FR2771170A1 - METHOD OF DETERMINING THE LEVEL OF A LIQUID IN A CLOSED CONTAINER AND ITS APPLICATION TO THE FUEL TANK OF A MOTOR VEHICLE - Google Patents

Abstract

The container can be selectively and sequentially communicated with a device for producing a pressure variation within the container or with the atmosphere and the pressure variation in the container is detected by means of a pressure sensor and is analyzed for the determination of the level, a pressure change is produced in the volume of gas, unfilled with liquid, of the container when the volume is closed to the atmosphere, and the pressure production is interrupted when a preset pressure value ps is reached and the gas volume is vented, so that a pressure equalization takes place up to the ambient pressure Pu. The time constant T is determined, proportional to the volume of gas, the variation of the pressure p as a function of time between the start and the end of the pressure equalization and, consecutively, the level FS1, FS2 of the liquid by exploiting the time constant tau.

Description

1. The invention relates to a method for determining the level of a liquid.

  in a closed container, according to which - the container can be put in communication selectively and successively with a device serving to produce a pressure variation inside the container or with the atmosphere and - the pressure variation in the container is detected by means of a pressure sensor and is analyzed with a view to determining the level, - a pressure variation is produced in the volume of gas V, not filled with liquid, of the container when the volume V is closed vis-à-vis vis the atmosphere, - the production of pressure is interrupted when a preset pressure value Ps is reached and the volume of gas is placed in communication with the atmosphere, so that a pressure equalization occurs until at ambient pressure Pu Numerous devices are already known by means of which the level of a liquid in a closed container, for example the fuel reserve contained in a

  motor vehicle tank, can be determined.

  In addition to the use of mechanical / electrical devices

  ques, in the case of which the displacement of a float is for example transmitted to the regulating member of a variable resistance (DE 26 27 865 A), or of purely electrical systems, in the case of which the level is measured at using different resistance elements connected in series and immersed in the liquid (DE 26 45 743 C2), using capacitive means (DE 28 35 744) or using ultrasound (DE 21 52 675 ), it is also possible to use pressure variations in the tank to

determine the level.

  To this end, DE 41 07 786 Ai discloses a device for measuring the proportion of liquid fuel in a tank, which device comprises a cylinder comprising two membrane chambers separated by a membrane. A first membrane chamber communicates with the tank and the other with the atmosphere. During the measurement, the membrane is moved into the membrane chamber communicating with the reservoir, from a defined position, under the action of the force of a spring and the pressure variation occurring when the reservoir is closed. vis-à-vis the atmosphere, or a quantity deduced from the displacement distance of the membrane, is considered to be a measure of the proportion of liquid fuel in the tank. Once the measurement has been made and when the tank is open with respect to the atmosphere, the membrane chamber is closed with respect to the tank and is opened with respect to a fuel circulation pump. , so that the diaphragm is moved into the defined position, against the action of the force of the spring, under the effect of the fuel which passes under pressure in the diaphragm chamber and which, during a new measurement and when the diaphragm chamber is closed vis-à-vis the pump and is open vis-à-vis the tank, is returned to the tank when the

  tank is closed vis-à-vis the atmosphere.

  DE 42 03 099 A1 discloses a method and a device for checking the level of the tank which, in addition to a device for checking the tightness of the tank, also includes a device for determining the gradient used to determine the value of a quantity. gradient of variation from at least one variation in pressure and a time interval associated therewith. There is also an output level supply device which receives the signals from the tank tightness control device, a gas evolution control device and the gradient determination device, in order to output the current value of the level using a known relationship between the magnitude of the pressure variation gradient and the level, when the tank is sufficiently tight and the fuel

gives off enough gas.

  DE 39 29 506 A1 describes a method and a device for measuring the liquid content of a closed liquid tank. The gas pressure in the gas volume is modified, at given time intervals, between preset pressure values and the time required for the pressure variation is measured. From the measured time, the volume of gas is determined and, from this volume of gas, the liquid content of the tank is in turn determined. Once the upper pressure value is reached, a measurement is made of the time which elapses until the lower pressure value is reached by letting the gas escape from the volume of gas and this time is in the selected set less than about 2 seconds, preferably less than

about 1 or 0.5 seconds.

  The invention aims to provide a method by which the amount of liquid in a closed container can be determined in a simple manner with great precision. To this end, the subject of the invention is a method for determining the level of a liquid in a closed container, of the generic type defined in the introduction, characterized by the following operations: - determination of the time constant T, proportional to the volume of gas V, of the variation of the pressure p as a function of the time between the start and the end of the pressure equalization and - consecutive determination of the level of the liquid by exploiting the time constant T. The process according to the invention may also have one or more of the following peculiarities: - the pressure variation is produced by means of a pump device which establishes a vacuum in the container, the pressure variation is produced by a pump device which establishes a overpressure in the container, - the values of the time constant r serve as input quantities of a characteristic table which is placed in a memory of a co measurement and measurement and in which associated values of the level are deposited, - the values of the level of the characteristic table are determined experimentally by means of tests on the test bench, - the time constant T is determined by the fact that 'is measured the time interval within the limits of which, from the prefixed pressure value Ps, the pressure P has equalized with the atmospheric pressure Pu and that one fifth of this time interval is considered as approximate value for the time constant T. - the time constant T is determined by the fact that the pressure variation Ap during the pressure equalization is described by means of a mathematical model in the form of a first order differential equation with the time constant, which differential equation has the following form: Ap =! -atps * (T with I Ap = Ap RC and 8 - i 1 V Po, mix 2 = RC = r4 7 '' p.'p PO, air = air density under normal conditions (PO, air = 1.29 kg / m3) P0, mix = density of fuel vapor under normal conditions (normal temperature To = 273.15 K, normal pressure po = 1013 hPa) Ap = pressure difference P - Pu 6 (t) = distribution of Dirac Apstart = pressure difference Ps - Pu Pu = ambient pressure p = tank pressure ps = threshold value V = volume above the liquid in the container 1 = length of the connecting pipe r = radius of the connection pipe = viscosity of air under normal conditions (y = 1.74.10-5 Ns / m2) T = time constant, - instead of the initial values (Ap, Ap, 6 (t)), filtered values (Apf, Aif, 6f (t)) are used, - the estimated values of the time constants (T) are determined by means of the method of least squares. The invention also relates to the use of the method according to the invention for r the level determination in the case of a fuel tank in a

motor vehicle.

  A very simple method for determining the level in the container is obtained by producing a pressure variation in the container by means of the establishment of a vacuum or an overpressure up to a pressure value. prefixed when the container is closed vis-à-vis the ambient environment, followed by an analysis of the time constant within the limits of which a pressure equalization with respect to the atmosphere takes place, this time constant being

  proportional to the level in the tank.

  An example of implementation of the invention is set out below in detail with reference to the drawings. We see: in Figure 1, a block diagram of the device for implementing the method according to the invention and, in Figure 2, a time course graph

level determination.

  The device shown in a simplified manner in FIG. 1 comprises a container 1, partially filled with liquid, a filling tube of which does not bear any particular mark can be hermetically closed by means of a plug 2. The container 1 communicates with a pump device 7 via a connecting pipe 4. So that, even when the container 1 is completely filled, liquid cannot pass through the connecting pipe 4 and therefore towards the pump device e 7 , this supply line is connected to the upper face of the container. This ensures that the pump device 7 communicates with the gas chamber of the container which is located above the liquid. In FIG. 1, the reference FS1 designates a first level of the liquid and the reference FS2 a second level lower relative to the first level. An electrically controlled shut-off valve 5 is interposed in the connecting pipe 4 between the container 1 and the pump device 7. Appropriate signals for controlling the shut-off valve 5 which are delivered by a control and measure 8 make it possible to modify the cross section of passage of the connecting pipe 4, this connecting pipe 4 being able in particular

  be completely closed or open.

  A vent valve 6 is interposed in the section of the connecting pipe 4 which connects the container 1 to the stop valve 5. Thus, appropriate control signals from the control and measurement device 8 allow to communicate with the atmosphere the

container gas chamber 1.

  A pressure sensor 3 is connected to the upper face of the container 1, the pressure connection of this

  sensor communicating with the gas chamber of container 1.

  It is also possible to use, as a pressure sensor 3, a differential pressure sensor whose first connector communicates with the gas chamber of the container 1 and the second connector, with the atmosphere. The signal from the pressure sensor 3 is sent, for formatting and analysis, to the control and

measure 8.

  The method according to the invention for determining the level in the container is described using the time sequence graph in FIG. 2. An electric vacuum pump, which is activated by means of control signals from the device. control and measurement 8, serves as a pump device 7. This makes it possible to produce a vacuum in the container 1 and the connecting line -7. When using the method according to the invention to determine the fuel reserve in a tank of a motor vehicle, the vacuum prevailing in the intake pipe in specific areas of operation of the internal combustion engine driving the vehicle is used to produce the vacuum and therefore to

level determination.

  Alternatively, it is also possible to use a pump device 7 which produces an overpressure in the container 1. Hereinafter, it is the system variant comprising a vacuum system which is exposed. There is no significant difference in the way of

  proceed for an overpressure system.

  In the initial state (instant to), the air release valve 6 is open with respect to the ambient environment (FIG. 2B) and the stop valve 5 is closed with respect to the

  pump (Figure 2A). Thus, the gas chamber located above

  above the liquid in the container communicates with the atmosphere. The pressure sensor 3 then indicates the ambient pressure Pu. At time t1, the pump device is started, the stop valve 5 is open and the vent valve 6 is closed, so that a vacuum is established in the container 1 and in the connecting line 4. If the vacuum p reaches a threshold value ps prefixed (instant t2), the vacuum establishment phase is terminated and the pump device 7 is switched off. At the same time, the stop valve 5 is closed and the vent valve 6 is open. This results in a pressure equalization between the container and the ambient medium. The air used for pressure equalization passes into container 1 through the

  air valve 6 and via the connecting line 4.

  Since the connecting line behaves as a resistor and the container as a capacity, the pressure rise takes place in a manner equivalent to

  the voltage rise of an RC circuit (resistance circuit-

  capacitor) of an electrical circuit. The pressure rise therefore has an exponential variation over time with the time constant T. This time constant T depends in a proportional manner on the volume of gas V contained in the container. If the content of liquid in the container varies, the volume of gas located above the liquid and therefore the time constant r vary linearly with respect to this variation. It is thus possible that the level in the container is deduced directly from the determination of this time constant T. In FIG. 2C, the time constant T is brought to the level FS1. After a time interval 5T, the pressure equalization is practically complete (instant t3). The variation in pressure over time for level FS2 is shown in broken lines. Since, for this level FS2, the volume of gas is greater than that corresponding to the level FS1, the time constant t is also greater and

  pressure equalization does not end until time t4.

  How the level can be determined from

  of the time constant T is set out below.

  A first possibility consists in starting a time counter when the threshold value ps is reached (instant t2). By means of the pressure sensor 3, the pressure p is monitored and the time counter is stopped when the pressure p again corresponds to the ambient pressure Pu (instant t3 for the level FS1). This time interval t3 t2 then corresponds to a value of 5T. This value, or the normalized value with respect to the time constant r, constitutes an input quantity for a characteristic table KF1 which is contained in a memory 9 of the control and measurement device 8 and in which associated values of the level MSDS are filed. These level values of the characteristic table are determined by means of tests on a test bench and therefore contain, among other things, the geometric parameters

  container and connecting line.

  If the method according to the invention is used in a motor vehicle to determine the content of the tank, the FS level read is indicated directly to the driver of the vehicle and also serves as a parameter for further processing, for example as quantity d input for the diagnosis of a tank venting system leak. It is also possible to determine the level of the container by the fact that the variation in pressure over time during the pressure equalization is described by means of a 1st order model. As already indicated in the introduction, it is possible to use for this purpose the following analogy between an electrical circuit and a

pneumatic circuit.

  electric pneumatic voltage U [V] pressure [p] current intensity [I] mass flow m [kg / s]

Resistance R R = 8-n-

  At Pu r4 Capacity CV = Pmix 'VC = - RT pu In the case of a vacuum, for the mass flow in container 1, it comes: m = -C-Ap (1) For the mass flow in the line link 4, it comes:

I

m = - Ap

R (2)

  The identification of (1) and (2) gives the differential equation of the pressure variation over time during the pressure equalization: Ap 1

* C (3)

  It is a first order differential equation with the time constant: 8. '1- V Po, mix t = RC = 4 (4) 7' Pu'Por and the condition at the origin at instant t = O: Ap (0) = -APstart in which -Apstart corresponds to the pressure

  differential Ps - Pu at time t2 in Figure 2C.

  With the condition at the origin and the distribution of Dirac 6 (t), we obtain the equation: À i AP-- AP -ApM.6 (t). (T Once the pressure value ps obtained at instant t2 (FIG. 2C), the pressure p in the container is recorded for a minimum time (instant t3) With the N pressure measurement values Ap (n.TA) occurring at analysis time steps TA (by example 50 ms), it is possible to present equation (5) at all times: Ap (O) = - -Ap (o) + Ap 8 (0) (pressure at time t = 0) * 1 (pressure 1 step of p (.l T) = _ ap0. TA) + p + .5 (.TA) X TA scan later) A p ((N - 1) TA) = ApN- 1) -TA) + Ap, -5 * 8 ((N-1) .TA) T or, expressed as a matrix equality: y = X. b with Ap (O) -Ap (O) (0) 3y = 0 Apa ( 'TA) X = -. [AT; b = [] 6)

Y = 'OTA (6)

  L-aP ((N- 1). TA 6 (N - 1) TA A p ((N - 1) - TA This results in the estimation formula: = (XTX) 'XTy () for the parameters b = A (8) î for parameters b =. (8) e Equation (5) is therefore used to obtain an estimated value for these parameters.This can for example take place using the LESS-SQUARE algorithm known in mathematics. As a result, a value is obtained for the time constant T. It follows from equation (4) that the volume V of the gas chamber located above the liquid in the container is proportional to the time constant t. Furthermore, this volume of gas is obtained from the difference between the total volume of the container, known and prefixed by the geometric parameters of the

  container, and the volume of liquid that determines the level.

  The value determined for the time constant t reflects

so the level.

  Equation (5) can also, in reverse, be used as a model equation: A. p. = -.APtart.8 (t) 9) The equivalent procedure to equation (5) then provides the estimated parameters: [= u] T (10) -Ir A Pis In the estimation equations used for the determination of the parameters b, the derivative A, (n.TA) at the instants of analysis TA is necessary. This must be calculated in the control and measuring device in a digital way, for example by difference formation

successive values.

  To reduce the high noise of digital differentiation and the high standard deviation of the estimation error which is linked to it, it is also possible to use, instead of the initial values (Ap, Ap, 6 (n. TA)), filtered values (Apf, Apf, âf (n.TA)). A condition is then that all the signals follow the same filter function. As a filter function, it is for example possible to use a 1st order low pass filter. The estimated value of the time constant T or 1 / T constitutes an input quantity of a characteristic table KF2 which is contained in a memory 9 of the control and measurement device 8 and in which associated values of the level FS are filed. The level values of the characteristic table are stored by means of tests on the test bench and therefore contain, among other things, the geometric parameters of the container and the connection pipe. The symbols used in the equations have the following meaning: Ap = pressure difference 6 (n.TA) = Dirac distribution Apf = filtered pressure difference A "f = filtered derivative of the pressure difference f (n..TA) = filtered Dirac distribution PO, air = density of air under normal conditions (P0, air = 1.29 kg / m ') P0omix = density of fuel vapor under normal conditions (normal temperature To = 273, 15 K, normal pressure po = 1013 hPa) Pmix = fuel vapor density Pu = ambient air density T = ambient temperature (= gas volume temperature) Pu = ambient pressure p = absolute tank pressure V = volume above the liquid in the container 1 = length of the connecting pipe r = radius of the connecting pipe = viscosity of the air under normal conditions (? = 1.74.10-5 Ns / m2) Since the pressure sensor has a certain offset, the displacement of the zero point of the heading signal tor can be read before the execution of the process in order to increase accuracy. This can be done by any method known per se. The signal from the tank pressure sensor can then be corrected by this determined offset value for other calculations. Due to the evolution of gas from the fuel, there is already a pressure rise which exhibits behavior similar to the pressure rise, due to the incoming air, which is used for the

level determination.

  A preferred application of this process according to the invention is the determination of the level in the case of a fuel tank of a motor vehicle. The tank systems of motor vehicles are currently often equipped with tank vent systems. In these, there is, among other things, a tank vent valve which is placed in a connecting pipe arranged between a canister, containing an activated carbon filter and serving for the intermediate storage of fuel vapors. , and the internal combustion engine intake pipe driving the vehicle. Since a vacuum prevails in the intake pipe in specific areas of engine operation, this intake pipe can be used as a device for producing vacuum. A shut-off valve, which is located on the canister and which is open for scanning the activated carbon filter, makes the tank system communicate with the ambient medium for venting. A pressure sensor is provided for the diagnosis of a tank system leak. Such a tank system including a diagnostic process is described

for example in DE 44 27 688 Al.

  Taking into account the level of the fuel tank in such a process makes it possible to increase the accuracy of the diagnosis of leakage, without additional means in terms of constituent structural elements, since the pressure sensor necessary for the level determination exists in any case. way already in the tank vent system. Knowledge of the level is especially necessary when leaks in the tank system of the order of magnitude of 0.5 mm

  must be ascertained by such a process.

Claims (10)

  1. Method for determining the level of a liquid in a closed container, according to which - the container can be placed in communication selectively and successively with a device serving to produce a pressure variation inside the container or with the atmosphere and the pressure variation in the container is detected by means of a pressure sensor and is analyzed with a view to determining the level, - a pressure variation is produced in the volume of gas (V), not filled with liquid, of the container (1) when the volume (V) is closed vis-à-vis the atmosphere, - the production of pressure is interrupted when a pre-set pressure value (ps) is reached and the volume of gas (V) is placed in communication with the atmosphere, so that pressure equalization takes place up to ambient pressure (Pu), characterized by the following operations: - determination of the time constant (T), proportional at volume d e gas (V), of the variation of the pressure (p) as a function of the time between the start and the end of the pressure equalization and - consecutive determination of the level (FS1; FS2) of   liquid by exploiting the time constant (t).   2. Method according to claim 1, characterized in that the pressure variation is produced by means of a pump device (7) which establishes a vacuum in the container (1).   3. Method according to claim 1, characterized in that the pressure variation is produced by means of a pump device (7) which establishes an overpressure in the container (1).   4. Method according to claim 1, characterized in that the values of the time constant (t) serve as input quantities of a characteristic table (KF1; KF2) which is placed in a memory (9) of a control and measuring device (8) and in which values   associated with the level (FS1; FS2) are deposited.   5. Method according to claim 4, characterized in that the values of the level (FS1; FS2) of the characteristic table are determined experimentally by means of tests on the test bench.   6. Method according to claim 1, characterized in that the time constant (T) is determined by the fact that the time interval (t3 - t2) is measured within the limits of which, from the pressure value (ps) prefixed, the pressure (p) has equalized with the atmospheric pressure (Pu) and that one fifth of this time interval (t3 - t2) is considered as an approximate value for the time constant (T).   7. Method according to claim 1, characterized in that the time constant (T) is determined by the fact that the pressure variation (Ap) during pressure equalization is described by means of a mathematical model in the form of '' a 1st order differential equation with the time constant (t), which differential equation has the following form: AP = AP- Apsi 8 (t) T with àp = - -Ap RC and x = R. c = 8 .; V Pomix nr Pu Pou P0, air = air density under normal conditions (PO, air = 1.29 kg / m3) PO, mix = density of fuel vapor under normal conditions (normal temperature TO = 273 , 15 K, normal pressure Po = 1013 hPa) Ap = pressure difference p - Pu 6 (t) = distribution of Dirac Apstart = pressure difference ps - Pu Pu = ambient pressure p = tank pressure Ps = threshold value V = volume above the liquid in the container 1 = length of the connecting pipe r = radius of the connecting pipe = viscosity of the air under normal conditions (l = 1.74.10-5 Ns / m2) t = time constant.   8. Method according to claim 7, characterized in that instead of the initial values (Ap, Ap, 5 (t)),   filtered values (Apf, Apf, Èf (t)) are used.   9. Method according to claim 7 or 8, characterized in that the estimated values of the time constants (T) are determined by means of the method of least squares.   10. Use of the process according to one of the   previous claims for level determination   in the case of a fuel tank in a motor vehicle.