Method for estimating the crank angle at which 50% of the fuel mass has been burnt in a cylinder of an internal combustion engine with spontaneous mixture ignition

Abstract

A method for determining the angular speed of a drive shaft of an internal combustion engine at each tooth event of a phonic wheel presenting a number N of teeth includes detecting the angular amplitude for each tooth; detecting the time of each tooth; and determining the raw angular speed of each tooth according to the corresponding angular amplitude and time. For each tooth a compensation value is determined which represents the difference between the actual angular amplitude of the tooth and the theoretic angular amplitude of the tooth; and determining the angular speed of each tooth by correcting the raw angular speed based on the corresponding compensation value.

Description

TECHNICAL FIELD

The present invention relates to a method for estimating the crank angle at which 50% of the fuel mass has been burnt in a cylinder of an internal combustion engine with spontaneous mixture ignition.

BACKGROUND ART

In internal combustion engines with controlled mixture ignition (i.e. in internal combustion engines operating according to the “Otto” cycle and fed with gasoline, methane, LPG or the like) it has been known for many years to use as control magnitude the MFB50 (50% Mass Fraction Burnt) combustion index which represents the crank angle at which 50% of the fuel mass has been burnt inside a cylinder. Indeed, in an internal combustion engine with controlled mixture ignition, the starting instant of combustion (i.e. the sparking instant of the ignition spark) is certainly and accurately known and is chosen beforehand by the engine control strategy by means of the spark advance control.

Instead, in internal combustion engines with spontaneous mixture ignition (i.e. in internal combustion engines operating according to the “Diesel” cycle and fed with oil or similar) the MFB50 combustion index is not used as control magnitude due to the considerable difficulties in estimating such MFB50 combustion index efficiently (i.e. with sufficient accuracy), effectively (i.e. rapidly and without employing excessive calculation power) and cost-effectively (i.e. without requiring the installation of additional components with respect to those normally present).

Currently, pressure sensors capable of directly measuring the pressure inside the cylinders are used to determine the MFB50 combustion index; however, such pressure sensors are very expensive and present a limited reliability in time and are thus only suitable for laboratory tests and not for an extensive use on standard production cars.

DISCLOSURE OF INVENTION

It is the object of the present invention to provide a method for effectively, efficiently and cost-effectively estimating the crank angle at which 50% of the fuel mass has been burnt in a cylinder of an internal combustion engine with spontaneous mixture ignition.

According to the present invention, there is provided a method for estimating the crank angle at which 50% of the fuel mass has been burnt in a cylinder of an internal combustion engine with spontaneous mixture ignition as claimed in the attached claims.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be described with reference to the accompanying drawings which illustrate a non-limitative embodiment thereof, in which:

FIG. 1 is a diagrammatic view of an internal combustion engine provided with a control unit which implements the estimating method objective of the present invention, and

FIG. 2 is a block chart which illustrates the operation of the control unit in FIG. 1 for estimating the MFB50 combustion index.

PREFERRED EMBODIMENTS OF THE INVENTION

In FIG. 1, numeral 1 indicates as a whole an internal combustion engine fed with oil (i.e. operating according to the Diesel cycle which contemplates the spontaneous ignition of the mixture) and mounted aboard a road vehicle (not shown) which is provided with a transmission line (not shown) for transmitting the torque generated by the internal combustion engine 1 to the ground. The internal combustion engine 1 comprises four cylinders 2, each of which accommodates a corresponding piston 3 mechanically connected by means of a connecting rod to a drive shaft 4 for transmitting the force generated by the combustion inside the cylinder 2 to the drive shaft 4 itself.

A phonic wheel 5 provided with a number N (e.g. 60) of teeth 6 is keyed onto the drive shaft 4 and is coupled to a sensor 7, which is adapted to detect the time elapsed between the passage of two consecutive teeth 6. The teeth 6 of the phonic wheel 5 are equally spaced out except for one pair of teeth 6 which are reciprocally arranged at a greater distance than the others to constitute a singularity which allows to accurately identify each tooth 6 and thus the angular position of the phonic wheel 5 (i.e. of the drive shaft 4). The engine 1 further comprises a control unit 8 which is connected to the sensor 7 and is adapted to estimate, for each cylinder 2, the corresponding MFB50 (50% Mass Fraction Burnt) combustion index which represents the crank angle at which 50% of the fuel mass has been burnt inside the cylinder 2.

With reference to FIG. 2, there is described the mode used by the control unit 8 to estimate the MFB50 combustion index in each cylinder 2 by using the information provided by the sensor 7 coupled to the phonic wheel 5.

As previously mentioned, the phonic wheel 5 is provided with a number N (e.g. 60) of teeth 6, which are equally spaced out except for one pair of teeth 6 which are reciprocally arranged at a greater distance than the others to constitute a singularity which accurately allows to identify each tooth 6; in this manner, each tooth 6 is identified by a characteristic sequential number i included between 1 and N.

The sensor 7 detects the number i of the current tooth 6 (i^th tooth), the time T_i of the current tooth 6, i.e. the interval of time elapsing between the detections of the two angular events which characterize the current tooth 6, the raw angular speed ω_raw,i of the current tooth 6, and the drive angle α (i.e. the angular position of the drive shaft 4 which is comprised between 0 and 4π radians). The raw angular speed ω_raw,i of the i^th tooth 6 is supplied by the following equation:ω_i,raw=Δα_i/T_i

• ω_i,raw raw angular speed of the i^th tooth 6 [rad/s];
• Δα_i angular amplitude of the i^th tooth 6 [rad];
• T_i time of the i^th tooth 6 [s].

The magnitudes detected by sensor 7 and listed above are transmitted to a compensation block 9, which corrects each raw angular speed ω_raw,i to compensate for the errors deriving from the inevitable manufacturing tolerances of the teeth 6. In principle, all teeth 6 have the same angular amplitude, but actually due to the inevitable manufacturing tolerances each tooth 6 presents a characteristic shape error, i.e. presents a characteristic angular amplitude more or less different from the theoretical angular amplitude; consequently, the angular speeds ω_raw,i are afflicted by an intense noise which makes their use without an appropriate compensation practically impossible. In order to compensate for the shape errors of teeth 6, there is stored a vector (δμ₁, δμ₂, . . . , δμ_i, . . . , δμ_N) composed of N compensation values δμ_i, each of which is associated to a corresponding i^th tooth 6 and represents the difference between the actual angular amplitude of the i^th tooth 6 and the theoretic amplitude of the i^th tooth 6.

The corresponding compensated angular speed ω_i is calculated for each tooth 6 in the compensation block 9 by using the following equation:ω_i=(Δα_i+δμ_i)/T_i

• ω_i compensated angular speed of the i^th tooth 6 [rad/s];
• Δα_i angular amplitude of the i^th tooth 6 [rad];
• δμ_i compensation value δμ_i of the i^th tooth 6 [rad];
• T_i time of the i^th tooth 6 [s].

The calibration mode used to determine the compensation values δμ_i of the teeth 6 is described below. The calibration mode contemplates calculating the raw angular speed ω_raw,i of the i^th tooth 6 by using the above-described equation and calculating the average angular speed ω_av,i per revolution of the i^th tooth 6 by means of a moving average filter on N values of the raw angular speeds ω_raw,i sequence; it is important to observe that the moving average filter on N values of the raw angular speeds ω_raw,i sequence introduces a phase delay equal to N/2 tooth intervals to the average angular speeds ω_av,i. Subsequently, the delayed speed error Δω_i-N/2 is calculated for the i^th tooth 6 by using the following equation:Δω_i-N/2=Cω_av,i−ω_raw,i-N/2

• Δω_i-N/2 delayed speed error of the i−N/2^th tooth 6 [rad/s];
• ω_av,i average angular speed per revolution of i^th tooth 6 [rad/s] presenting a phase delay equal to N/2 tooth intervals;
• ω_raw,i-N/2 raw angular speed ω_raw,i of the i−N/2^th tooth 6 [rad/s].

The raw angular speeds ω_raw,i sequence is intentionally delayed by a phase delay equal to N/2 tooth intervals to equal the phase delay introduced by the moving average filter.

Subsequently, the delayed angular error δα_i-N/2 for the i^th tooth 6 is determined by using the following equation:δα_i-N/2=Δω_i-N/2*T_i-N/2

• δα_i-N/2 delayed angular error of the i−N/2^th tooth 6 [rad];
• Δω_i-N/2 delayed angular speed of the i−N/2^th tooth 6 [rad/s];
• T_i-N/2 time of the i−N/2^th tooth 6 [s].

Finally, the delayed angular errors vector δα_i-N/2 may be used either to determine (in the case of first absolute calibration operation) or to update the compensation values vector δμ_i. According to a preferred embodiment, before actually using a delayed angular errors vector δα_i-N/2, it is checked that each delayed angular error δα_i-N/2 of the vector itself is lower as absolute value than an acceptability threshold which indicates the maximum possible tolerance to the shape of the teeth 6; only if each delayed angular error δα_i-N/2 of the vector is lower as absolute value than the acceptability threshold may the delayed angular errors vector δα_i-N/2 actually be used. If a compensation value vector δμ_i is updated, it is preferably used a weighed average between the previous compensation values δμ_i and the corresponding delayed angular errors δα_i-N/2 which normally assigns a higher weight to the previous compensation values δμ_i (e.g. higher the value as higher is the confidence in the previous compensation values δμ_i). For example, the following equation could be used:δμ_i,j=K*(δα_i−δμ_i,j-1)+δμ_i,j-1

• δμ_i,j updated i^th compensation value [rad];
• k non-dimensional weight included between 0 and 1;
• δα_i i^th angular error [rad];
• δμ_i,j-1 previous i−1^th compensation value [rad].

Preferably, the above-described calibration operation for determining/updating the compensation values vector δμi is performed only in presence of certain circumstances, e.g. when the internal combustion engine 1 is in cut-off phase and the revolution speed of the drive shaft 4 is included in a certain calibration range so that the reciprocating, indicated inertial torques approximately are balanced with the internal combustion engine 1 in cut-off phase.

The compensated angular speeds ω_i calculated by the compensation block 9 are supplied to an analysis block 10, in which a frequency analysis of the known type is performed on the compensated angular speeds ω_i to determine at least one speed signal harmonic Ω_n. The harmonic generic Ω_n of the n^th speed signal is always a complex number, characterized by its own module |Ω_n| and its own phase Arg(Ω_n).

The harmonics Ω_n of the speed signal calculated by the analysis block 10 are supplied to a calculation block 11, in which there is implemented an algorithm based on an inverse mechanical model of the transmission which is symbolically indicated by the equation H(j{acute over (ω)})=T/Ω and represents, in the {acute over (ω)} frequency domain, the relation between the Fourier transforms of the angular speed Ω and of the torque T of the internal combustion engine 1. The function H(j{acute over (ω)}) is a complex value function and allows to obtain the n^th torque harmonic T_n by means of the equation:T_n=H(jΩ_n)*Ω_n

• T_n n^th torque harmonic;
• Ω_n n^th speed signal harmonic.

As the speed signal harmonic Ω_n, the torque harmonic T_n is a complex number characterized by its own module |T_n| and by its own phase Arg(T_n). It is important to observe that in order to be able to obtain precise estimations, the inverse mechanical model of the transmission must consider the inertial torques caused by the masses in reciprocating motion (e.g. the masses of the pistons 3).

The module |T_n| of the torque harmonics T_n calculated by the calculation block 11 is supplied to a calculation block 12, in which there is implemented an algebraic function Γ which puts the average indicated torque T_ind into relation with the module |T_n| of the n^th torque harmonic T_n according to the following equation:T_ind=Γ(|T_n|,P)

• T_ind average indicated torque T_ind;
• |T_n| n^th torque harmonic module T_n;
• P set of engineering parameters of the internal combustion engine 1.

Normally, the engineering parameters P of the internal combustion engine 1 comprise the revolution speed of the drive shaft 4 and the intake pressure (i.e. the pressure present inside an intake pipe).

The phase Arg(T_n) of the torque harmonics T_n calculated by the calculation block 11 is supplied to a calculation block 13, in which there is implemented an algebraic function Ψ which puts the MFB50 combustion index into relation with the phase Arg(T_n) of the n^th torque harmonic T_n according to the following equation:MFB50=Ψ(Arg(T_n),P

• MFB50 MFB50 combustion index;
• Arg (T_n) phase of the n^th torque harmonic T_n,
• P set of engineering parameters of the internal combustion engine 1.

By way of non-binding example, the equations of the direct mechanical model of the transmission in the frequency domain may be the following:

Torque Engine Side

$T_{e,n}=j{\tilde{\omega }}_e·\frac{J_1{J}_2{\tilde{\omega }}_e^2-\left(J_1+J_2\right)\left(j{\tilde{\omega }}_e{\rho }_{\mathrm{eb}}+K_{\mathrm{eb}}\right)}{J_2{\tilde{\omega }}_e^2-j{\tilde{\omega }}_e{\rho }_{\mathrm{eb}}-K_{\mathrm{eb}}}·{\omega }_{e,n}$ ${\tilde{\omega }}_e=\frac{n}{2}{\omega }_e$

Inertial Torque

$T_{\mathrm{alt},n}=\frac{m_{\mathrm{eq}}·r^2}{2}·{\omega }_e^2·e^{{\mathrm{j\theta }}_{\mathrm{alt}}}$

Cylinder TorqueT_ind,n=T_e,n−T_alt,n

• J₁ moment of inertia of the internal combustion engine 1;
• J₂ moment of inertia of the vehicle referred to the drive shaft 4;
• K_eb stiffness of the transmission line;
• ρ_eb damping coefficient of the transmission line;
• j imaginary unit;
• n harmonic order;
• {acute over (ω)}_e pulse;
• ω_e angular speed of the drive shaft 4; T_e torque generated by the internal combustion engine 1;
• T_alt,n torque due to the reciprocating motion masses;
• m_eq reciprocating motion masses;
• θ_alt inertial torque phase factor (reciprocating motion);
• T_ind average indicated torque.

The inverse mechanical model of the transmission (previously symbolically indicated by the equation H(j{acute over (ω)})=T/Ω) is obtained from the three equations of the direct mechanical model of the transmission shown above by replacing the first and the second equation in the third equation and so as to put the average indicated torque T_ind,n into relation with the angular speed ω_e,n of the drive shaft 4; the relation between the average indicated torque T_ind,n and the angular speed ω_e,n of the drive shaft 4 supplies the inverse mechanical model of the transmission (the final formula of the inverse mechanical model of the transmission is not shown because it does not add information and is complex to annotate). It is important to observe that in practice an experimental methodology is performed to experimentally obtain by points the transfer function of the inverse mechanical model of the transmission.

By way of non-binding example, the algebraic function Ψ which puts the combustion index MFB50 into relation with the phase Arg(T_n) of the harmonic T_n of the n^th torque could be as follows:MFB50=c₀+c₁Φ+c₂Φ²+c₃PΦ+c₄P²Φ

• MFB50 combustion index;
• Φ phase of the n^th torque harmonic T_n;
• P intake pressure;
• c_i coefficients.

The values of the coefficients c_i may be obtained experimentally by minimising the mean square error between a series of measured values of the combustion index MFB50 and a corresponding series of estimated values of the combustion index MFB50.

By way of non-binding example, the algebraic function Γ which puts the average indicated torque T_ind into relation with the module |T_n| of the harmonic T_n of the n^th torque could be as follows:T_ind=b₀+b₁|Tn|+b₂P+b₃N

• T_ind average indicated torque T_ind;
• |T_n| n^th torque harmonic module T_n;
• P intake pressure;
• N revolution speed of the drive shaft 4;
• b_i coefficients.

The values of the coefficients b_i may be obtained experimentally by minimising the mean square error between a series of measured values of the average indicated torque T_ind and a corresponding series of estimated values of the average indicated torque T_ind.

The above-described estimation mode of the combustion index MFB50 in each cylinder 2 presents many advantages, because it is efficient (i.e. it allows to estimate the combustion index MFB50 with a considerable accuracy), effective (i.e. it allows to estimate the combustion index MFB50 rapidly and without employing excessive calculation power), and cost-effective (i.e. it does not require the installation of additional components with respect to those normally present in a modern internal combustion engine operating according to the “Diesel” cycle).

In virtue of the possibility of estimating the combustion index MFB50 in each cylinder 2, it is possible to feedback control the combustion process in virtue of which it is possible to reduce the fuel consumption and the polluting emissions, the performances being equal. For example, the combustion index MFB50 in each cylinder 2 may be advantageously used to govern the force on the connecting rods and knock phenomena.

Furthermore, the above-described estimation mode of the combustion index MFB50 in each cylinder 2 allows to obtain the average indicated torque T_ind produced by each cylinder 2 which may be effectively used to further improve the combustion process control.

Claims

1. A method for determining the angular speed (ωi) of a drive shaft (4) of an internal combustion engine (1) at each tooth event of a phonic wheel (5) presenting a number N of teeth (6); the method comprises the steps of: detecting the angular amplitude (Δαi) for each tooth 6;detecting the time (Ti) of each tooth (6); anddetermining the raw angular speed (ωraw,i) of each tooth (6) according to the corresponding angular amplitude (Δαi) and time (Ti);the method is characterized in that it comprises the further steps of:determining for each tooth (6) a compensation value (δμi) which represents the difference between the actual angular amplitude of the tooth (6) and the theoretic angular amplitude of the tooth (6); anddetermining the angular speed (ωi) of each tooth (6) by correcting the raw angular speed (ωraw,i) by means of the corresponding compensation value (δμi).

2. A method according to claim 1, wherein the angular speed (ωi) of each tooth (6) is determined according to the following equation: ωi=(Δαi+δμi)/Ti ωi compensated angular speed of the ith tooth (6) [rad/s];Δαi angular amplitude of the ith tooth (6) [rad];δμi compensation value δμi of the ith tooth (6) [rad];Ti time of the ith tooth (6) [s].

3. A method according to claim 1, wherein the step of determining the compensation value (δμi) for each tooth (6) comprises the further steps of: calculating the raw angular speed (ωraw,i) of the tooth (6);calculating an average angular speed (ωav,i) per revolution of the tooth (6) by means of a moving average filter on N values of the raw angular speeds sequence (ωraw,i) ;calculating a delayed speed error (Δωi-N/2) of the tooth (6) as difference between the average angular speed (ωav,i) per revolution and the delayed raw angular speed (ωraw,i) of one phase delay equal to N/2 tooth intervals;determining a delayed angular error (δαi-N/2) of the tooth (6) by multiplying the delayed speed error (Δωi-N/2) of the tooth (6) by the time (Ti) of the tooth (6) delayed by a phase delay equal to N/2 tooth intervals; andusing the delayed angular error (δαi-N/2) for either determining or updating the compensation value (δμi) of the tooth (6).

4. A method according to claim 3, wherein the delayed speed error (Δωi-N/2) of the tooth (6) is calculated by using the following equation: Δωi-N/2=ωav,i−ωraw,i-N/2 Δωi-N/2 delayed speed error of the i−N/2th tooth (6) [rad/s];ωav,i average angular speed per revolution of ith tooth (6) [rad/s] presenting a phase delay equal to N/2 tooth intervals;ωraw,i-N/2 raw angular speed ωraw,i of the i−N/2th tooth (6) [rad/s].

5. A method according to claim 3, wherein the delayed angular error (δαi-N/2) of the tooth (6) is calculated by using the following equation: δαi-N/2=Δωi-N/(2)*Ti-N/(2) δαi-N/2 delayed angular error of the i−N/2th tooth (6) [rad];Δωi-N/2 delayed speed error of the i−N/2th tooth (6) [rad/s];Ti-N/(2) time of the i−N/2th tooth 6 [s].

6. A method according to claim 3, wherein a vector of N delayed angular errors (δαi-N/2) is used to either determine or update a compensation values vector (δμi) only if each delayed angular error (δαi-N/2) of the vector itself is lower as absolute value than an acceptability threshold which indicates the maximum possible tolerance on the shape of the tooth (6).

7. A method according to claim 1, wherein the compensation values (δμi) of the teeth (6) are either determined or updated only when the internal combustion engine (1) is in cut-off phase and the revolution speed of the drive shaft (4) is comprised in a certain calibration range so that the reciprocating, indicated inertial torques are approximately balanced with the internal combustion engine (1) in cut-off phase.