the torque indicated in each cylinder of a gasoline or diesel engine. For this purpose, it proposes a method for correcting the average torque indicated in a cylinder of a vehicle engine, in which: a) a set torque value (Cc) is determined; b) estimating a value of the indicated average torque (Ci) in at least one cylinder i; c) a comparison is made between the torque reference value (Cc) and the estimated torque value (Ci); d) and controlling a quantity of fuel to be injected (Qia) taking into account the comparison made in step c). In a first embodiment, between steps c) and d), a corrected torque reference value (Ccorr) is calculated and the quantity of fuel to be injected (Qia) is determined by means of a mapping. In a second embodiment between steps c) and d), the quantity of fuel to be injected (Q ;,) corresponding to the torque setpoint (Cc) is determined by means of a mapping; from the comparison made in step c), the correction value of the fuel quantity Qi error is determined, the amount of fuel to be injected (Q, a) is calculated from Q, c, and of error. The method is implemented in real time. In a first variant, the method is implemented cylinder 20 after cylinder. In a second variant, the method is implemented simultaneously on all the cylinders. Finally, the process may be carried out cycle to cycle or on q cycles. 2. Preferably, the value of the indicated average torque of the cylinder is determined from at least one characteristic variable of the rotational movement of the engine. . Thus, in an engine comprising a position sensor composed of a target provided with patterns and integral with a moving motor element in rotation, and a sensitive element fixed to the engine block, said sensor delivering an alternating frequency signal. proportional to the speed of movement of the target patterns in front of the sensing element, a relative value is estimated for a torque generated by a cylinder i from the equation:

  k, jk, j + a0, k = 9; wherein: C is the indicated average torque of the cylinder i during a combustion cycle; Nk, j is a function of at least one magnitude characteristic of the rotational movement of the motor; io ak, j is a weighting coefficient of the magnitude / 3k, first-order dependent of the average engine speed; ao, 1 is a variable dependent on the first order average engine speed; - q; and r; respectively denote the number of the first pattern and is the number of the last pattern perceived by the sensing element of the position sensor during the combustion of the cylinder i defining the angular window for analyzing the engine torque associated with the combustion of the cylinder i; 8; is a weighting coefficient. According to a first variant, a value relative to a torque generated by a cylinder i is estimated from the equation: k = g; kLtk + ao (E4) in which: - Ci is the indicated average torque of the cylinder during a combustion cycle; Atk is a period of rotational movement of the motor; ak is a weighting coefficient of the rotational movement time of the engine, depending first-order on the average engine speed; ao is a variable dependent on the first order average engine speed; q; and r; respectively denote the number of the first pattern and the number of the last pattern perceived by the sensing element of the position sensor during the combustion of the cylinder i defining the angular window for analyzing the engine torque associated with the combustion of the cylinder i. According to a second variant, a relative value is estimated for a torque generated by a cylinder i from the equation:

  Ci = akwk + ao (E3) = 9r in which: C is the indicated average torque of the cylinder i during a combustion cycle; - Wk is an instantaneous speed of rotation of the motor; ak is a coefficient of weighting of the instantaneous speed of rotation of the motor, dependent on the first order of the average engine speed; ao is a variable dependent on the first order average engine speed; qi and r; respectively denote the number of the first pattern and the number of the last pattern perceived by the sensing element of the position sensor during the combustion of the cylinder i defining the angular window of analysis of the engine torque associated with the combustion of the cylinder i. Other features and advantages of the invention will become apparent in the following description of a preferred embodiment of the invention, given by way of non-limiting example, which description is made with reference to the accompanying drawings on which: - Figure 1 is a diagram of a part of an engine to which the method according to the invention is applied; FIG. 2 is a flowchart illustrating the principle of the unfolding of the method according to the invention within the engine of FIG. 1; FIGS. 3, 4 and 5 illustrate the method for estimating the average torque indicated for a cylinder and FIG. 6 illustrates a step of the method illustrated in FIG. 2. FIG. 7 illustrates an alternative embodiment of FIG. the step illustrated in Figure 6. The device relating to this invention is shown in Figure 1. It applies to a diesel or gasoline internal combustion engine comprising a movable piston 4, a connecting rod 5, a combustion chamber 3, a gas intake device 2, an air flow sensor 13 making it possible to estimate the mass of gas admitted by the engine, a gas exhaust device 1, a wealth sensor 14 allowing the measurement of the richness of the air-fuel mixture, a means 16 for measuring the angular position of the crankshaft 7, an injector 8, a device connecting the injector to pressurized fuel 12, a means for measuring the fuel pressure 10, a means for to measure the fuel temperature 11 and u n electronic computer 6. The position of the accelerator pedal 9 allows the electronic computer 6 to control the engine according to the will of the driver. A preferred embodiment of implementation of the method according to the invention is detailed in Figure 2. The instantaneous position and speed of the crankshaft are measured via the position sensor 16. This instantaneous position is connected to the calculator electronic 6.

  On the other hand, in the electronic computer 6, a certain number of computations and algorithms are made, as detailed in FIGS. 2, 6 and 7. The position of the accelerator pedal 9 makes it possible to interpret the will driver in the form of a setpoint of power, torque or engine speed. In step 20, the method interprets the will of the driver and results in the development of a torque command Cc to be provided by the crankshaft. More specifically, the will driver is arbitrated against the competing demands of other computers of the vehicle (calculator of BVA, ESP, lo ACC etc ...). After arbitration and regardless of the type of interpretation of the will of the driver or vehicle requests, the electronic computer 6 always leads to the development of a torque setpoint Cc. The set torque can be the torque of a particular cylinder or the general torque that all cylinders will have to reach. In practice, the setpoint torque is common to all cylinders. On the other hand, it is independent of the choice of the type of calculation of the estimated torque: setpoint torque on n cylinders or setpoint torque of the cylinder i. In the next step 21, the electronic computer 6 determines from the set torque value Cc a quantity of fuel to be injected 20 Qia. This determination is also made from the indicated torque measured as will be explained below with reference to Figures 6 and 7. It thus adjusts if necessary the fuel charge to be injected so as to achieve the torque setpoint. In the next step 22 of FIG. 2, the method receives the fuel quantity instruction Qia resulting from the preceding step, as well as information on the operating point of the engine. This information corresponds to all the variables of state such as the speed, the torque, the air pressure, the air temperature, the water temperature, the state of the pollution control systems, the loading of the filter particulate matter, 30 NOx trap, etc. In step 22, the method determines the injection pattern that will be used for fuel injection into the cylinders. This pattern includes the number of injections and for each of these injections a flow and a phasing. In the next step 23, the electronic computer 6 interprets the fuel mass setpoint to be injected Qia to determine the duration of activation of the injector Ti. For this, a map is used which uses as input data the pressure P, and the temperature Tc of the fuel from the sensors 22 and 24 and the amount of fuel to be injected. During the activation time thus determined, the injector 8 is then activated. The electronic computer 6 also synchronizes and angularly positions the control of the injector according to the angular position 8 of the crankshaft. From the instantaneous speed w of the crankshaft or the evolution of the angular position 6 of the crankshaft over time, the electronic computer 6 then estimates at block 24 the gas torque also called the indicated torque of each of the n cylinders constituting the engine. The embodiment of this step 24 will be described with reference to FIGS. 3, 4 and 5. The method for estimating the average indicated torque C, produced in its own right by each combustion in a cylinder i, uses at least one characteristic magnitude of the movement. motor rotation. It implements a position sensor composed of a target provided with patterns and integral with an element of the moving motor in rotation, and a sensitive element fixed to the engine block, said sensor delivering an alternating frequency signal proportional to the speed of movement of the patterns of the target in front of the sensing element. This method is based on an equation using a characteristic quantity of the rotational movement of the motor. This characteristic quantity is transmitted by the angular position sensor.

  The general equation of the estimation method is: Cl = J, J r k,. + A0 ,. wherein: G is the indicated average torque of the cylinder i during a combustion cycle; j3k, j is a function of at least one characteristic quantity of the rotational movement of the motor; ak, l is a weighting coefficient of magnitude / 3k dependent on the first order of the average engine speed; a, ~ is a variable dependent on the first order average engine speed; q; and r; denote respectively the number of the first pattern and the number of the last pattern perceived by the sensitive element of the position sensor during the combustion of the cylinder i defining the angular window for analyzing the engine torque associated with the combustion of the cylinder i; (Si is a weighting coefficient A particular embodiment of the measurement of the torque will be described with reference to FIGS 3 to 5. A target 37 connected to the flywheel is integral with the crankshaft, so rotates with it, and presents teeth 39 on its periphery which pass opposite the sensitive element of the magneto-reluctance-based angular position sensor 16. The sensor 16 measures the duration of passage of each tooth 39 of the ring gear in front of the sensitive element of the sensor Within the calculator 6, the inverse of the value obtained is calculated and the result is multiplied by the value of the angular sector of the corresponding tooth, more specifically, the duration Atk corresponds to the time which elapses between a front (amount or downstream) of the signal emitted by the position sensor and the following homologous front as

  9 is illustrated in FIG. 5. This figure illustrates a portion of the target 37 with teeth 39 in the lower part and then above, the raw signal emanating from the sensor, approaching a sinusoid and finally, again above, the signal of the sensor after treatment and allowing the detection on rising edge. This duration is associated with the tooth Dk, occupying the angular position 8k, and angular width Aek of the target 37. As illustrated in block 38 of FIG. 3, the angular velocity Wk associated with the tooth Dk is then obtained by the formula : = 08k wk (E2) Otk Next, the computer 6 will relate the different instantaneous speeds thus obtained as shown in block 40 of FIG. 3 or 4. For this, the speeds are added after having been weighted by coefficients ak . The calculation is made of the average torque indicated (or gas torque) developed by the cylinder i of the engine comprising p cylinders according to the following formula: akwk + ao (E3) The calculation of the average torque indicated according to formula (E3) has advantages . Thus, the angles 6k of the teeth Dk of the target 37 can be arbitrary. The calculation of the indicated torque thus achieved is not affected by angular defects of the steering wheel, problems of roundness, the size of the long tooth traditionally arranged on this type of steering wheel, or possible defects of the filter electronics of the sensor signal (problem of the fronts after a long tooth, for example). These advantages come from the taking into account of these defects in the coefficients ak. These coefficients are predetermined and here dependent on the first order of the average engine speed wo. To determine the coefficients ak, one can use a calculation function or a map dependent on the engine speed. A good development of the coefficients i0

  ak makes them independent of environmental parameters of the engine, which is an important advantage. These parameters will be for example:

  - the rate of recirculated exhaust gas; - the phasing of injections;

  - the quantity of fuel injected;

  the temperature of the air leaving the compressor;

  - the temperature of the flue gases at the exhaust;

  - the temperature of the recirculated exhaust gas;

  io - the engine water temperature;

  - the engine oil temperature;

a temperature before the turbine;

  - intake manifold pressure or exhaust manifold pressure.

  This calculation also makes it possible to estimate the average torque indicated with great precision. Thus, it is possible to achieve accuracy with a risk of error of less than 1%.

  In a variant of the implementation of the torque determination steps, the computer 6 measures the instantaneous duration IMk required for

  20 passage of each tooth in front of the sensor. This duration corresponds to the time that elapses between a front of the signal emitted by the position sensor and the next homologous front. As previously, this duration is associated with the tooth Dk, occupying the angular position ek and angular width Mk of the target 37. As before, a block 40 is

  Weighted values thus obtained using this time the formula:

  VS; lakAtk + ao (E4) k = q; The Atk times and Aek angles associated with the teeth Dk of the target 37 may be arbitrary and the same advantages are found as in the previous embodiment. The coefficients ak have the same properties

  And are obtained in the same way. 2917460 It Thus, the estimate C; the resultant torque resulting from step 24 is used via the feedback loop referenced in step 21 to equalize the setpoint torque C, and the measured torque C ;. This adjusts the amount of fuel to be injected. This step 21 will now be explained in detail with reference to FIG. 6. In the present example, the setpoint torque value C, is used in a mapping to determine a fuel quantity setpoint to be injected Q. This mapping has has been established in advance by development tests. This mapping step is illustrated in block 26. This mapping takes into account the phasing of the injections, the desired injection mode of the other engine parameters that can influence the relationship between the torque and the fuel to be injected. The estimated value C; the indicated torque arrives via the feedback loop 25 to a subtracter or comparator 27 which makes the difference between this value and the torque setpoint Cc. The difference thus obtained, called regulation error, is transmitted to a regulator 28. On the basis of this value, the regulator 28 performs a certain number of calculations such as typically integration, derivation, multiplication, filtering, etc. This leads to the development of a correction value to be applied. The error value of the injection quantity Q, c from the block 26 and the error correction value Q1 from the regulator 28 are then applied to an adder 29 which in turn deduces the applied quantity of injection quantity Qia which constitutes the output of the block 21. Thus, through the adder 29, the setpoint Q, receives an error correction which makes it possible to equalize the indicated torque measured at the reference torque. This amounts to canceling the control error of the indicated torque. In the present example, the method according to the invention also includes a learning process through the referenced loop.

  12 which sends the error correction value outgoing from the regulator 28 in the block 26 where the mapping determination is provided. In this case, a learning process is also carried out through the loop referenced 31 which sends the quantity order to be injected to apply Qia coming from the adder 30 to the loop referenced 31. The step 24 illustrated in FIG. 2 is applicable. to all the n cylinders composing the engine from an indicated torque information measured over a number q of cycles that can vary from 1 to infinity. It also applies to the regulation of the torque indicated cylinder cylinder, cycle to cycle or over several cycles. This last approach requires a different map for each cylinder. This is the mapping of block 26 (FIG. 6) or block 32 (FIG. 7). The estimation method described with reference to FIGS. 3 to 5 makes it possible to calculate the average torque indicated on any angular window.

  This window can represent several cycles. Therefore, the general equation (E1) to (E4) makes it possible to know the torque on the n cylinders of the motors and on q cycles. On the other hand, the calculation of the torque of the cylinder i on q cycles requires the calculation of the average torque indicated on the angular window associated with the combustion of the cylinder i. Then it requires a calculation of average over q cycles. In other words, it is then necessary to average the average pairs measured at each cycle. In the case of the indicated average torque estimated on n cylinders, the quantity correction to be injected is an average correction. It will act on the setpoint of the quantity to be injected, essentially the setpoint of the main injection, globally and globally correct the drift of the injectors. This corrected amount of injected quantity is perceived in the same way by all the injectors, it is therefore identical on all the cylinders.

  This correction is not the best suited for each of the cylinders. Indeed, each cylinder has a drift of its own, that is why it is possible to correct the setpoint of the amount injected per cylinder, using the measurement of the average torque indicated cylinder i. The first solution may be of interest when the strategy associated with a cylinder generates too many instabilities or disparities because of poor control of the torque measurement, for example. In this case, a general correction may be more robust. FIG. 7 illustrates an implementation variant of step 21 illustrated in FIG. 6. The implementation is identical to that of FIG. 6 except for the following differences.

  The comparator 27 and the regulator 28 extend this time upstream of the block 26 corresponding to the use of a mapping to go from a set torque value to a fuel setpoint to be injected. Thus, the result of the comparison made to the comparator 27 is sent to the regulator 28 which itself sends its result in the form of a 1s torque setpoint to the block 26. The block 26 therefore does not receive a feedback loop as that identified by the reference 30 or 31 in Figure 6. Thus, the torque setpoint C, from block 20 by interpreting the will of the driver and the arbitration of the vehicle requests is compared to the indicated torque C; estimated by the torque measuring device. The error calculated by the comparator 27, between these two values, is communicated to the regulator 28 which processes this signal in the form of a series of integrations, derivatives and gain multiplications. The regulator output supplies a new torque set value Ccorr to block 26. Block 26 outputs the amount of fuel to be injected Q, which will correct the average torque indicated. A method for learning the torque error is provided at the block 32. The latter receives, on the one hand, the torque setpoint from the block 20 and, on the other hand, the torque setpoint coming out of the regulator 28, to supply feedback to the summator 33, 30 information which will be added to the output signal of the regulator 28.

  It is also possible to provide a method for learning the corrected torque setpoint at block 32. The latter then receives, on the one hand, the torque setpoint from block 20 and, on the other hand, the corrected torque setpoint. Ccorr from the summoner 33.

  As can be seen, the method of the invention interprets the torque setpoint into a quantity of fuel to be injected. It also compares the torque setpoint with the indicated torque measured and adjusts if necessary the fuel setpoint to be injected so as to achieve the torque setpoint. The invention thus provides a good control of the torque actually provided by the crankshaft. Of course, we can bring to the invention many changes without departing from the scope thereof.

Claims (17)

  A method of correcting the average torque indicated in a cylinder of a vehicle engine, wherein: a) determining a set torque value (Cc); b) estimating a value of the indicated average torque (C;) in at least one cylinder i; c) a comparison is made between the torque reference value (Cc) and the estimated torque value (C;); d) and controlling a quantity of fuel to be injected (Qia) taking into account the comparison made in step c).   2. Method according to claim 1, in which: between steps c) and d), a corrected torque reference value (Ccorr) is calculated and the quantity of fuel to be injected (Qia) is determined via mapping.   3. Method according to the preceding claim, wherein: - from the comparison made in step c), the value of the torque error is determined by means of a regulator (28).   4. Method according to one of claims 2 or 3, wherein: - the value of the corrected torque (Ccorr) obtained by a learning method is sent to the mapping for determining the quantity of fuel to be injected ( Qia).   5. Method according to claim 1, in which: between steps c) and d), the quantity of fuel to be injected (Q ;,) corresponding to the torque setpoint (Cc) is determined by means of mapping; From the comparison made in step c), the correction value of the amount of fuel Qi error is determined, - the quantity of fuel to be injected (Qia) is calculated from Q, and the error .   6. Method according to the preceding claim, wherein: - the correction value of the amount of fuel Qi error is determined by means of a regulator (28). 10   7. Method according to one of claims 5 or 6, wherein: - is carried out by sending the fuel quantity correcting value (Qierreur) to the mapping to determine the amount of fuel to be injected (Q ;vs). 15   8. Method according to one of claims 5 to 7, wherein: - a learning is carried out by sending the value of the amount of fuel to be injected (Q; a) to the mapping to determine the amount of fuel to inject (Q c). 20   9. Method according to one of claims 1 to 8, implemented in real time.   10. Method according to one of claims 1 to 9, implemented cylinder by cylinder.   11. Method according to one of claims 1 to 9, implemented on all cylinders.   12. The method of claim 10 or 11, implemented ring cycle. 17   13. The method of claim 10 or 11, implemented on q cycles, q> _ 2.   14. A method according to any one of claims 1 to 13, wherein the value of the indicated average torque of the cylinder is determined from at least one characteristic value of the rotational movement of the engine.   15. The method of claim 14, wherein the motor lo comprises a position sensor composed of a target provided with patterns and secured to a movable motor element in rotation, and a sensitive element attached to the engine block, said sensor delivering an alternating frequency signal proportional to the speed of movement of the patterns of the target in front of the sensitive element, characterized in that a value relative to a ls generated by a cylinder i generated from the equation:, Jfl k, J + a0, i wherein: Ci is the indicated average torque of the cylinder i during a combustion cycle; Nk, J is a function of at least one characteristic quantity of the rotational movement of the motor; ak, ~ is a weighting coefficient of the first-order dependent magnitude Nk of the average engine speed; a01 is a variable dependent on the first order average engine speed; - q; and r; respectively denote the number of the first pattern and the number of the last pattern perceived by the sensing element of the position sensor during the combustion of the cylinder defining the angular window for analyzing the engine torque associated with the combustion of the cylinder i; 8j is a weighting coefficient.   16. The method according to claim 15, wherein a value relative to a torque generated by a cylinder i is estimated from the equation: k = 9, in which: lo Ci is the indicated average torque of the cylinder i at during a combustion cycle; Atk is a period of rotational movement of the motor; ak is a weighting coefficient of the rotational movement time of the engine, depending on the first order of the average engine speed; ao is a variable dependent on the first order average engine speed; q; and r; denote respectively the number of the first pattern and the number of the last pattern perceived by the sensing element of the position sensor during the combustion of the cylinder i defining the angular window for analyzing the engine torque associated with the combustion of the cylinder i.   17. A method according to claim 15, wherein a value relative to a self generated torque by a cylinder i is estimated from the equation: ## EQU1 ## where: Ci is the indicated average torque of the cylinder i in the course of a combustion cycle; wk is an instantaneous speed of motor rotation; ak is a coefficient of weighting of the instantaneous speed of rotation of the engine, dependent on the first order of the average engine speed; ao is a variable dependent on the first order average engine speed; qi and r; respectively denote the number of the first pattern and the number of the last pattern perceived by the sensing element of the position sensor during the combustion of the cylinder i defining the angular window for analyzing the engine torque associated with the combustion of the cylinder L 19