METHOD OF CONTROLLING A COMBUSTION ENGINE FROM ESTIMATION OF THE BURNT GAS MASS FRACTION IN THE INTAKE MANIFOLD

Abstract

In a method of controlling the combustion of a combustion engine from an estimation of the burnt gas mass fraction in the intake manifold, a measurement relative to a fresh air flow rate or to a burnt gas flow rate is performed upstream from the mixing space where fresh air and burnt gas mix. The burnt gas mass fraction present in the mixing space is then estimated from the measurement and from a model of the mixing dynamics in this space. A transport delay between the space and the intake manifold is estimated. The mass fraction of burnt gas in the intake manifold is then deduced in real time. Finally, combustion is controlled from the burnt gas mass fraction in the intake manifold.

Description

FIELD OF THE INVENTION

The present invention relates to the sphere of engine control and more particularly to the estimation of the burnt gas recirculation rate for a gasoline engine provided with an EGR circuit.

BACKGROUND OF THE INVENTION

Downsizing gasoline engines currently appears to be the preferred option for reducing the consumption of gasoline engines. In fact, this technology allows to shift the working points of the engine to zones of higher efficiency and thus to limit the pumping losses inherent in the operation of an internal-combustion engine. This type of engine then requires the presence of a compressor driven by a turbine arranged in the exhaust line. Such a device is used to improve air filling of the cylinder and to provide a torque equivalent to that of an engine of conventional displacement. It is thus possible to have the same performances while drastically reducing the consumption.

However, using such a technology greatly increases the risk of engine knock appearance. When the engine runs under full load conditions, the thermodynamic conditions in the combustion chamber may be detrimental to the stability of the mixture and generate auto-ignition thereof. This phenomenon can eventually greatly deteriorate the combustion chamber.

To solve this problem, the ignition advance is usually degraded. This option generates an increase in the gas temperature at the end of the combustion cycle and therefore over the entire exhaust line. Thus, in order to compensate for this phenomenon, the mixture is richened at the intake.

Such a method involves two drawbacks: first, it increases the engine consumption; furthermore, it deteriorates the efficiency of the catalyst arranged downstream from the exhaust manifold, which provides optimal conversion of the pollutants resulting from the combustion when the mixture is in stoichiometric proportions.

In this context, burnt gas recirculation (EGR) from the exhaust to the intake is a promising option. Indeed, feeding burnt gas that has not reacted during combustion into a cylinder of the engine allows to decrease the overall combustion temperature and to limit engine knock appearance. The advantages of downsizing in terms of efficiency and consumption are thus preserved. Besides, introduction of burnt gas also allows to reduce the temperature of the exhaust gas and therefore to limit the impact thereof on the catalyst or the turbine.

However, such a strategy has a great influence on the overall running conditions of the engine. For example, the air mass trapped in the cylinder is smaller in an EGR configuration since burnt gas takes the place of fresh air in the cylinder. To operate under stoichiometric conditions, it is necessary to adjust the fuel loop to the air loop, and thus to have a precise estimation of the composition of the gas in the cylinder. Such a reasoning also applies to the strategies for controlling variable valve actuators (VVT type) or to the ignition advance strategy. Besides, this estimation is essential for torque transient management, notably at low loads where too high a proportion of burnt gas can extinguish the combustion.

The presence of a gas composition estimator in the intake manifold is therefore necessary to allow proper combustion control (ignition advance control, gas composition control for example), in particular in transient states.

The composition of the mixture at the intake is thus an essential datum. The method according to the invention allows to estimate in real time such a composition, notably the burnt gas mass fraction present in the mixture in the intake manifold, then to control the combustion through control of the ignition advance or of this fraction by means of EGR valve (6).

SUMMARY OF THE INVENTION

In general terms, the invention relates to a method of controlling the combustion of a combustion engine (1) comprising at least one cylinder (2), an intake circuit including an intake manifold (3) and a burnt gas recirculation circuit including an EGR valve (6). The method comprises the following stages:

• • performing a measurement relative to a gas, fresh air or burnt gas flow rate, upstream from a mixing space V_bp wherein fresh air and burnt gas mix,
  • estimating a burnt gas mass fraction BGR_bp present in said mixing space, from said measurement and a model of the mixing dynamics in said space,
  • estimating a transport delay from said space to said intake manifold,
  • estimating in real time a burnt gas mass fraction BGR(t) in the intake manifold from said estimation of the burnt gas mass fraction present in the mixing space and said transport delay,
  • controlling the combustion from said burnt gas mass fraction BGR(t) in the intake manifold.

According to the invention, the model of the mixing dynamics in the mixing space can be a function of an atmospheric pressure, of a temperature upstream from the EGR valve, of said space V_bp, of a mass flow rate D_gb(t) of burnt gas fed through the EGR valve and of a fresh air mass flow rate D_air(t) at the intake circuit inlet.

According to the invention, the pure delay τ can be calculated by solving a laminar flow law iteratively over intake circuit zones defined in such a way that thermodynamic conditions in each zone are homogeneous. Three zones can therefore be defined as follows: a first zone where the thermodynamic conditions correspond to the atmospheric conditions; a second zone where pressure and temperature increase during gas transport, and a third zone where the temperature decreases, which consists of the remaining zone of the intake circuit.

According to one embodiment, the measurement is a measurement of the mass flow rate of fresh air at the intake line inlet. According to another embodiment, the measurement is a measurement of a pressure difference (P at the EGR valve, and the burnt gas mass fraction BGRbp present in said space is estimated from pressure difference (P and from a punctual pressure drop (i.e., pressure drop at a point of the system) relation applied at said EGR valve.

According to an embodiment, combustion control is achieved by controlling an ignition advance, by adding a proportional term to the burnt gas mass fraction BGR(t) in the intake manifold.

According to another embodiment, combustion control is achieved by controlling an air mass contained in the cylinder, by controlling said EGR valve (6) so as to minimize the difference between BGR(t) and a burnt gas fraction set point in said intake manifold (3) BGR^sp.

The invention also relates to a combustion engine (1) comprising at least one cylinder (2), an intake manifold (3) and a burnt gas recirculation circuit comprising an EGR valve (6), said combustion engine being provided with a pressure detector (5) at the level of said EGR valve (6) or with an air flow meter (5′) at the start of the intake line. The engine comprises control means suited for applying the control method according to the invention.

BRIEF DESCRIPTION OF THE FIGURES

Other features and advantages of the method according to the invention will be clear from reading the description hereafter of non limitative embodiment examples, with reference to the accompanying figures wherein:

FIG. 1 illustrates a supercharged gasoline engine comprising an EGR circuit,

FIG. 2 illustrates a result of the estimation of the burnt gas mass fraction present in the mixture at the level of the intake manifold, using the method according to the invention, and

FIG. 3 shows a process flow diagram of the method according to the invention.

DETAILED DESCRIPTION

FIG. 1 shows a gasoline engine (1) provided with an EGR burnt gas recirculation circuit. At least one cylinder (2) of combustion engine (1) is supplied with air and burnt gas from an intake manifold (3). The air intake circuit is provided with a cooler (4) and a turbocharger (7) compressor. The exhaust line consists of an exhaust manifold (13), a turbocharger (7) turbine, a bypass line, for injecting part of the burnt gas into the air intake circuit. This part of the circuit is notably provided with a cooler (4′) and a controlled valve, referred to as EGR valve (6), which controls the amount of burnt gas injected into the air intake circuit. This engine is in particular equipped with either a pressure difference detector (5) at the level of the EGR valve or an air flow meter (5′) at the start of the intake line. Engine (1) as shown in FIG. 1 is also provided with a direct injection device and a variable valve timing system, these elements being usually present in a downsized engine (reduced engine displacement) but whose presence is not taken into consideration in the method according to the invention.

The burnt gas recirculation circuit considered withdraws burnt gas at the engine exhaust downstream from the turbine of turbocharger (7) and catalyst (11), and reinjects it into the intake of a cylinder (2) upstream from the compressor of turbocharger (7). The amount of burnt gas reinjected into the intake line is controlled by controlled EGR valve (6) arranged downstream from the EGR burnt gas recirculation circuit.

Since a gasoline engine runs under optimum conditions when the proportion of air/gasoline mixture allows to provide complete combustion of the fuel without excess air, one generally considers that the exhaust gas fully consists of burnt gas. Thus, the EGR burnt gas recirculation circuit is filled with burnt gas only.

Notations

In the description, the terms upstream and downstream are defined with respect to the direction of flow of the fluids in air loop (10). Furthermore, the main notations used are as follows:

• • t: time
  • x: points in the intake circuit.

Engine Parameters

• • P_atm,T_atm: atmospheric pressure and temperature. They can be considered to be constant.
  • T_am: upstream temperature at the EGR valve inlet. This temperature is imposed by the passage through exchanger (4′) arranged in the burnt gas recirculation circuit.
  • V_bp: the fresh air and burnt gas mixing space, downstream from the EGR valve. It is the volume occupied by the lines at the intersection of the fresh air delivery lines and the burnt gas delivery lines, this space extending up to the compressor of turbocharger (7). This space corresponds to the hatched zone (12) of FIG. 1.
  • V_adm: volume of the intake manifold.
  • P_adm,T_adm: pressure and temperature in the intake manifold. Conventionally, the intake temperature is considered to be constant. Indeed, the exchanger arranged upstream from the manifold is so dimensioned as to provide such a regulation.
  • Φ_adm, Φ_ech: position of the intake (8) and exhaust (9) valve actuators. These variables quantify a phase difference with respect to a reference position.
  • N_e: engine speed.

Variables of the Air Loop Model

• • BGR: burnt gas mass fraction in the air mixture (fresh air and burnt gas) at the level of intake manifold (3). It conditions the mass of burnt gas present in the cylinder upon closure of intake valve (8).
  • BGR_bp: burnt gas mass fraction in the air mixture (fresh air and burnt gas) at the level of the volume downstream from EGR valve (12).
  • D_gb(t): mass flow rate of burnt gas fed through EGR valve (6).
  • D_air(t): fresh air mass flow rate at the intake line inlet. This quantity can, depending on the instrumentation, be measured by an air flow meter (5′).
  • D_asp: mass flow rate of cylinder filling with air mixture (fresh air and burnt gas)
  • D_thr: air mixture (fresh air and burnt gas) mass flow rate upstream from the compressor of turbocharger (7).
  • τ: gas transport delay between the time of fresh air and burnt gas mixing, and the delivery in the intake manifold.
  • S: effective surface area of the EGR valve. This quantity characterizes the surface area of fluid that can flow through the valve, it is linked with the opening of the valve via a characteristic map of the valve considered.
  • P_am: upstream pressure at the EGR valve inlet.
  • P_av: downstream pressure at the EGR valve outlet.
  • ΔP: pressure difference between upstream and downstream from the EGR valve: ΔP=P_am−P_av. This quantity can, depending on the instrumentation, be measured with instrumentation (5) of the EGR valve.
  • P,T: current pressure and temperature, i.e. at a point x and at a time t.
  • ν: punctual velocity (i.e., velocity a point of the system) of the air mixture (fresh air and burnt gas) in the intake circuit.
  • L_cana: length of the intake circuit line considered.
  • A: effective cross section of the intake circuit line considered.
  • ρ: density of the gas mixture.

Constants of the Air Loop Model

• • r: specific ideal gas constant, which is the same for all the gases concerned here (exhaust gas and air), and has the value 288 J/kg/K.
  • γ: gas specific heat ratio. The gases being assumed to be ideal, this ratio is an identical constant for all the gases concerned, and it has the value 1.4.

These notations, with index --^sp, represent the set point values associated with the quantities considered.

The method according to the invention allows to estimate the burnt gas mass fraction present in the mixture at the level of the intake manifold, then to control this fraction by controlling EGR valve (6) of the exhaust gas recirculation circuit. It is based on the use of either a pressure difference detector at the level of the EGR valve or a flow meter at the start of the intake line. FIG. 3 shows a process flow diagram of the method according to the invention that comprises the following stages:

1. Acquisition of a measurement relative to a gas mass flow rate upstream from the mixing zone in the intake circuit.

2. Real-time estimation of the burnt gas mass fraction present in the mixture at the level of the intake manifold:

i. estimation of the burnt gas mass fraction present in the gas mixing space

ii. estimation of the transport delay from this space to the intake manifold

3. Combustion control from the estimated BGR.

1—Acquisition of a Measurement Relative to the Gas Mass Flow Rate in the Intake Circuit

The first stage consists in performing a measurement relative to a gas mass flow rate (fresh air or burnt gas) upstream from the (fresh air and burnt gas) mixing zone in the intake circuit. The mixing space corresponds to a volume downstream from the EGR valve. It is the volume of the lines at the intersection of the fresh air delivery lines and the burnt gas delivery lines, this volume extending up to the compressor of turbocharger (7). This volume corresponds to hatched zone (12) in FIG. 1.

It can be a value relative to the burnt gas flow rate, or a value relative to the fresh air mass flow rate at the intake line inlet.

According to one embodiment, the fresh air mass flow rate at the intake line inlet is measured using an air flow meter (5′) arranged at the start of the intake line.

According to another embodiment (FIG. 3), the difference between the pressure downstream from the EGR valve and the pressure upstream from the EGR valve is measured using a pressure difference detector (5) at the level of the EGR valve.

This value is correlated with the phenomenon occurring at the level of the valve, notably at the level of the mixing space, where mixing of fresh air and of burnt gas occurs.

To obtain an estimation of the burnt gas mass fraction at the level of the intake manifold, the burnt gas mass fraction at the level of the mixing space is estimated, from this pressure difference measurement or from the measurement given by the air flow meter, then a delay due to the distance traveled from the mixing space to the intake manifold is applied.

2—Real-time estimation of the burnt gas mass fraction present in the mixture at the level of the intake manifold

The burnt gas recirculation circuit (EGR circuit) withdraws burnt gas at the exhaust downstream from the turbine and reinjects it into the intake upstream from the compressor. The amount of burnt gas reinjected into the intake line is controlled by a valve arranged at the junction between the intake line and the EGR recirculation circuit, thus downstream from the EGR circuit. This actuator (EGR valve) being spatially distant from the intake manifold, a subsequent variable transport delay is induced on the intake line between the burnt gas mass fraction at the level of the mixing space, downstream from the EGR valve, and that at the level of the intake manifold. The mixing space is the volume of the lines at the intersection of the fresh air delivery lines and the burnt gas delivery lines, this volume extending up to the compressor of turbocharger (7). This volume corresponds to hatched zone (12) of FIG. 1.

Since a gasoline engine runs under optimum conditions when the proportion of air/gasoline mixture allows to provide complete combustion of the fuel without excess air, one generally considers that the exhaust gas fully consists of burnt gas. The EGR circuit thus is filled only with burnt gas. It is then possible to model the dynamics of the reinjected burnt gas in two stages:

• • a mixing dynamics in the mixing space:

$\begin{array}{cc}\frac{}{t}{\mathrm{BGR}}_{\mathrm{bp}}\left(t\right)=\frac{{\mathrm{rT}}_{\mathrm{am}}}{P_{\mathrm{atm}}V_{\mathrm{bp}}}\left(-\left[D_{\mathrm{gb}}\left(t\right)+D_{\mathrm{air}}\left(t\right)\right]{\mathrm{BGR}}_{\mathrm{bp}}+D_{\mathrm{gb}}\left(t\right)\right)& \left(\mathrm{E1}\right)\end{array}$

• • a dynamics of transport from this space to the intake manifold, represented by a transport over time:

BGR(t)=BGR_bp(t−τ(t))  (E2)

Thus, according to the invention, real-time estimation of the burnt gas mass fraction BGR(t) present in the mixture at the level of the intake manifold is achieved by estimating the burnt gas mass fraction present in the mixing space. The transport delay from this space to the intake manifold is then estimated by segmenting the intake line according to the thermodynamic conditions.

i. Estimation of the Burnt Gas Mass Fraction Present in the Mixing Space

In order to provide an estimation of the burnt gas composition in the intake manifold, we first estimate the burnt gas composition in the mixing space. We therefore want to solve Equation 1:

$\frac{}{t}{\mathrm{BGR}}_{\mathrm{bp}}\left(t\right)=\frac{{\mathrm{rT}}_{\mathrm{am}}}{P_{\mathrm{atm}}V_{\mathrm{bp}}}\left(-\left[D_{\mathrm{gb}}\left(t\right)+D_{\mathrm{air}}\left(t\right)\right]{\mathrm{BGR}}_{\mathrm{bp}}+D_{\mathrm{gb}}\left(t\right)\right).$

To achieve this, we know the following parameters:

• • r: specific ideal gas constant, which is the same for all the gases concerned here (exhaust gas and air), and has the value 288 J/kg/K.
  • T_am: upstream temperature at the EGR valve inlet. This temperature is imposed by the passage through exchanger (4′) arranged in the burnt gas recirculation circuit.
  • P_atm: atmospheric pressure. It can be considered to be a first-order constant.
  • V_bp: mixing space. It is a manufacturer's datum.

And the following parameters have to be estimated to solve Equation (E1):

• • D_air(t): fresh air mass flow rate at the intake line inlet.
  • D_gb(t): mass flow rate of burnt gas fed through the EGR valve.

To estimate D_gb+D_air, the total mass flow rate of gas (fresh air and burnt gas) leaving the mixing space, we can introduce D_thr, the gas (fresh air and burnt gas) mass flow rate upstream from the compressor of turbocharger (7), and write D_thr=D_gb+D_air=real total flow rate. This flow rate D_thr can be estimated by a model constructed by combining a static cylinder filling model with a dynamic model of the punctual pressure drop occurring at the level of the fresh air butterfly valve (13 in FIG. 1). For example:

$\begin{array}{cc}{D}_{\mathrm{thr}}=D_{\mathrm{asp}}+\frac{r}{V_{\mathrm{adm}}T_{\mathrm{adm}}}\frac{P_{\mathrm{adm}}}{t}& \left(\mathrm{E6}\right)\end{array}$

where P_adm, T_adm and V_adm are known, and where D_asp is the mass flow rate of cylinder filling with gas (fresh air and burnt gas). We consider a static filling model estimating the burnt gas mass in the cylinder as a function of the engine speed, the temperature and the pressure in the intake manifold, and the actuator positions.

$\begin{array}{cc}\{\begin{array}{c}{M}_{\mathrm{asp}}^{\mathrm{int}}={\alpha }_1\frac{P_{\mathrm{adm}}}{{\mathrm{rT}}_{\mathrm{adù}}}V_{\mathrm{ivc}}-M_{\mathrm{asp}}^{\mathrm{ech}}\\ M_{\mathrm{asp}}^{\mathrm{ech}}={\alpha }_2\frac{\mathrm{OF}}{N_e}+{\alpha }_3V_{\mathrm{evc}}\end{array}& \left(\mathrm{E3}\right)\end{array}$

with:

• • M_asp^int: burnt gas mass in the cylinder.
  • M_asp^ech: burnt gas mass at the cylinder exhaust.
  • α₁, α₂, α₃: known maps, functions of P_adm and N_e (experimentally determined on the test bench).
  • V_ivc: volume of the cylinder upon intake valve closure (ivc). This volume is a function of the position of the intake valve actuator Φ_adm, and of the engine dimensions.
  • V_evc: volume of the cylinder upon exhaust valve closure (evc). This volume is a function of the position of the exhaust valve actuator Φ_ech, and of the engine dimensions.
  • OF: overlap factor. It is a function of the positions of the intake and exhaust valve actuators Φ_adm and Φ_ech. It can be determined by the relation:

OF=∫_θivo^θiv=θevA_intdθ+∫_θiv=θev^θevcA_echdθ

with:

• • A_adm and A_ech: intake and exhaust valve opening areas, they are engine parameters.
  • θ: crank angle.
  • θ_ivo: crank angle upon intake valve opening (ivo). This angle is a function of the position of the intake valve actuator Φ_adm.
  • θ_evc: crank angle upon exhaust valve closure (evc). This angle is a function of the position of the exhaust valve actuator Φ_ech.
  • θ_iv=θ_ev: crank angle where both valves have the same opening area.

From this system of equations (E3), we express the mass of burnt gas in the cylinder from engine parameters P_adm, N_e, Φ_adm and Φ_ech and from overlap factor OF.

By combining the equations of system (E3), we can define a function g between the burnt gas mass as a function of the three parameters P_adm, Φ_adm and Φ_ech:

$\begin{array}{cc}\begin{array}{c}{M}_{\mathrm{asp}}^{\mathrm{int}}=g\left(P_{\mathrm{adm}},{\Phi }_{\mathrm{adm}},{\Phi }_{\mathrm{ech}}\right)\\ ={\alpha }_1\left(P_{\mathrm{adm}},N_e\right)\frac{P_{\mathrm{adm}}}{{\mathrm{rT}}_{\mathrm{adm}}}V_{\mathrm{ivc}}\left({\Phi }_{\mathrm{adm}}\right)-{\alpha }_2\left(P_{\mathrm{adm}},N_e\right)\frac{\mathrm{OF}\left({\Phi }_{\mathrm{adù}},{\Phi }_{\mathrm{ech}}\right)}{N_e}-\\ {\alpha }_3\left(P_{\mathrm{adm}},N_e\right)V_{\mathrm{ech}}\left({\Phi }_{\mathrm{ech}}\right)\end{array}& \left({\mathrm{E3}}^{\prime }\right)\end{array}$

We then determine the mass flow rate of cylinder filling with air by means of relation (E4):

$\begin{array}{cc}{D}_{\mathrm{asp}}=\frac{\mathrm{Ne}}{30}M_{\mathrm{asp}}^{\mathrm{int}}& \left(\mathrm{E4}\right)\end{array}$

We then combine Equations (E3′) and (E4) and, for the sake of clarity, the engine speed and the intake temperature are disregarded in the expression. We can thus define a function f between the cylinder filling mass flow rate and the three parameters P_adm, Φ_adm and Φ_ech (E5):

$\begin{array}{cc}\begin{array}{c}{D}_{\mathrm{asp}}=\frac{N_e}{30}M_{\mathrm{asp}}^{\mathrm{int}}\\ =\frac{N_e}{30}g\left(P_{\mathrm{adm}},{\Phi }_{\mathrm{adm}},{\Phi }_{\mathrm{ech}}\right)\\ =f\left(P_{\mathrm{adm}},{\Phi }_{\mathrm{adm}},{\Phi }_{\mathrm{ech}}\right)\end{array}& \left(\mathrm{E5}\right)\end{array}$

The mass flow rate of the air mixture (fresh air and burnt gas) upstream from the compressor of turbocharger (7) can be deduced therefrom using a relation of the type as follows:

$\begin{array}{cc}{D}_{\mathrm{thr}}=D_{\mathrm{asp}}+\frac{r}{V_{\mathrm{adm}}T_{\mathrm{adm}}}\frac{P_{\mathrm{adm}}}{t}& \left(\mathrm{E6}\right)\end{array}$

To estimate D_gb(t), the burnt gas mass flow introduced through the EGR valve, the measurement of the difference between the pressure downstream from the EGR valve and the pressure upstream from the EGR valve (stage 1) can be used if the engine is provided with a dedicated detector:

D _gb=φ(ΔP)  (E7)

We apply a punctual load (i.e., load at a point of the system) relation at the level of the EGR valve, for example the Barré-Saint Venant relation. We can write a relation between the mass flow rate of burnt gas fed through the EGR valve, D_gb, the effective surface area S of the EGR valve, the temperature upstream from the valve, T_am, and the pressure difference at the level of the valve (Equation 11):

$D_{\mathrm{gb}}=\frac{{\mathrm{SP}}_{\mathrm{am}}}{\sqrt{{\mathrm{rT}}_{\mathrm{am}}}}\{\begin{array}{cc}{\left(\frac{P_{\mathrm{av}}}{P_{\mathrm{am}}}\right)}^{\frac{1}{\gamma }}\sqrt{\frac{2\gamma }{\gamma -1}\left(1-{\left(\frac{P_{\mathrm{av}}}{P_{\mathrm{am}}}\right)}^{\frac{\gamma -1}{\gamma }}\right)}& \mathrm{if}\frac{P_{\mathrm{av}}}{P_{\mathrm{am}}}>{\left(\frac{2}{\gamma +1}\right)}^{\frac{\gamma }{\gamma -1}}\\ \sqrt{{\gamma \left(\frac{2}{\gamma +1}\right)}^{\frac{\gamma +1}{\gamma -1}}}& \mathrm{otherwise}\end{array}$

This relation is obtained by applying the Barré-Saint Venant relation for a fluid flowing from a point 1 (upstream) to a point x (downstream), which allows to determine the velocity V of the fluid at point x:

$V\left(x\right)=\sqrt{\frac{2\gamma }{\gamma -1}{{\mathrm{rT}}_1\left(1-\frac{P\left(x\right)}{P_1}\right)}^{\frac{\gamma }{\gamma -1}}}$

• • P₁ being the pressure at point 1.

Then the velocity is related to the mass flow rate by the relation:

$V\left(x\right)=\frac{D_{\mathrm{gb}}}{\rho S}$

The formula as follows is then applied to obtain Equation 11:

$\rho =\frac{P_1}{\mathrm{rT}}$

The pressure downstream from the valve is considered to be the atmospheric pressure, and the relation below can therefore be expressed only by means of quantity ΔP=P_am−P_av and P_atm. In an advantageous embodiment of the invention, this model can be linearized for the small values of ΔP (for example for ΔP<10% P_atm).

Alternatively, to estimate D_gb(t), the burnt gas mass flow introduced through the EGR valve, it is possible to use an air flow meter (5′) that measures the fresh air mass flow rate at the intake line inlet, (D_air), then to use the flow rate model given in ((E3)-(E6)) to estimate the burnt gas flow rate, by writing D_gb=D_thr−D_air.

Using model ((E6) and (E5)) in the mixing dynamics (E1) in place of the real total flow rate D_thr=D_gb+D_air=real total flow rate, we construct an estimation of the amount of burnt gas present in the mixing space, by writing:

$\frac{}{t}{\mathrm{BGR}}_{\mathrm{bp}}\left(t\right)=\frac{{\mathrm{rT}}_{\mathrm{am}}}{P_{\mathrm{atm}}V_{\mathrm{bp}}}\left(-D_{\mathrm{thr}}{\mathrm{BGR}}_{\mathrm{bp}}+D_{\mathrm{gb}}\left(t\right)\right)$ $\mathrm{with}$ $D_{\mathrm{thr}}=g\left(D_{\mathrm{asp}},\frac{P_{\mathrm{adm}}}{t}\right)\mathrm{and}$ $D_{\mathrm{asp}}=f\left(P_{\mathrm{adm}},{\Phi }_{\mathrm{adm}},{\Phi }_{\mathrm{ech}}\right)\mathrm{and}$ $D_{\mathrm{gb}}=\varphi \left(\Delta P\right)$

or by writing:

$\frac{}{t}{\mathrm{BGR}}_{\mathrm{bp}}\left(t\right)=\frac{{\mathrm{rT}}_{\mathrm{am}}}{P_{\mathrm{atm}}V_{\mathrm{bp}}}\left(-D_{\mathrm{thr}}{\mathrm{BGR}}_{\mathrm{bp}}+\left[D_{\mathrm{thr}}-D_{\mathrm{air}}\left(t\right)\right]\right)$ $\mathrm{with}$ $D_{\mathrm{thr}}=g\left(D_{\mathrm{asp}},\frac{P_{\mathrm{adm}}}{t}\right)\mathrm{and}$ $D_{\mathrm{asp}}=f\left(P_{\mathrm{adm}},{\Phi }_{\mathrm{adm}},{\Phi }_{\mathrm{ech}}\right)$

according to the engine instrumentation.

We thus estimate the burnt gas mass fraction BGR_bp present in the mixing space. Estimation of D_thr is achieved from known parameters, and estimation of D_gb(t) is achieved either by means of the measurement of the difference between the pressure downstream from the EGR valve and the pressure upstream from the EGR valve (stage 1) and of the application of a relation of punctual load at the level of the EGR valve, the Barré-Saint Venant relation for example, or by means of the measurement performed by an air flow meter (stage 1) and of a mass balance using estimation D_thr.

The estimation obtained by this methodology can be initialized (or calibrated) when the EGR valve is closed and the proportion of burnt gas downstream from the valve is zero.

Ii. Estimation of the Transport Delay from this Space to the Intake Manifold

In order to obtain an estimation of the burnt gas mass fraction at the level of the intake manifold, from the burnt gas mass fraction at the level of the mixing space, a pure delay due to the distance traveled from the mixing space to the intake manifold is applied to the estimator of BGR_bp.

This gas transport delay between the time of fresh air and burnt gas mixing, and the delivery in the intake manifold is denoted by τ. This delay is a function of the distance traveled and of the velocity of the air mixture. The length of the line considered is denoted by L_cana and the punctual velocity of the air mixture by ν. The effective cross section of the line considered, in which the mixture circulates, is denoted by A. Finally, D represents the current mass flow rate.

Pure delay τ can be defined implicitly from a laminar flow law:

$\begin{array}{cc}{L}_{\mathrm{cana}}={\int }_{-\tau \left(t\right)}v\left(s\right)s& \left(\mathrm{E7}\right)\\ v\left(s\right)=\frac{D\left(s\right)\mathrm{rT}\left(s\right)}{\mathrm{AP}\left(s\right)}& \left(\mathrm{E8}\right)\end{array}$

where the thermodynamic conditions (pressure P and temperature T) are involved in the relation between velocity and mass flow rate (E8) and where s is an integration variable (mute variable).

In order to estimate this pure delay from Equations E7 and E8, the intake line is divided according to the prevailing thermodynamic conditions. The goal is that the thermodynamic conditions within each zone are stable (constant, or varying linearly). This segmentation therefore depends on the engine architecture.

Thus, according to an embodiment, the definitions of the zones with respect to the thermodynamic conditions lead to the segmentation as follows:

• • the first zone is located between the EGR valve and the compressor of turbocharger (7). The thermodynamic conditions of this zone correspond to the atmospheric conditions,
  • the second zone extends from the compressor of turbocharger (7) to intercooler (4). The pressure and the temperature in this zone are considered to be homogeneous. They are measured by a dedicated detector,
  • the last zone considered is located between intercooler (4) and air butterfly (13). The cooler being so dimensioned as to drastically decrease the temperature, the latter is considered to be homogeneous and equal to the temperature measured in the manifold. The pressure is also considered to be homogeneous and equal to that of the second zone, upstream from the cooler.

On each one of these segments, the flow relation of (E7)-(E8) type is solved by iteration from the start of the intake line (EGR valve). This solution can be performed by discretization of the integral and direct numerical computation. This relation uses thermodynamic data that are obtained either through direct measurement or via models, according to the technical definition of the engine detectors. Besides, the value of mass flow rate D is approximated by model (E3).

Once delay T determined, relation (E2) is applied so as to obtain in real time an estimation of the burnt gas mass fraction present in the mixture at the level of the intake manifold:

BGR(t)=BGR_bp(t−τ(t))  (E2)

FIG. 2 illustrates a result of the estimation of the burnt gas mass fraction present in the mixture at the level of the intake manifold, using the method according to the invention. The working points considered cover a wide range: the engine speed ranges between 1000 rpm and 3000 rpm with increments of 500 rpm and, for each engine speed, the torque requirement is 20, then 220 Nm. The estimated BGR value (BGR^e) is shown in full line and the measured real value (BGR^r) in dotted line. The ordinate axis gives the BGR value (estimated or measured) and the abscissa axis the time t in s.

3. Engine Control from the Estimated BGR

Estimation of the composition of the gas in the intake manifold then allows the combustion to be controlled. It is thus possible to control the ignition advance or the composition of the gas entering the cylinder.

I. Ignition Advance Control

In a combustion engine, combustion of the air/gasoline mixture usually starts after the spark. The flame front propagates and its blow drives part of the mixture against the cylinder walls and the piston top. The pressure and temperature rise become so significant that the fuel pressed against the walls reaches its auto-ignition point and ignites in several places. This phenomenon is referred to as “engine knock”. Thus, engine knock is first and foremost an abnormal combustion phenomenon in spark-ignition engines, externally perceptible by a metallic noise coming from the engine. It results from the appearance of shock waves in the combustion chamber.

In order to prevent such a phenomenon, ignition is controlled as a function of the engine speed and of the sucked air mass so as to guarantee that no combustion will occur. Thus, conventionally, ignition advance control is written as follows:

AaA=f(Ne,M_air)

where:

• • function f is a static map
  • AaA is the crank angle at which ignition has to be achieved
  • Mair is the air mass (fresh air and burnt gas)
  • Ne is the engine speed.

In cases where low-pressure EGR is used, the presence of burnt gas (thus neutral) allows the combustion to be slowed down and this abnormal combustion phenomenon to be avoided. Thus, we can re-optimize the ignition advance in order to improve the combustion efficiency. This re-optimization is based on the knowledge of the gas composition in the intake manifold, while adding a proportional term to the burnt gas mass fraction BGR(t) in the intake manifold. The advance control is thus rewritten as follows:

AaA=f(Ne,M_air)+g(Ne,M_air)BGR

where function g is a static map.

Functions f and g are determined on the test bench as it is known to specialists.

In this engine control strategy, the BGR estimation method is very important for controlling the combustion.

Ii. Gas Composition Control

This thus allows the composition of the burnt gas in the cylinder to be controlled. The following stages are therefore carried out:

• • selecting a burnt gas fraction set point BGR^sp in intake manifold (3),
  • estimating the burnt gas mass fraction BGR(t) present in the mixture at the level of the intake manifold, according to stages 1 and 2,
  • calculating an opening set point O^sp of EGR valve (6) allowing the difference between BGR(t) and BGR^sp to be minimized, and
  • controlling said EGR valve (6) as a function of the opening said point of EGR valve (6).

According to an embodiment, an opening set point O^sp of EGR valve (6) is calculated with a proportional integral type controller. Applying opening set point O^sp to the valve can be done using manual, hydraulic, pneumatic, electric, electronic or mechanical controls.

In this engine control strategy, the BGR estimation method is vital for controlling the composition of the gas so as to ensure smooth progress of the combustion.

The invention also relates to a combustion engine (1) comprising at least one cylinder (2), an intake manifold (3) and a burnt gas recirculation circuit comprising an EGR valve (6). The engine is provided with a pressure detector (5) at the level of EGR valve (6) or with an air flow meter (5′) at the start of the intake line. The engine also comprises control means (controllers, softwares) suited for applying the control method according to the invention.

Claims

1. A method of controlling the combustion of a combustion engine comprising at least one cylinder, an intake circuit including an intake manifold and a burnt gas recirculation circuit including an EGR valve, characterized in that it comprises the following stages: performing a measurement relative to a gas, fresh air or burnt gas, flow rate, upstream from a mixing space Vbp wherein fresh air and burnt gas mix,estimating a burnt gas mass fraction BGRbp present in said mixing space, from said measurement and a model of the mixing dynamics in said space,estimating a transport delay from said space to said intake manifold,estimating in real time a burnt gas mass fraction BGR(t) in the intake manifold from said estimation of the burnt gas mass fraction present in the mixing space and said transport delay,controlling the combustion from said burnt gas mass fraction BGR(t) in the intake manifold.

2. A method as claimed in claim 1, wherein said model of the mixing dynamics in the mixing space is a function of an atmospheric pressure, of a temperature upstream from the EGR valve, of said space Vbp, of a mass flow rate Dgb(t) of burnt gas fed through the EGR valve and of a fresh air mass flow rate Dair(t) at the intake circuit inlet.

3. A method as claimed in claim 1, wherein pure delay τ is calculated by solving a laminar flow law iteratively over intake circuit zones defined in such a way that thermodynamic conditions in each zone are homogeneous.

4. A method as claimed in claim 3, wherein three zones are defined as follows: a first zone where the thermodynamic conditions correspond to the atmospheric conditions; a second zone where pressure and temperature increase during gas transport, and a third zone where the temperature decreases, which consists of the remaining zone of the intake circuit.

5. A method as claimed in claim 1, wherein the measurement is a measurement of the fresh air mass flow rate at the intake line inlet.

6. A method as claimed in claim 1, wherein the measurement is a measurement of a pressure difference ΔP at the EGR valve, and the burnt gas mass fraction BGRbp present in said space is estimated from pressure difference ΔP and from a punctual pressure drop relation applied at said EGR valve.

7. A method as claimed in claim 1, wherein combustion control is achieved by controlling an ignition advance, by adding a proportional term to the burnt gas mass fraction BGR(t) in the intake manifold.

8. A method as claimed in claim 1, wherein combustion control is achieved by controlling an air mass contained in the cylinder, by controlling said EGR valve (6) so as to minimize the difference between BGR(t) and a burnt gas fraction set point BGRsp in said intake manifold.

9. A combustion engine comprising at least one cylinder, an intake manifold and a burnt gas recirculation circuit comprising an EGR valve, said combustion engine being provided with a pressure detector at the level of said EGR valve or with an air flow meter at the start of the intake line, characterized in that said engine also comprises control means suited for applying the control method as claimed in claim 1.