FR3128974A1 - Method for low-frequency estimation of a flow rate of exhaust gases recycled to the inlet of an internal combustion engine - Google Patents

Abstract

According to this method for calculating the exhaust gas recirculation flow rate at the intake of an internal combustion engine: The minimum Πmin and maximum Πmax values of an expansion rate Π are measured, The mean value of the rate of expansion Πmoy, We calculate the Barré de Saint-Venant function applied to the average value of the expansion rate, We calculate the second derivative of the Barré de Saint-Venant function, and We calculate the recirculation flow rate of the gases of exhaust. Figure for abstract: Figure 2

Description

Method for low-frequency estimation of a flow rate of exhaust gases recycled to the intake of an internal combustion engine

The present invention relates to motor vehicle internal combustion engines of the spark ignition type (gasoline) or of the compression ignition type (diesel), and relates in particular to a partial exhaust gas recirculation circuit at the exhaust from such an engine, called the EGR circuit, for “Exhaust Gas Recirculation” in Anglo-Saxon terms or “Recirculation des Gaz d'Echappement” in French.

It relates more particularly to a method and a system for low-frequency calculation of the flow rate of the exhaust gases recycled in such an EGR circuit, also known as the EGR flow rate.

Prior techniques

Classically, exhaust gas recirculation is a process that limits the production of nitrogen oxides (NOx) in the combustion gases of a diesel engine, or reduces the fuel consumption of gasoline engines. This EGR process consists of taking gases from the exhaust and sending them to the intake, for example downstream of an engine air flow control valve. On diesel engines, the action on the formation of NOx can go up to 50% of overall reduction and is interpreted by a reduction in the temperatures of the combustion gases due to the dilution and a consequent slowing down of the kinetics of formation of the polluting. On gasoline engines, the reduction in fuel consumption is due to a lesser sensitivity of the engine to knocking which makes it possible to increase the ignition advance and to the reduction of the losses by pumping of the engine, the contribution of recycled gases making it possible to increase the pressure in the intake manifold of the engine for the same value of air flow necessary for the production of the torque.

Generally the quantity of recycled gas is estimated using a differential pressure sensor at the level of a valve for regulating the flow of the EGR circuit, called the EGR valve. This sensor makes it possible to estimate the flow which has just passed through the valve, called the EGR flow. Good control of the EGR flow is essential because it determines the EGR rate, i.e. the ratio between the flow of exhaust gases at the intake and the total flow of gases admitted into the engine. engine (recycled air and gas), which affects the level of NOx pollutants or fuel consumption. Poor control of the EGR flow rate can lead to undesired effects that could compromise the operation of the vehicle. For example, too high an estimate of the EGR flow rate can lead to higher NOx emissions in diesel engines or knocking phenomena in spark ignition engines. This can have drastic consequences in terms of degraded reliability or non-compliance with regulatory pollution control standards.

The difficulty in estimating the EGR flow rate is due to the fact that a flow sensor is not used at the EGR valve itself, which would be expensive and inaccurate. Instead of such a flowmeter, a differential pressure sensor is used which measures the difference between the pressure downstream and upstream of the valve. The flow through the valve is estimated based on a model that uses the measured differential pressure and the valve opening level. Even if this model can provide very precise results, the differential pressure information has invariably a disturbed and very dynamic character because presenting high frequency oscillations, difficult to follow using electronic control units whose capacities of calculation are limited. Thus, it is not possible in practice to make calculations with information from the differential pressure sensor captured at high frequency, i.e. with a periodicity of 1 millisecond.

The invention consists in developing a low-frequency calculation method for estimating an EGR flow rate, i.e. with a periodicity of approximately 20 milliseconds, while maintaining a level of precision equivalent to a high-frequency calculation.

There are, in the state of the art, different methods for estimating an EGR flow rate.

The first method is to make a model of the total flow of the engine intake gases, on the one hand, and to determine the only air flow, on the other hand. For this, the total flow can be determined by a filling model, from the pressure and the temperature in the engine intake manifold, and from a filling efficiency value which is itself a function of a set of parameters comprising at least the speed, the pressure in the intake manifold, and possibly other parameters such as the timing position of the intake and exhaust valves. The fresh air flow is determined by a flow meter. The EGR flow is then equal to the total intake gas flow from which the fresh air flow is subtracted. This approach is described in particular in document US 2016/0069285, or in document FR 2 938 016.

These approaches are simple, but have several disadvantages.

For example, for engines that have two EGR circuits, namely a low pressure EGR circuit and a high pressure EGR circuit, they cannot distinguish between low or high pressure EGR rates, only the overall EGR rate that can be estimated using this method.

Then, the estimate is based on knowing the total intake gas flow, which is conventionally a volumetric pump model, and the fresh air flow, which is directly measured by a sensor. The available precision of this information is approximately +/-5% for the measurement per sensor and approximately +/-3% for the total flow, which can generate inconsistent results, particularly on low flow rate values. EGR and transient values.

A second method consists in using a valve cross-section model, which generally uses the Barré de Saint-Venant equation or “Throttle Equation” in Anglo-Saxon equivalent. The equation gives very precise results, if one controls the position of the valve and the pressure difference which exists at its terminals.

On the other hand, it is extremely sensitive to the pressure values at the valve terminals and to the angular position of the valve, when the pressure difference is low or when the valve is close to closing.

In order to improve its accuracy, it is common to add corrective terms taking into account the dispersions, or even to prefer the first method mentioned above on the problematic operating points. This cross section method is used for example in the document EP 1 416 138 with the addition of a correction map taking into account the upstream pressure, the downstream pressure and the cross section of the valve, or in the document EP 3 434 888, with the addition of a dynamic correction map depending on the positions of the valve.

The cross section method has the disadvantage of being highly non-linear in the very closed or very open positions of the valve. On the other hand, the pressure variations are very fast dynamic phenomena, of the order of a millisecond, which are generally filtered for use in the management of the EGR system, whose order of magnitude of the constant of time is around 10 to 20 milliseconds. However, given the strongly nonlinear character of the equation on certain positions, significant errors are introduced by using filtered pressure values.

In view of the foregoing, the object of the invention is to improve the accuracy of the estimation of an EGR flow rate, in particular for use in the management of the low-frequency EGR system.

The subject of the invention is a method for calculating the flow rate of recirculation of exhaust gases at the intake of an internal combustion engine allowing the control of said engine.

According to the process:

• The minimum Πmin and maximum Πmax values of an expansion rate Π are measured, defined as the ratio between the pressure measured upstream and the pressure measured downstream of the exhaust gas recirculation valve,
• We measure the average value of the rate of expansion Πmoy,
• We calculate the Barré de Saint-Venant function applied to the average value of the expansion rate Πmoy,
• We calculate the second derivative of the Barré de Saint-Venant function applied to the average value of the rate of expansion Πmoy,
• The exhaust gas recirculation rate is calculated.

The method for calculating the EGR flow rate thus makes it possible to calculate the EGR flow rate precisely using the Barré de Saint-Venant function and the second derivative of this function, applied to the mean value of the expansion rate Πmoy.

Advantageously, the second derivative of the Barré de Saint-Venant function applied to the average value of the expansion rate Πmoy is calculated using three terms obtained by applying the Barré de Saint-Venant function to the envelope values Πmin and Πmax and at the average value Πavg.

Advantageously, the exhaust gas recirculation rate is calculated by summing a first term directly proportional to the Barré de Saint-Venant function applied to the mean value of the expansion rate Πmoy and a second term directly proportional to the second derivative of the Barré de Saint-Venant function applied to the mean value of the rate of expansion Πmoy.

For example, the minimum Πmin and maximum Πmax values can be measured at high frequency.

For example, the minimum Πmin and maximum Πmax values can be measured at low frequency.

For example, the mean value Πmoy can be measured at low frequency.

Advantageously, the internal combustion engine is equipped with at least one high pressure partial exhaust gas recirculation circuit and at least one low pressure partial exhaust gas recirculation circuit.

The invention also relates to a system for calculating the flow rate of recirculation of exhaust gases at the intake of an internal combustion engine allowing the control of said engine.

The flow rate calculation system comprises means for measuring the minimum Πmin and maximum Πmax values of an expansion rate Π, defined as the ratio between the pressure measured upstream and the pressure measured downstream of the gas recirculation valve d exhaust, means for measuring the average value of the rate of expansion Πmoy, means for calculating the Barré de Saint-Venant function applied to the average value of the rate of expansion Πmoy, means for calculating the second derivative of the Barré de Saint-Venant function applied to the mean value of the expansion rate Πmoy, and means for calculating the flow rate of recirculation of the exhaust gases.

Other aims, characteristics and advantages of the invention will appear on reading the following description, given solely by way of non-limiting example, and made with reference to the appended drawings in which:

schematically illustrates the structure of an internal combustion engine of a motor vehicle equipped with a high pressure partial exhaust gas recirculation circuit, a low pressure partial exhaust gas recirculation circuit exhaust and an EGR flow calculation system according to the invention;

illustrates a flowchart of the method for calculating the EGR flow rate, implemented by the calculation system, according to an embodiment of the invention;

schematically illustrates the information accessible at the level of the EGR valves.

is a logic diagram used for the classic estimation of the EGR flow rate.

is a flowchart used in the invention for estimating the EGR flow rate.

Detailed description of at least one embodiment

In the example shown in the , the internal combustion engine 10 comprises, without limitation, four cylinders 12 in line, a fresh air intake manifold 14, an exhaust manifold 16, a turbo-compression system 18, a recirculation circuit partial high pressure exhaust gas (“high pressure EGR circuit”) and a partial low pressure partial exhaust gas recirculation circuit (“low pressure EGR circuit”).

The cylinders 12 are supplied with air via the intake manifold 14, or intake distributor, itself supplied by a pipe 20 provided with an air filter 22 and the compressor 18b of the turbocharger 18 of the engine 10 .

The turbocharger 18 essentially comprises a turbine 18a driven by the exhaust gases and a compressor 18b mounted on the same shaft as the turbine 18a and providing compression of the air distributed by the air filter 22, with the aim of increasing the quantity (mass flow) of air admitted into the cylinders 12 of the engine 10.

A heat exchanger 24 is placed after the outlet of the compressor 18b equipping the supply line 14a of the intake manifold 14 with fresh air.

The internal combustion engine 10 comprises an intake circuit Ca and an exhaust circuit Ce.

The intake circuit Ca comprises, from upstream to downstream in the direction of air circulation:

- the air filter 22 or air box;

- a flow meter 26 disposed in the intake duct 20 downstream of the air filter 22 to measure the actual value of the air flow entering the engine 10;

- an air inlet valve 28;

- the compressor 18b of the turbocharger 18;

- a throttle body 30 or a gas inlet valve in the engine;

- a heat exchanger 32 to cool the intake gases corresponding to a mixture of fresh air and recirculated exhaust gases after their compression in the compressor 18b; And

- the intake manifold 14.

The compressor is associated with a bypass circuit with an inlet relief valve 55 which opens in the event of sudden closure of the throttle valve 30, to prevent the compressed air, located between the compressor 18b and the valve butterfly 30, does not cross the compressor 18b and does not degrade it, when for example, the driver of the vehicle suddenly lifts his foot from the accelerator pedal.

The exhaust circuit Ce comprises, from upstream to downstream in the direction of circulation of the burnt gases:

- the exhaust manifold 16;

- The turbine 18a of the turbocharger 18; And

- A system 40 for depolluting the engine combustion gases.

As regards the exhaust manifold 16, the latter recovers the exhaust gases resulting from combustion and evacuates them to the outside, via a gas exhaust duct 34 leading to the turbine 18a of the turbocharger 18 and by an exhaust line 36 mounted downstream of said turbine 18a.

The engine 10 here comprises two partial exhaust gas recirculation circuits at the intake, called “EGR” circuits (“exhaust gas recirculation” in Anglo-Saxon terms), namely a high-pressure EGR circuit 15 and a circuit Low pressure EGR 38.

Circuit 38, here a low-pressure exhaust gas recirculation circuit, called “EGR BP”, originates at a point on exhaust line 36, downstream of said turbine 18a, and in particular downstream of the system 40 for gas pollution control and returns the exhaust gases to a point in the fresh air supply pipe 20, upstream of the compressor 18b of the turbocharger 18, in particular downstream of the flow meter 26. The flow meter 26 only measures the flow fresh air alone.

As illustrated, this recirculation circuit 38 comprises, in the direction of circulation of the recycled gases, a cooler 38a, a filter 38b, and a valve 38c intended to regulate the flow of low-pressure exhaust gases. Valve 38c is disposed downstream of cooler 38a and upstream of compressor 18b.

It will be noted that the air intake valve 28 can also be used to force the circulation of a low pressure exhaust gas flow in the EGR circuit BP in the case where the depression between the exhaust circuit and the intake circuit would be insufficient. In this case, closing the valve 28 makes it possible to create a depression downstream thereof, able to suck in gases from the EGR circuit BP.

The high-pressure EGR circuit 15, called “EGR HP”, originates at a point on the exhaust circuit Ce, upstream of said turbine 18a and returns the exhaust gases to a point on the intake circuit Ca, downstream of the heat exchanger 32.

As shown, this recirculation circuit 15 includes a valve 15a configured to regulate the flow of high pressure exhaust gases.

The engine is associated with a fuel circuit comprising, for example, fuel injectors (not referenced) injecting gasoline directly into each cylinder from a fuel tank (not shown).

The engine comprises an electronic control unit 70 configured to control the various elements of the internal combustion engine from data collected by sensors at various locations of the engine.

The electronic control unit 70 comprises a calculation module 72, a measurement module 73 and a control module 74.

As illustrated in the , several pieces of information are accessible at the level of the EGR valves 15a and 38c.

For example, the gas pressure downstream of the EGR valve called Paval is measured using a relative pressure sensor. The differential pressure ΔP is measured using a differential pressure sensor. The opening angle of the θegr valve is measured using a position sensor placed on the electric motor that drives the valve.

It is generally accepted that the flow through the Qegr valve can be described according to the following equation:

(1)

With :

, mass flow, kg/s

S _e , the cross section of the valve, in mm²

BSV, the Barré de Saint-Venant function calculated according to equation (2), below and expressed as

Tamont, the temperature upstream of the valve, in K

Pamont, the pressure upstream of the valve, expressed in Pa.

The Barré de Saint-Venant function is expressed when the flow is non-sonic, i.e. for all the calculations that interest us, by the following expression:

(2)

With :

Π, the rate of expansion, i.e. the ratio between Pamont and Paval, dimensionless

γ, the adiabatic gas index, dimensionless

r, the ratio between the ideal gas constant divided by the molar mass of the gas in question, expressed in J kg ⁻¹ K ⁻¹

By multiplying equation (1) on each side by Π, we obtain the following equation:

(3)

With :

(4)

As mentioned earlier, the pressure information is extremely dynamic due to the acyclic behavior of an internal combustion engine. This acyclism creates strong volatilities of the differential pressure signal and by that implies strong variations on the rate of expansion Π.

In order to limit the noise on the total bit rate, it is necessary to filter this information. However, filtering the pressure information has a direct impact on the accuracy of the flow rate recalculated by equation (1) or (3). Indeed, these equations use the Barré de Saint-Venant (BSV) function which is non-linear.

It is therefore not equivalent to calculate the average of the function over all the pressures encountered, which corresponds to a true average of the flow, than to calculate the average of the pressures and to apply the Barré de Saint-Venant function to it. The difference between these two calculations can be quite significant, sometimes greater than 20% according to certain calculations, which risks leading to an overestimation of the flow rate interpreted by the rest of the motor control.

It is thus proposed to get as close as possible to the real value of the EGR flow rate while performing all the calculations at the lowest frequency, i.e. every 10 milliseconds for example.

To achieve this, the variations in the rate of expansion Π presented above are assimilated to a sinusoid which is added to a mean signal varying as a function of time, the sinusoid having an unknown amplitude A and an unknown pulsation ω as described in the following equation:

(5)

To calculate the EGR flow from equation (3), it is necessary to correctly estimate the function next :

(6)

To this end, we use the Taylor expansions in their approximate forms, namely:

(7)

Which we apply here with And .

We then obtain at second order the following approximation:

(8)

The classic estimation method, which consists, as explained above, of first filtering or averaging the rate of expansion and then applying the Barré de Saint-Venant function would lead to the result in the following form:

(9)

Which corresponds only to the first term of the approximation (8).

The present invention corresponds to a method which consists in using the following terms of the Taylor expansion in order to complete the information available without excessively increasing the computational load.

To calculate the average flow over a given period T, we integrate equation (3) over this period:

(10)

If we consider that the calculation period is quite short, for example of the order of a pressure oscillation, we can assume that the position of the valve And is a constant. We can also assume that the pressure remains constant and that most of the oscillations appear on .

We can ask:

(11)

And rewrite equation (10) as:

(12)

By using equations (12) and (8) we then obtain:

(13)

Which can be rewritten as:

(14)

Over a period T large compared to the pulsation , we obtain

And

By choosing a filter with an adequate time constant, we then obtain:

(15)

To calculate the value of the average EGR flow, it is necessary to calculate the second derivative of the function BSV ₂ .

Using a numerical scheme of the centered differences type, we can then evaluate the second derivative of the function BSV ₂ as follows:

(16)

We then obtain the following expression for the average EGR flow rate:

(17)

This amounts to considering the classical estimate to which we add a correction related to the estimates of maximum throughput and minimum throughput. We note that the corrective is not very greedy in calculation resources because it consists in carrying out weightings for the following three information items only: ¼ for the maximum bit rate, ½ for the average bit rate and ¼ for the minimum bit rate.

We can also notice that all the odd derivatives of the sine functions lead to a zero throughput average because . Expression (17) is thus a third-order approximation, which ensures acceptable precision even when the differences between And become important.

Figure 4 illustrates the procedure used in the conventional estimation, in order to better appreciate the contribution of the invention, the procedure of which is presented in Figure 5. Conventionally, the information of the rate of expansion is captured at high frequency, i.e. all the milliseconds. The information is filtered by taking the average at low frequency, i.e. every 10 milliseconds. We then apply the BSV2 function to the average value and multiplying by the coefficient α to obtain the estimate of the EGR flow rate.

Referring to Figure 5, the information is captured at high frequency, ie every millisecond. We then retrieve the maximum values , minimum and average at low frequency, i.e. every 10 milliseconds. We then apply the BSV2 function to these three maximum values , minimal and average with a weighting of ¼ for the maximum value, ½ for the average value and ¼ for the minimum value and a multiplication is carried out by the coefficient α to obtain the estimate of the EGR flow rate.

The organizational chart represented in the , illustrates the method for calculating the EGR flow, implemented by the calculation system 70.

During a first step 61, the maximum and minimum values of the rate of expansion are measured, then, during the following step 62, the average value of the rate of expansion is measured.

The method 60 further comprises a step 63 for calculating the function BSV2 applied to the average expansion rate, according to equations (2) and (4), and a step 64 for calculating the second derivative of the function BSV2 applied to the rate of expansion. medium expansion, according to equation (16).

The method 60 continues with a step 65 of calculating the average flow, according to equation (17).

The method 60 finally comprises a step 66 of engine control, through an EGR flow setpoint Q.

Thus, the invention proposes a method for estimating the EGR flow rate used for engine control, which consumes little computing resources and which can use the pressure values filtered at low frequency, with sufficient accuracy of the results.

Claims (8)

Method for calculating the flow rate of recirculation of exhaust gases at the intake of an internal combustion engine (10) allowing the control of said engine (10), characterized in that:

1. The minimum (Πmin) and maximum (Πmax) values of an expansion rate Π are measured, defined as the ratio between the pressure measured upstream and the pressure measured downstream of the exhaust gas recirculation valve (15a and /or 38c),
2. We measure the average value of the rate of expansion (Πmoy),
3. We calculate the Barré de Saint-Venant function applied to the average value of the rate of expansion (Πmoy),
4. We calculate the second derivative of the Barré de Saint-Venant function applied to the average value of the rate of expansion (Πmoy),
5. The exhaust gas recirculation rate (Q) is calculated.

Method according to claim 1, in which the second derivative of the Barré de Saint-Venant function applied to the mean value of the rate of expansion (Πmoy) is calculated using three terms obtained by applying the Barré de Saint-Venant function - Coming to the envelope values (Πmin) and (Πmax) and to the mean value (Πmoy). Method according to claim 1, in which the exhaust gas recirculation rate is calculated by summing a first term directly proportional to the Barré de Saint-Venant function applied to the average value of the expansion rate (Πmoy ) and a second term directly proportional to the second derivative of the Barré de Saint-Venant function applied to the mean value of the rate of expansion (Πmoy). Method according to one of Claims 1, 2 or 3, in which the minimum (Πmin) and maximum (Πmax) values are measured at high frequency. Method according to any one of Claims 1, 2 or 3, in which the minimum (Πmin) and maximum (Πmax) values are measured at low frequency. Method according to any one of Claims 1, 2 or 3, in which the mean value (Πmoy) is measured at low frequency. Method according to any one of the preceding claims, in which the internal combustion engine (10) is equipped with at least one high pressure exhaust gas recirculation circuit (15) and at least one low pressure circuit for exhaust gas recirculation (38). System for calculating the flow rate of recirculation of exhaust gases at the intake of an internal combustion engine (10) allowing the control of said engine (10), characterized in that it comprises:

1. Means for measuring the minimum (Πmin) and maximum (Πmax) values of an expansion rate Π, defined as the ratio between the pressure measured upstream and the pressure measured downstream of the exhaust gas recirculation valve (15a and/or 38c),
2. Means of measuring the average value of the rate of expansion (Πmoy),
3. Means of calculating the Barré de Saint-Venant function applied to the average value of the rate of expansion (Πmoy),
4. Means of calculating the second derivative of the Barré de Saint-Venant function applied to the average value of the rate of expansion (Πmoy),
5. Means of calculating the exhaust gas recirculation flow (Q)