DE60201570T2 - Method for regulating the fuel tank vacuum of a vehicle - Google Patents

Description

The The present invention relates to a method for the determination and for controlling the negative pressure in a fuel tank for automobiles, which is generated by the cleaning / emptying of the absorber for fuel vapors becomes. The motor vehicle is equipped inter alia with a motor, operated with fuel stored in the fuel tank is with a dipstick for the level fuel, an absorber for fuel vapor, with Lines that are necessary for the emptying / cleaning of the absorber for fuel vapors, a electromechanical actuator, which is provided, the cleaning / emptying the absorber for Fuel vapors to regulate, and which between the latter and the intake manifold of the Motors on one of the mentioned Lines lies and with an electronic calculator, which one Control signal of the electromechanical actuator determines and generates.

Within the goal of complying with the standards for the emission of hydrocarbons by a motor vehicle, in addition to improving the tightness hydrocarbons-containing agents, it is known that those with internal combustion engines and in particular engines with controlled ignition, called "gasoline engines" equipped are, an absorber for Fuel vapors to use. This allows the fuel vapors coming from the fuel tank go out, which will be called simply tank in the following, too store to discharge in Avoid the surrounding air, whether the vehicle and / or its engine stopped / is or in operation.

Of the Absorber for Fuel vapors hereinafter simply absorber (in English: canister), appears as a container, which the fuel vapors stores, by means of absorption elements which are permeable, z. B. by activated carbon. He is generally with three openings equipped: one for connection to the tank, one for suction and one for release.

The opening to the Connection of the absorber is to the tank by means of a permeable pipe connected, called the connection line. Because of his Construction, and in the absence of leaks, the tank allows the Fuel vapors, to escape alone to the absorber. The release opening is an opening the absorber to the atmosphere and so the tank is put under atmospheric pressure, over the Absorber and in particular about the absorption elements.

Of the Absorber has one certain storage capacity. If this capacity is reached, it is said that the absorber is saturated. To saturation and therefore the ejection of Hydrocarbons in the surrounding air through the release opening to avoid, it is necessary to regenerate the absorption elements. It is known that the regeneration takes place when the engine is in operation, and that it is appropriate, a system of wires to use what is called cleaning / emptying pipe which the suction opening of the absorber with the intake manifold the engine connects. Movement of pistons of the engine is made a negative pressure in the intake manifold, which allows, over the Difference to the atmospheric Pressure to create an airflow between the release port and the intake manifold, which sweeps over the absorption elements; the air is charging with fuel contained in the absorber and that so obtained fuel-containing mixture is sucked by the engine, to be burned there.

If the engine is stopped, and thus the absorber as effective as possible is, it is trivial that the openings are spaced for release and for connection and by the Absorption elements are separated from each other. Also, to the To clean absorber, it is necessary that the openings for absorption and release by the same absorption elements are separated. Such are the openings therefore, close to each other for connection and suction; she can in the same zone of the absorber above the absorption elements be placed; or one (both) openings can easily penetrate into the absorption elements, while the separation between the openings for connection and suction in relation to the separation of these openings remains weak from that of the release.

It is known that the evaporation of fuel vapors from the tank is dependent on the temperature inside it and its level. Further, a low fuel tank typically emits more fumes than a heavily filled one. The following reasons favoring the vaporization of fuel may be mentioned, the contact surface between the liquid fuel and the overlying gas is generally the most important, for a filling of the tank in the order of 40% of the total volume, the thermal effects ( Heating of the fuel due to, for example, the return of excess fuel, which emanates from the injection line and which is hot or because of the thermal radiation of the exhaust pipe of the engine), their transition phases (increase in temperature the fuel) are much faster when the amount of fuel charged is reduced, the movements of the liquid fuel in the tank caused, inter alia, accelerations and decelerations of the motor vehicle, which are much more significant when the tank is only partially filled.

The fuel-containing mixture, which by the cleaning / emptying the absorber is not allowed in the cylinders of the Motors cause differences in the mixing ratio, which are harmful to compliance of standards the emission of pollutants by the engine are. So it is known that this contribution is controlled and limited by a device which comprises an electromechanical actuator, z. B. an electric valve, which by the electronic computer is controlled and which in the cleaning / emptying line integrated, programmed control means in the electronic computer, which, inter alia, the signal for controlling the electromechanical Actuator determine and generate the following simply actuator is called.

These Device allowed to pass in the area of passage of the gases of the cleaning / emptying line regulate, the loss of charge of this line is therefore variable dependent from the control of the actuator. The actuator can have several intermediate positions own, between a complete Closure, d. H. an infinite loss of charge of the cleaning / emptying pipe, and, a maximum opening, d. H. a minimum loss of charge (but not 0) of the purge / dump line. It is therefore appropriate to the opening of the actuator to reduce the contribution of fuel-containing Reduce the mixture. The following is appropriate to the general Consider case in which the actuator has several intermediate positions the opening has; and those which are controlled by a cyclic aperture ratio are equal to them, the opening position is through the Given RCO value of the cyclic opening ratio given.

The opening of the Actuator is progressively through an opening step causes. So it is also for closing, except in certain cases, not executed, in which an abrupt and complete close is effected.

It is also known that certain vehicles, depending on the applicable standard for Pollutant emissions in the country of marketing, in operation one Diagnosis of the tightness of the fuel tank and associated elements have to effect if a loss, equal or higher as determined by the norm, it is therefore necessary to determine given information of a bad functioning to the driver be, for. B. by lighting a warning light on the dashboard of the vehicle is placed. The well-known diagnostic principle is a normally open vent valve to use that placed on the release opening of the absorber is, and a pressure sensor on the fuel tank or on one of the elements that has the function of containing fuel are assigned as the absorber is placed. The procedure for diagnosis is divided into several phases, the tank is in Negative pressure is offset by the closing of the ventilation valve and the opening of the Actuator. After a certain delay, the actuator is closed, will, as the vent valve remains closed, then a vacuum is obtained in the tank and this develops depending from a possible loss. The output by the pressure sensor Information allows re-rising the pressure in the tank process and assess the significance of the loss. If the test is complete, the vent valve is open and the Actuator is opened, to clean / empty the absorber.

If the motor is in operation and the actuator is open, even partially, then become static negative pressures at the height the openings for sucking and connecting the absorber generated (possibly the values are identical). This vacuum at the height of the opening to Connection exists, whether a ventilation valve on the release opening the absorber is present or not, and if it is present, whether it is open or not. Since the fuel tank is tight, developed the pressure inside it to balance itself with that the at height the connection opening of the absorber prevails, due to a flow, which in the connecting line arises. The negative pressure caused by the cleaning / emptying of the Absorber was thus produced in the tank, the latter is capable and / or damaging the connected elements, such as the fuel pump.

To limit the negative pressure in the tank, to a reasonable value, so that the latter is not damaged, there are several solutions. It either has the disadvantage of increasing the cost of the vehicle or reducing the cleaning / emptying of the absorber. As examples of solutions may be mentioned: the integration of a flap or a safety valve in the connecting line and which has the task to close this line when the negative pressure at the level of the connection approach of the absorber is too significant. The use of a flap or safety valve placed on the Tank, and which sets the tank in communication with the atmosphere, when the negative pressure in the same becomes too substantial, in addition to the cost of the flap, the drawback is the risk of sucking up external particles which subsequently occlude the elements of the fuel pump or filters can. The reduction of the loss of charge between the joints of the connection and the suction by means of the use of absorbent elements which are more permeable but more expensive. However, increasing the loss of charging the charge / discharge line (eg, by increasing that of the actuator) allows to reduce the negative pressure of the tank, but it also limits the cleaning / depleting of the absorber, and this permanently also be the existing in the tank vacuum.

The problem is even more critical if the vehicle is equipped with a vent valve and which, in the event of a failure, does not open when the leak diagnosis of the tank is completed. The patent US 54996613 , applied to a vehicle equipped with a vent valve and a pressure sensor intended for this leak diagnosis, teaches a method which uses information derived from a pressure sensor to diagnose a failure of the vent valve and, if so, the Case is, and with the aim of avoiding the damage of the fuel tank, and which closes the actuator to terminate the connection between the tank and the source of the negative pressure, which is the engine. Meanwhile, the device is not applicable to a vehicle which does not have a pressure sensor mounted on the tank, further, a total closure of the actuator does not allow to clean / dump the absorber, finally, the described method is only correct under the condition in that the pressure sensor is not defective.

The The present invention aims to overcome these disadvantages of the prior art to remedy the technique by providing a method of estimating and is proposed for regulating the pressure in the tank of the motor vehicle, which allows the negative pressure in a tank to be specific Limit value while a possibility for cleaning / emptying of the absorber is obtained. Applied to a vehicle equipped with a pressure sensor, the for one Leakage diagnosis of the tank is determined, it also allows the to diagnose proper functioning of the pressure sensor. In fact, can too much negative pressure damages the fuel tank and associated elements, such as the fuel pump.

Further Particularities and advantages of the present invention become clearer in the following description, by reference was made on the attached drawings, in which:

1 Figure 12 is a schematic of the systems involved in the purification / depletion of the absorber and by the present invention.

2 a system for measuring the negative pressure .DELTA.P _{can represents} at the height of the connecting piece.

3 the results of the measurement of ΔP _can with the system 2 represents.

4 a simplified diagram of the initialization process of the strategy for controlling the electromechanical actuator represents.

5 a simplified diagram of the method for calculating the strategy for controlling the electromechanical actuator (solenoid valve, controlled by RCO) with modification thereof by the method according to the invention.

6 FIG. 4 is a diagram of the initialization procedure of the strategy for estimating the negative pressure in the tank. FIG.

7 a simplified diagram of the method for calculating the strategy for controlling the electromechanical actuator (solenoid valve, controlled by RCO) with modification based on the invention represents.

8a . 8b illustrate a diagram of the method for calculating the estimation of the negative pressure in the tank.

Before in detail the method according to the invention is described, it is appropriate to the phenomenon of the generated in the tank Low pressure to remember.

If the engine is in operation and the actuator is open, even partially, then a negative pressure on height the openings created for sucking and for connecting the absorber. Because the tank is dense, the pressure inside it develops to enter the To come to equilibrium with the one at the height of the opening to Connection of the absorber prevails, due to a current, the arises in the connecting line. The direction and the throughput the flow in the connecting line vary over time with the Developments of the control signal of the actuator and the pressures in the tank and the intake manifold. The negative pressure in the tank is minimal and of a value of 0 mbar, after closing the actuator or stopping the engine and this, if a delay, the for the stabilization of this vacuum is necessary has expired is. The negative pressure in the tank is maximum, for a maximum negative pressure in the tank elbow and a minimum loss of charge to the line for cleaning / emptying and this, after a delay, the for the stabilization of this vacuum is necessary has expired is.

Of the static negative pressure, at altitude the opening is generated to connect the absorber is mainly dependent on the static pressure in the intake manifold of the engine, the static Pressure on height the release port, trivial to the atmospheric Pressure, the loss of charge between the openings for suction and the Release dependent from the permeability the totality of the absorption elements, the construction properties of the absorber), the loss of the charge (possibly not 0) between the openings for suction and connection, loss of charge of the purge / drain line dependent from the opening of the actuator, of a variable and for the stabilization of this Negative pressure necessary delay, physical properties and the temperature of the gas mixture, which traverses the cleaning / emptying line.

The Direction and flow rate of the flow, which in the connecting line arises, are mainly dependent from the negative pressure to height the opening for connecting the absorber, from the negative pressure in the tank, from Loss of charge in this pipe and physical properties and the temperature of the gas mixture, which is the connecting line crosses.

So, and because the opening of the actuator dependent is from the control signal generated by the computer, are therefore the direction and the flow rate of gases in the connecting pipe mainly depending on the following parameters: the static pressure in the inlet of the Motors, the static pressure at the release port, trivially to the atmospheric pressure, the corresponding loss of charge between the openings for aspiration and release, loss of charge (possibly not 0) between the openings for suction and connection, the loss of charge of the cleaning / emptying line dependent from the control signal of the actuator, the negative pressure in the tank, the loss of charge of the interconnector and the physical Properties and temperatures of the gas mixture containing the pipes to connect and the cleaning / emptying traverses, due the variable and for stabilizing the static negative pressure at height the connection opening the absorber necessary delay.

The for the Stabilization of the negative pressure in the tank is necessary delay variable and is mainly dependent of the following parameters: the mass flow rate, which the connection line crossed and therefore mainly from the new parameters shown above, the volume of the gas in the tank, taking into account, that the variation of the volume of a liquid, depending on the pressure, negligible is opposite that of a gas, the physical properties and the temperature the gases in the tank.

The Fact that the tank at the same time of its liquid fuel is emptied, for the requirements of the supply of the engine, does not change the described phenomena, since this is a parameter which determines the volume of the tank in the tank Gases determined.

The Method according to the invention has the task, in particular, in the open loop and permanently to estimate the vacuum present in the tank. This is an iterative model in which the time step is carefully chosen, which is the mass calculated on gases which traverses the connecting line and which, associated with a calculation of the volume of the tank contained gases, allowing the mass of existing in the tank To calculate gases and consequently the level of negative pressure.

The Construction of the model is based on the aforementioned preliminary analysis, nonetheless, it is possible certain simplifications. This is shown below Model takes into account certain Parameters as constants, since their variations are not significant estimation error entail.

Around To simplify the model, the gas mixture containing the pipes flows through for connection and for cleaning / emptying, as well as that in the tank, as from a homogeneous and constant Composition considered, at a same temperature, which is regarded as a constant, and the atmospheric pressure is also assumed to be constant. Therefore, the model uses one Value of the gas density, a gas constant, a temperature of the gas and an atmospheric pressure, which carefully are constants to be defined. Finally, the model is based on following theses: the gas is perfect, the transformations are isoterm and the variation of the volume of the tank depends on the Pressure is negligible (ie: negligible Deformation of the walls).

The Volume of gases in the tank and its connected elements, can easy from the level of the tank and the known characteristics of its construction will be like: maximum volume of fuel contained can, dead volume (remaining volume of the gases when the tank maximum filled is), etc. ... The height the level of the fuel tank can be known: 1) by means of the dipstick for fuel, already present in the vehicle, the information obtained by the computer where this strategy is programmed, be it directly over one Wired connection, be it indirectly via another computer, which is mounted on the vehicle and multiplexing the two computers, 2) possibly by means of the fuel consumption of the engine.

There the model works in the open loop, it may be desirable be a reserve negative "worst Case "to calculate. So can consequently the values influenced by the constants to be defined be selected under realistic values. Also will in one of the following formulas, a correction term for eventual consideration the tolerance regarding the accuracy of the fuel dipstick integrated information integrated. It is also understood that if the mass of fuel vapors, in the tank during of the operation of the engine is neglected, this then in the direction this model "worst Case "goes.

The calculation model begins with the initialization of the negative pressure in the tank, denoted ΔP _res . A value 0 can be used for ΔP _res , taking into account that when the engine is started, the negative pressure in the tank is 0: ΔP _res (t ₀ ) = 0, or a value equal to 0, taking into account that Negative pressure already exists in the tank, z. In a case where the possibility of initializing the reserve negative pressure to the maximum allowable ΔP _res = ΔPmax _res or ΔPmax _res is the threshold of the negative pressure in the engine Tank, which should not be exceeded, which is defined and programmed in the electronic computer.

Thereafter, the iterative calculation is made, at each step Δt of the calculation, according to the following sections:
the first section consists in calculating the volume of gases in the tank Vgas _res by: VGas res (t i ) = V res - V carb (t i ) - ε jauge + V mort (E.1)

With:
Vgas _res : to calculate the volume of gases in the tank, of which the unit is m ³ .
V _carb : volume of fuel in the tank whose unit is m ³ , and which is known by the information _{coming from} the fuel _dipstick , which should preferably be filtered.
V _res : maximum volume of fuel that the tank can contain, whose unit is m ³ , which is a known constant.
V _mort : minimum volume of gases contained in the tank, whose unit is m ³ , which is a known constant.
ε _jouge : error of the tank _gauge , homogeneous at a volume whose unit is m ³ , which is a known constant.

From ΔP _res (t _i-1 ) and VGas _res (t _i ), a reference value of the mass of gases in the tank, denoted by Mgaz_ref _res (t _i ), is derived by: Mgaz_ref res = VGas res (t i ) × [P atmo - ΔP res (t i - 1 )] / (r × T) (E.2) or trivially, t _{i - 1} = t ₀ in the first step of the calculation.

With:
Mgaz_ref _res : the calculated reference mass of the gases contained in the tank whose unit is Kg.
T: Temperature of gases in Kelvin, considered as a constant whose value is to be defined.
_At atmospheric pressure, considered as a constant whose value is to be defined.
r: gas constant whose unit is J / Kg / K, considered as a constant whose value is to be defined.

The reference mass of the gases Mgaz_ref _res is calculated at each calculation step to take into account: 1) the evolution of the fuel level in the tank, and 2) the level fluctuations that may be present, even though the information from the fuel gauge is present is smooth. It is believed that the variations will vary over time, to an actual level, as the reference mass of the gases varies by their actual value over time. Consequently, the negative pressure in the tank, which will be calculated below, will remain around its actual value.

To the reference term Mgaz_ref _res , a gas mass ΔMgaz _{gen is} added, which represents the mass that is generated by evaporation during the time step .DELTA.t, and a value of the mass flow rate .DELTA.Mgaz _res , which crosses the connec tion line during this same time step, depending on the direction of the Flow of gases in the connection line added or subtracted. ΔMgaz _res results from the cleaning / emptying of the canister. The mass of gases in the tank Mgaz _res , at time t _i , is obtained by: Mgaz res (t i ) = Mgaz_ref res (t i ) + ΔMgaz gene (t i ) ± ΔMgaz res (t i ) (E.3)

With:
Mgaz _res: mass of the gases in the tank is calculated by calculating the mass of the gases generated by evaporation of the fuel and estimation, the unit is Kg.

Which then allows, at time t _i , to derive the static negative pressure in the tank ΔP _res by:

The iterative calculation is therefore fed back at the output of the calculation of ΔP _res by means of the above equation E.4. It is appropriate to explain below how ΔMgaz _gen and ΔMgaz _res are obtained.

ΔMgaz _conditions can in various ways, not carried out by means of z. A formulation which may be empirical, semiempirical or experimental.

Before carrying out the calculation of ΔMgaz _res , it is appropriate to clarify that the following formulas (E.9), (E.5) and (E.5 ') are given by means of the equations of Barré St. Venant and of Bernoulli, suppose that first the variation of the height between the fuel tank and the absorber is negligible, secondly that the idea of a reservoir implies that the velocity of the gas in the tank is 0, thirdly the fact that the negative pressure in the tank, the negative pressure at the level of the connection opening of the absorber and the negative pressure at the level of the intake manifold have as a common reference the atmospheric pressure.

If a vent valve is present in the closed position on the absorber release port, the gas drawn in by the engine is withdrawn alone into the tank (ignoring the volume of gases in the absorber and the purge / dump lines and the connection in proportion to the volume of the gas contained in the tank). ΔMgaz _res can therefore be calculated directly by means of the values of the negative pressure in the tank ΔP _res and the value of the negative pressure in the intake manifold ΔP _coll . Taking into account the circumstances of the observed pressures, the formula of Barré St. Venant is used. A correction coefficient β in the equivalent section is used to stabilize the model.

so ist die Strömung subsonisch und der Durchsatz wird berechnet durch:

andernfalls ist die Strömung sonisch und der Durchsatz wird berechnet durch:If

so the flow is subsonic and the throughput is calculated by:

otherwise the flow is sonic and the flow rate is calculated by:

The both equations, the inequalities, and the conditional execution in a single formula (E.9).

And, considering at each time point t _i that the occurrence of calculations in the time interval Δt is sufficiently small and much less than the occurrence of observed inversions of the flow rate, P _amont and P _{aval are} defined according to the flow direction by: if ΔP coll (t i )> ΔP res (t i - 1 ), then P aval = P atmo - ΔP coll (t i ) and P amont = P atmo - ΔP res (t i - 1 ) unless P amont = P atmo - ΔP coll (t i ) and P aval = P atmo - ΔP res (t i - 1 )

With:
ΔP _coll : static vacuum at the intake manifold of the engine.
γ: ratio of the mass heats of gases, considered as a constant whose value is to be defined.
Se: equivalent cross section of the connection line, which is a constant whose value is known.
Δt: time step, or recurrence of calculations, Δt = t _i - t _{i - 1} , whose choice is decisive for the precision of the calculation.
β: correction coefficient.

If the device does not include or has one vent valve and this is in the open position, then the mass flow rate that passes through the solenoid valve is drawn on the one hand into the atmosphere via the release of the absorber and, on the other hand, the tank. ΔMgaz _res is therefore calculated inter alia on the basis of values of the negative pressures at the ends of the connecting line.

und, zu jedem Zeitpunkt t_i, berücksichtigend, dass das Auftreten von Berechnungen in dem Zeitintervall Δt hinreichend klein gewählt ist und deutlich weniger als das Auftreten von beobachteten Umkehrungen des Durchsatzes, wird angenommen, dass die Strömungsrichtung durch das Vorzeichen bestimmt wird von: ΔPres(ti – 1) – ΔPcan(ti) (E.6) The generally observed ratios of the pressures at the ends of the connection line are weak, so the formula of Bernoulli can be used to the mass flow rate ΔMgaz _res during a time .DELTA.t to calculate:

and, at each time t _i , taking into account that the occurrence of calculations in the time interval Δt is chosen to be sufficiently small and significantly less than the occurrence of observed inversions of the flow rate, it is assumed that the flow direction is determined by the sign of: .DELTA.P res (t i - 1 ) - ΔP can (t i ) (E.6)

With:
ΔP _can : static negative pressure at the level of the connection opening, which is calculated.
ρGas: gas density, considered as a constant whose value is to be defined.
α: correction coefficient.

Eventually, to replace (E.5) and (E.6), the formula of Barré St. Venant can be used, resulting in the following equation to calculate the mass flow rate ΔMgaz _res during a time .DELTA.t.

so ist die Strömung subsonisch und der Durchsatz wird berechnet durch:

andernfalls ist die Strömung sonisch und der Durchsatz wird berechnet durch:If:

so the flow is subsonic and the throughput is calculated by:

otherwise the flow is sonic and the flow rate is calculated by:

The both equations, the inequalities, and the conditional execution in a single formula (E.9).

And, considering at each time point t _i that the occurrence of calculations in the time interval Δt is sufficiently small and much less than the occurrence of observed inversions of the flow rate, P _amont and P _{aval are} defined according to the flow direction by: if ΔP can (t i )> ΔP res (t i - 1 ), then P aval = P atmo - ΔP can (t i ) and P amont = P atmo - ΔP res (t i - 1 ) Unless: P amont = P atmo - ΔP coll (t i ) and P aval = P atmo - ΔP res (t i - 1 )

ΔP _res is the sought value, so it is appropriate to use the value estimated in the previous step of the calculation.

ΔP _can should be known and can be determined by means of the calculations and the parameters properly programmed into the computer, a procedure of which is explained below:

The parameters are the results of the measurements carried out on the elements which coincide with those used by the vehicle in the following way: under the chosen atmospheric pressure P _atmo and the gas temperature, the test is carried out either on a test _stand or directly in the vehicle; the purge / drain line containing the actuator connects the intake port of the absorber to the intake manifold of the engine as previously shown. The absorber may be equipped with a vent valve on the release opening, in this case it is in the open position. The connection line is omitted and replaced by a static pressure sensor which is fixed on the connection opening of the absorber. For various control signals of the actuator, each characterizing an opening position of the latter, and, for various negative pressures in the manifold ΔP _coll , the negative pressure ΔP _{can is} measured after stabilizing its value to a constant. The control signal, denoted by SC, may be characterized by one or more parameters, e.g. B: voltage, control frequency, cyclic opening ratio, etc ... One remembers PSC _j the j-th of the n detected parameters to describe the signal SC. From the measurements a value matrix of ΔP _{can is} obtained dependent on several vectors of values of ΔP _coll and of the PSC _j , parameters (n) of the control signal: SC: [ΔP can ] = F {[ΔP coll ], [PSC 1 ] ... [PSC j ], [PSC n ]}

The parameters programmed in the calculator are the matrix [ΔP _can ], the vector [ΔP _coll ] and the n Vectors [PSC _j ]. The value of ΔP _can is determined at the time t _i depending on these parameters and the values ΔP _coll and the PSC _j determined by the computer in advance in the same step of the calculation t _i . .DELTA.P can (t i ) = f {[ΔP can ], [ΔP coll ], [PSC 1 ] ... [PSC j ], [PSC n ] ΔP coll (t i ), PSC1 (t i ) ... PSC j (t i ) ... PSC n (t i )} E.7

It is therefore considered that the occurrence (or the variation between two calculation steps) of the negative pressure at the connection port ΔP _{can is} instantaneous (all effects of the dynamic behavior of the gas are neglected here). The correction coefficient α of the equivalent cross section just allows to stabilize the model. This allows to obtain a simple, fast and easy to calibrate model.

The negative pressure in the intake manifold ΔP _coll is obtained by the difference between the previously defined atmospheric pressure P _atmo and the static pressure in the intake manifold, which is known for the requirements of the control of the engine and either from the detection of information of the sensor for the static pressure located on the intake manifold or originating from the recovery of information from one or more other sensors.

In the context of a vehicle equipped with a pressure sensor, which is placed on the absorber near the intake port, for example for the requirements of an on-board diagnosis of a fuel leak, the information coming from this sensor can then be used as a value of ΔP _can .

Further can, according to one Variant of the invention, the constant Se by two constants Se1 and Se2 are replaced, and one or the other depending on the flow direction which is determined by the equation (E.6) the connection line is generally one (several) permeable mechanical (s) System (s) includes, such as a safety valve when overturning the vehicle (English acronym: ROV), which (s) a different equivalent Cross section according to the flow direction with can (can) bring. These equivalent Cross sections can be measured. In the same way, the coefficient α can be replaced are given by two different coefficients α1 and α2.

Likewise, according to another variant of the invention, the constant Se and the coefficient α are replaced by the term Se2 and α3 when the closure of the solenoid valve is abrupt. In fact, it is preferable to no longer calculate ΔMgaz _res between the ends of the connection line, but between the release opening of the absorber and the tank.

According to one embodiment of the invention, it is possible to integrate into the model a leak of a previously defined cross-section. For example, a leak with a maximum cross-section defined by the applicable standard and which is an on-board diagnostic. For this purpose, the equation E.3 is modified by subtracting a term which is representative of the mass of the lost through the leak gas, designated ΔMgaz _fuite (t _i), during the time step .DELTA.t: Mgaz res (ti) = Mgaz_ref res (t i ) + ΔMgaz gene (t i ) ± ΔMgaz res (t i ) - ΔMgaz fuite (t i ) (E.3 ')

And the mass of the gas lost by the leak during the time step Δt, between the tank and the atmosphere, can be modeled similarly to the calculation of ΔMgaz _res :

in the As part of the method of regulation according to the invention, it is appropriate a value for to use the cross section of the leak from zero to a negative pressure "in the worst Case ".

After all it is according to the invention possible, to improve the prediction of the model in the following way. It are considered as variables, for example by means of sensors originating information, the following parameters as constant were adopted, such as the temperature of the gas, the atmospheric pressure. It may also be the composition of the gas in the absorber and the tank, for example by means of a strategy, based on the deviations of the mixing ratio, which by the exhaust gas probe to be watched. It can at each step of the calculation, the gas constant and the gas density can be inferred.

An object of the invention is therefore achieved by means of the initialization of the value of ΔP _res (t0) followed by an iterative calculation of the negative pressure in the tank which is fed back at each iteration of the occurrence Δt with the following use of the detection of the level of the fuel Vcarb , then equations E.1, E.2, E.7, E.6, E.5, E.8, E.3 'and E.4 (or equations E.1, E.2, E. 7, E.6, E.5, E.3 and E.4 for the first application of this invention with a calculation of a "worst case" negative pressure).

The method for controlling the negative pressure in the tank is to compare, at each time point ti, the estimated level of the vacuum in the tank ΔPres with a threshold value of the negative pressure, which is defined in the electronic computer and program, which Apmax _res is called. When the threshold is reached or exceeded, the strategy then calculates, at each instant t _i , a threshold above one of the parameters of the signal controlling the actuator, which ensures a value of ΔP _can equal to the threshold ΔPmax _res . By default, the first parameter of the signal SC is considered to be the one for which a threshold value, designated PSCmax1, is calculated. It can be determined by means of matrices of parameters [ΔP _can ], [ΔP _coll ] and [PSCj] with j = 1, n previously defined and which are programmed in the calculator, the threshold of negative pressure ΔPmax _res and the values ΔP _coll and PSCk with k = 2, n known to the electronic computer, by: PSCmax1 (ti) = g {[ΔP can ], [ΔP coll ], [PSC1] .. [PSCj] .. [PSCn], ΔPmax res (ti), ΔP coll (t i ), PSC2 (t i ) ... PSCK (t i ) .. PSCN (t i )}

If not, if ΔP _{res is} less than ΔPmax _res , then the invention sets the value of PSCmax1 to the maximum value that can be used, ie, the full opening of the electromechanical actuator.

According to a variant of the invention, two threshold values of the negative pressure in the computer 10 defined and programmed. In fact, a first threshold may be defined which corresponds to a target for the negative pressure which is not desirable to exceed, and a second threshold (whose magnitude is greater than the first) which absolutely should not be exceeded.

Of the threshold to be used is selected according to parameters. To the Example is the recorded threshold value of the first, except in one the following cases: 1) in the absence of information about the level of the dipstick, becomes a standard level of the dipstick is used as a lowered mode (eg, the full one Tank corresponding level) and to the cleaning / emptying of the Absorbing absorbers by this method becomes the second threshold 2) if information about the degree of charge of the Absorbers available is (for example by means of a method which the charge of the Absorber with fuel estimates) and if the grade is higher is defined as a charge defined and programmed in the computer, which may correspond to the case in which the absorber in the course of saturation with fuel, then to avoid the emission of fuel into the atmosphere, the second threshold is applied to the cleaning / emptying of the absorber should not be restricted by this procedure.

Thus, the invention is to generate a signal for controlling the actuator by taking into account the previously determined threshold PSCmax1 (t _i ). The new control signal is determined in the following calculation step t _i + 1. As previously stated, the basic function of the actuator is to limit the supply of fuel-containing mixture by the cleaning / emptying of the absorber, PSC1 is determined to be closed at each calculation step to regulate. After the initial calculation of PSC1 (t _i - 1) is added to an additional condition: if PSC1 (t _i + 1)> PSCmax1 (t _i), then PSC1 (t _i + 1) = PSCmax1 (t _i). The signal for controlling the actuator is then generated with the consequence of reducing the opening of the actuator and a much greater than intended restriction on the supply of fuel.

The calculations concerning the estimation of the negative pressure in the tank in step t _i + 1 are then carried out, and so on, the whole of the strategy is therefore fed back.

Thus, it is by PSC1 (t _i + 1) is limited to PSCmax1 (t _i), keeping the negative pressure in the container at a value of ΔPmaxres at t _i + 1, if not, the negative pressure will decrease at closing of the actuator.

In a further embodiment The described method of regulation is not based on one estimate the height the negative pressure as defined in the formulas described above, but by directly measuring the level of negative pressure over one Pressure sensor in the tank or on one connected to the tank Element lies.

The The invention will now be described in more detail with reference to FIG the adjacent figures.

The vehicle, according to 1 , uses an absorber 1 for fuel vapors, which allows to catch the discharge of hydrocarbons coming from the fuel tank 2 come by means of a connecting line 3 which the tank 2 with the connection opening 101 of the absorber 1 combines. The absorber 1 is equipped with two further openings, a release opening 102 and an intake opening 103 , The release opening 102 is from the connection opening 101 through absorption elements 104 separated, which are permeable. The inlet opening 103 is with the intake manifold 7 the engine (not illustrated) downstream of the throttle 9 connected, via a cleaning / emptying line 4 into which an electrovalve 6 is integrated, which is an electromechanical actuator, by the electronic injection computer 10 the motor is controlled by means of a control signal defined by an amplitude which is equal to the battery voltage designated Ubat and by two more by the computer 10 certain parameters, a control frequency designated FC and a cyclic opening ratio denoted by RCO. The Tank 2 is with a dipstick 5 equipped for the level of fuel, which with an electronic calculator 11 of the vehicle is connected. The calculator 10 and 11 are interconnected via a multiplex system 12 connected, through which they can exchange information and so the computer attacks 10 to the information of the level of fuel in the tank to, which by the dipstick 5 is delivered. The negative pressure in the manifold 7 is obtained via the difference to a level of atmospheric pressure, which in the calculator 10 is determined and programmed, by means of the pressure, which by a sensor of the static pressure 13 is detected, with the computer 10 is connected and on the manifold 7 lies.

For given battery voltages Ubat and frequency FC, the negative pressure ΔPcan is at the level of the connection opening 101 characterized in dependence on the negative pressure in the manifold and the RCO for controlling the solenoid valve 6 , z. By means of the system of measurements according to 2 , The device of 2 picks up the elements of the conduit of the absorber 1 on. It essentially comprises a vacuum pump 20 , which simulates the negative pressure in the intake manifold, a sensor 21 of the differential pressure prevailing upstream of the solenoid valve allows to measure the negative pressure in the manifold, this value being displayed on a screen. A generator of the RCO signal to control the solenoid valve, a differential pressure sensor located on the line 3 is placed to connect the absorber. This connection sensor allows the negative pressure at the level of the connection opening 101 of the absorber, this value being displayed on a screen. The experiment is carried out under an atmospheric pressure equal to that in the computer 10 is programmed.

The results obtained by means of the device are in the form of the graph 3 represented by template values [ΔPcan], obtained depending on vectors [ΔP _coll ] and [RCO]. On the abscissa RCO varies from 0 to 100%. The solenoid valve is fully closed when the RCO is equal to or less than an orifice offset, which is mainly dependent on the negative pressure in the manifold 7 and the Steuerfrequnz FC. The solenoid valve is fully open for a RCO of 100%. Between the two values, the opening of the solenoid valve is partial and dependent on the value of RCO used (and in this illustrated embodiment it is linearly dependent on RCO), and consequently, in 3 it can be stated that the negative pressure ΔP _{can is} also proportional to the RCO used. It is found that it is trivial that the RCO is the parameter of the control signal to be used to control the negative pressure at the communication port 101 to limit to a desired threshold of negative pressure. Since then, a table of values, called Table_Dpcan_100, has been defined, which is a vector of the heights of the negative pressure at the level of the connection opening 101 is, for a RCO of 100% dependent on the negative pressure in the manifold ΔPcoll; the values are identical, which is also the control frequency FC. By means of new measurements according to the scheme in 3 It is possible for other battery voltage values and other control frequencies, new heights of negative pressure ΔP _can at the level of the connection opening 101 and to determine new opening offsets. From the results, it is easy to work in the calculator 10 Derive values to be programmed.

The algorithms, the variables and the constants that are necessary for the control of the engine, of those that control the cleaning / emptying of the absorber 1 are dedicated and of those dedicated to the method according to the invention are properly in the computer 10 programmed.

After under voltage networks of the computer 10 and on the verge of other steps necessary for the operation of the engine, the procedures for initializing the various strategies, of which the initialization of the control strategy of the solenoid valve 6 for cleaning / emptying, shown in 4 , and the initialization of the estimation strategy of the negative pressure in the tank 2 represented in 6 , executed.

According to 4 The various variables necessary to operate the purge / dump strategy will be in phase 201 initialized, wherein the RCO is initialized with its minimum value, denoted by RCO _{lim_min} . This value corresponds to the physical limit of the component or the limit due to a need for electrical diagnosis.

According to 6 Start the initialization of the variables necessary for the estimation of the negative pressure in the tank through the phase 301 with the initialization of the variable ΔP _res , then with the initialization of the variable RCOmax with 100% of the opening in phase 302 ,

According to 5 After the phases of initialization, the procedures for calculating the strategy for cleaning / emptying the absorber 1 and the negative pressure estimation strategy (later with reference to FIG 8a and 8b described) in the tank, by the computer at each repetition of the calculation, which is set to a duration of Δt. The procedure for estimating the negative pressure in the tank 2 begins exactly after the end of the procedure for calculating the absorber cleaning / emptying strategy 1 , The procedure for controlling the solenoid valve starts through the phase 401 which is to check that the conditions of the cleaning / draining are met. As long as the different ones are not yet in common to activate the cleaning / emptying, the algorithm remains in phase 401 , If the conditions are common then the phase becomes 402 activates and determines the control frequency FC of the solenoid valve 6 , then it will be in phase 403 the opening offset, referred to as OFF_RCO, of the solenoid valve 6 determined, depending on the control frequency FC and the initial negative pressure in the manifold ΔP _res . A target value for RCO for control, labeled OBJ_RCO, then an opening step, labeled PAS_RCO, will be successively in the phases 404 and 405 certainly. Then it will be in phase 406 run a test. If the RCO in the previous calculation step is greater than the calculated target value for RCO, then phase becomes 407 otherwise, it is phase 408 which is completed. The phases 407 and 408 have to task the RCO to control the solenoid valve 6 decreasing or increasing, taking into account the limits of RCO, denoted by RCO _{lim_min} and RCO _{lim_max} (physical limits of the device or limits due to requirements of electrical diagnosis).

The basic strategy 5 can coincide with 7 be modified by a phase 409 will be added. The RCO is capable of being limited to the value of RCOmax which is in phase 302 has been initialized and which is then determined, as will be seen, in phase 514 or phase 515 , The procedure terminates and then, with parameters of frequency and RCO, the computer generates 10 the control signal, which is the solenoid valve 6 opens and a negative pressure arises at the level of the connection opening 101 one.

According to the 8a and 8b The diagram of the method for estimating the negative pressure in the tank then begins with the detection of the level of the dipstick 5 in the phase 501 What the calculation of the volume of gas in the tank 2 in phase 502 allowed according to formula E.1. The phase 503 calculated, depending on the negative pressure in the tank (value in phase 301 initialized, then after each repetition of 513 calculated) and the volume of the gas in the tank, a reference mass of the gas Mgaz_ref _res according to formula E.2 with K, which one in the computer 10 programmed constant and which K = r × t has to the value. The phase 504 calculated, depending on the negative pressure in the manifold .DELTA.P _coll , the negative pressure .DELTA.Pcan_100 at the level of the connection opening 101 for a RCO of 100% using the table of values labeled Table_Dpcan_100. Since this embodiment uses an electrovalve whose throughput, for a same negative pressure in the manifold, is pseudo-linearly dependent on the value of the RCO, the negative pressure in the connection opening becomes 101 ΔP _can in phase 505 calculated by means of a proportionality rule depending on the RCO for controlling the electrovalve in phase 409 was determined and offset by the RCO, which is in phase 403 was determined. A difference of the negative pressures ΔP _res -ΔP _can exists between the ends of the connection line 3 which are in phase 506 is calculated according to the formula E.6. The result is stored in the variable ΔΔP. The invention assumes that the sign of this difference, in phase 507 determines the direction of flow results. If the sign is positive, then calculate the phase 508 the variation of the mass ΔMgaz _res for a flow in the direction of the tank 2 to the absorber 1 by means of interpolations in a value table denoted Table_DM_1 depending on the absolute value of ΔΔP. This value table is determined by calculations using equation E.5. Then, in phase 509 , the new mass will be in the tank 2 derived existing gas. If the sign of the in phase 507 certain difference is negative, then becomes phase 510 executed. It is similar to the phase 508 but with a table of values designated Table_DM_2 which is specific to a flow in the direction of the absorber 1 to the tank 2 , Then, in phase 511 , the new mass will be in the tank 2 derived gas according to formula E.3, where the term ΔMgazgen (ti) is considered zero. After completion the phase 509 or the phase 511 , will be the phase 512 carried out in accordance with equation E.4 to calculate the negative pressure in the tank ΔPres with K, which is one in the computer 10 programmed constant and which has K = r × T to the value. The value of ΔPres is in phase with a threshold value of the negative pressure ΔPmaxres 513 compared. If the value of ΔPres is less than the threshold ΔPmaxres, then phase becomes 514 and the maximum value of 100% is given to the variable RCOmax, otherwise the phase becomes 515 and calculates the value to be given to the variable RCOmax such that the application of this value implies that the negative pressure at the level of the connection opening 101 is equal to the threshold value of the negative pressure ΔPmax _res . In this last case, in the following calculation step, the RCO becomes in phase 409 limited and so the negative pressure in the tank does not exceed the value of ΔPmax _res .

The Method according to the invention Can be applied to a vehicle that has a ventilation valve and a pressure sensor which is placed on the tank is. So it is possible the value of the negative pressure detected by the sensor with the value of the pressure estimated by the method of the invention.

• – Ein Leck in dem Tank oder auf der Verbindungsleitung oder auf der Reinigungs-/Entleerungsleitung. Tatsächlich bewirkt das Leck eine Zufuhr von Luft, welche die Unterdrucksetzung des Tanks vermindert.
• – Eine Verstopfung in der Reinigungs-/Entleerungsleitung. Tatsächlich erhöht diese Verstopfung, selbst teilweise, den Verlust an Ladung der Reinigungs-/Entleerungsleitung. So ist der Unterdruck ΔP_can viel schwächer und infolgedessen ist der Unterdruck in dem Tank viel schwächer.
• – Die vorigen zwei Mängel sind gleichzeitig vorhanden.

If the measured negative pressure is less than the estimated negative pressure, either:

• - A leak in the tank or on the connecting pipe or on the cleaning / emptying pipe. In fact, the leak causes a supply of air, which reduces the pressurization of the tank.
• - A blockage in the cleaning / emptying pipe. In fact, this obstruction, even partially, increases the loss of charge to the purge / dump line. Thus, the negative pressure ΔP _{can is} much weaker and as a result, the negative pressure in the tank is much weaker.
• - The previous two defects are present at the same time.

If the measured negative pressure is greater as the esteemed Negative pressure, then there is a blockage of the release opening of the Absorber. This clogging can be due to contamination or a Failure of the solenoid valve to vent due to the solenoid valve is closed instead of open when it should be open.

Claims (8)

Method for regulating the negative pressure in a fuel tank ( 2 ) for vehicles, which by the cleaning / emptying of an absorber ( 1 ) for fuel vapors by means of an electromagnetic actuator 6 generated in the cleaning / emptying line ( 4 ) of the absorber, the method being characterized in that it comprises: a step of estimating the negative pressure prevailing in the tank in the open loop and in a certain time interval (t _i ), according to the formula:

where ΔP _res (t _i ) of the estimate of the level of vacuum at time t _i is P _atmo the atmospheric pressure Mgaz _res (t _i ) is an estimate of the mass of gas present in the tank, depending on ΔP _res (t _{i - 1} ) and from the direction of movement of the gas between the tank and the absorber r the constant of the gas present in the tank is T the temperature of the gas in Kelvin is Vgaz _res (t _i ) the volume of gas present in the tank. A method according to claim 1, characterized in that it comprises a step of comparing the level of the negative pressure with a certain threshold value, wherein when the threshold value is exceeded, the method comprises a step of generating a signal for controlling the electromechanical actuator ( 6 ) to keep the level of negative pressure below or equal to the predetermined threshold. A method according to claim 1 or 2, characterized in that the mass of the gas present in the tank is defined by the formula: Mgaz res (t i ) = Mgaz_ref res (t i ) + ΔMgaz gene (t i ) ± ΔMgaz res (t i ) With Mgaz_ref res (t i ) = Vgaz res (t i ) × [P atmo - ΔP res (t i - 1 )] / (r × T) where Mgaz_ref _res (t _i ) corresponds to the reference mass of the gas to be calculated in the tank, ΔMgaz _gen (t _i ) corresponds to the mass of the gas generated by evaporation during the time Δt, Mgaz _res (t _i ) corresponds to the mass of the gas containing the Connecting line during the time .DELTA.t traversed .DELTA.t corresponds to the duration of the particular interval. A method according to claim 3, characterized in that the mass of the gas which traverses the connecting line during the time Δt is defined by the formula:

by taking α as a correction coefficient and Se corresponding to an equivalent section of the connecting line between the absorber and the tank ΔP _can (t _i ) corresponds to the static negative pressure at the connection opening between the absorber and the connecting line, this negative pressure being determined by a cartography, which depends on the pressure in the intake manifold and on at least one of the parameters of the signal for controlling the electromagnetic actuator. Method according to 3 or 4, characterized in that the formula of the mass of the gas present in the tank is corrected by subtracting a term ΔMgaz _fuite (t _i ) which is representative of the mass of the gas which an escape of gas from a certain section S _fuite corresponds, which is defined by the formula:

by taking α as a correction coefficient. Method according to one of claims 1 to 5, characterized in that the volume of the gas present in the tank is determined by the formula: Vgaz res (t i ) = V res - V carb (t i ) - ε jauge + V mort where V _res corresponds to the maximum volume of fuel that the tank may contain, V _carb (t _i ) - ε _{jauge corresponds to} the actual volume of fuel present in the tank, this volume corresponding to information _resulting from the fuel gauge Mounted in the tank, V _mort corresponds to the minimum volume of gas present in the tank. Method according to one the claims 2 to 6, characterized in that the method comprises a step the measurement of the prevailing in the tank vacuum by a Pressure sensor includes that in the tank or one connected to the tank Element is placed, a step of comparing between the estimated level of negative pressure and the amount of negative pressure measured by the sensor. Internal combustion engine equipped with an absorber ( 1 ) for fuel vapors, which is mounted in connection with the fuel tank, the engine comprising a computer, which is intended to realize the method according to one of claims 1 to 8.