FR2987074A1 - Method for measuring leak level of particle filter in gaseous exhaust circuit of internal combustion engine of car, involves estimating soot concentration in exhaust gas flow and leak level of filter from variation of resistance of sensor - Google Patents

Abstract

The method involves regenerating an exhaust sensor to destroy agglomerated soots between the electrodes of the sensor. Maximum biasing voltage greater than or equal to 12 V is applied to the sensor for duration. Variation of resistance is measured at terminals of the sensor. Soot concentration in exhaust gas flow and a leak level of filter are estimated from the variation of resistance. The voltage is modified according to the estimation of soot concentration to obtain a signal-to-noise ratio greater than or equal to 100 and a constant saturation time during a measurement cycle.

Description

The present invention relates to the technical field of on-line measurement of physical parameters or concentration of chemical elements in the exhaust gases from a combustion, such as for example in internal combustion engines or boilers.

More particularly, the invention relates to a method of measuring the level of leakage of a particulate filter housed in an exhaust circuit of an internal combustion engine or boiler. The invention is particularly applicable to the measurement of the concentration of carbonaceous particles, so-called soot, in a gaseous exhaust stream, in order, more particularly, to verify and qualify the state of a particulate filter placed in this exhaust flow. The method of the invention is preferably implemented using a measurement sensor positioned downstream of the particulate filter to diagnose the integrity thereof. Such sensors are in particular known in the automotive field. These sensors are in particular of resistive type, comprising two conductive electrodes separated from each other between which the variation of electrical resistance is measured with the gradual fouling of the sensor subsequent to the deposition of soot between said electrodes. The resistance measurement is carried out automatically according to a determined cycle controlled by an electronic control unit connected to the terminals of the electrodes of the sensor. This electronic control unit then performs a calculation of soot concentration from the measured resistance to qualify the state of the particle filter and trigger if necessary a replacement alert thereof.

The resistance measured across the electrodes of the sensor varies globally decreasing with the gradual increase in the amount of soot deposited on the sensor. This decreasing variation results from the establishment of carbon link bridges between the sensor electrodes due to the deposit of soot on it when the sensor is placed in the exhaust stream of the vehicle when the particulate filter is defective.

The resistance curve across the sensor generally comprises three distinct parts: a first horizontal section of constant resistance corresponding to the percolation phase of the sensor, that is to say the time required to obtain the formation of first carbon bridges between the electrodes, followed by a second decreasing part corresponding to the loading phase of the soot sensor, this second part tending towards a second horizontal asymptote constituting the third part of the curve and corresponding to the saturation of the soot sensor.

When the sensor is saturated, it is then necessary to regenerate it for the realization of new measurement cycles. This operation is performed by means of a heating resistor for burning the soot deposited between the electrodes by heating the surface of the sensor and soot and thus regenerate the sensor for new measurement cycles. During resistance measurement cycles across the sensor, the electrodes thereof are biased through a bias voltage. It has been found that this bias voltage has both an influence on the percolation time and the loading time of the sensor insofar as the electric field produced between the electrodes charges the soot particles present in the gas stream. exhaust, which facilitates their deposit on the sensor. In parallel, the bias voltage applied to the electrodes also influences the duration and quality of regeneration of the sensor when it is saturated.

To allow an effective measurement of the soot concentration in an exhaust stream downstream of the particulate filter and thus to qualify the leakage level of the particulate filter, it is extremely important to regenerate the sensor as quickly as possible and to obtain , cycles in measurement cycles between each regeneration, percolation times and substantially constant loading in order to have sufficient measurement fidelity. The achievement of these optimum conditions for the measurement of soot concentration and therefore the level of leakage of a particulate filter is the main objective pursued by the present invention. In response to this objective, the invention proposes a method for measuring the level of leakage of a particulate filter placed in an exhaust stream, in particular an internal combustion engine, in which the concentration of soot in the exhaust gas flow as a function of the resistance measured at the electrode terminals of a resistive-type exhaust sensor positioned in the exhaust stream, characterized in that it consists in carrying out a cycle i, i being an integer, comprising at least the following consecutive steps: a) Regeneration of the sensor to destroy soot agglomerated between the sensor electrodes, b) Application of a maximum bias voltage Vp equal to or greater than 12V to the exhaust sensor for a time c) Measurement of the resistance variation across the sensor during the duration ti of application of the bias voltage Vp, said duration ti covering phases of percol and d) Estimation of the concentration of soot in the exhaust gas stream and the leakage level of the particulate filter from the measured resistance variation during the period ti saturation of the sensor, e) Modification of the maximum bias voltage Vp as a function of the first soot concentration estimate to obtain, during the measurement cycle i + 1, a signal-to-noise ratio greater than or equal to 100 and a response time of saturation ti + 1 substantially constant. According to one characteristic of the method of the invention, the duration ti of application of the bias voltage Vp is adjusted as a function of the modification of the maximum value of the bias voltage Vp in step e).

According to another characteristic of the method of the invention, the bias voltage Vp during the regeneration steps of the sensor is kept minimum but not zero. According to another characteristic of the process of the invention, the concentration estimates made are compiled after a number n of measurement cycles carried out, n being an integer, and a consolidated estimate of the soot concentration in the exhaust flow and the level of leakage of the particulate filter. According to another characteristic of the method of the invention, the bias voltage Vp is varied during the duration ti to obtain different voltage values Vp during the percolation and charging phases of the sensor. According to another characteristic of the method of the invention, the value of the bias voltage Vp during the percolation phase of the sensor is greater than the bias voltage value Vp during the loading phase of the sensor. According to another characteristic of the method of the invention, during the percolation phase the bias voltage Vp is maintained at a value between 30V and 750V continuously and 30V and 500V alternately and during the charging phase until saturation of the resistance of the sensor, the value of the bias voltage Vp is less than or equal to the bias voltage of the percolation phase and greater than 12V. According to another characteristic of the method of the invention, during the regeneration phases of the sensor, the bias voltage Vp applied to the electrodes of the sensor is less than 12V. According to another characteristic of the method of the invention, the biasing voltage Vp is applied to the electrodes of the exhaust sensor during the duration by modulating the pulse width with a duty ratio of between 0 and 100% and a frequency between 100. Hz and 100 Khz. According to a last characteristic of the method of the invention, the effective regeneration of the sensor after each regeneration phase is verified by measuring the resistance across the sensor when the bias voltage Vp is at a minimum value. Various other characteristics appear from the description given below with reference to the accompanying drawings which show, by way of non-limiting examples, embodiments of the subject of the invention. In the attached figures: - Figure 1 schematically shows the active structure of a resistive exhaust sensor for measuring the concentration of soot in a motor vehicle exhaust system - Figure 2 schematically shows the resistance curve measured at terminals of a sensor according to FIG. 1 during a cycle for measuring the leakage level of a motor vehicle particle filter; FIG. 3 shows resistance curves measured at the terminals of a sensor of a motor vehicle; Exhaust according to FIG. 1 as a function of the bias voltage of the electrodes of said sensor. FIG. 4 is a block diagram of the method for measuring the leakage level of a particle filter in accordance with the present invention. The present invention provides a novel method of measuring the leakage level of a particulate filter using a resistive type sensor placed in a gaseous exhaust circuit of a boiler or, more typically, of an internal combustion engine of a motor vehicle. The method of the invention aims in particular to apply in the automotive field to check the state and performance of particulate filters placed in the exhaust systems of motor vehicles. In order to check the state of the particulate filter of a motor vehicle, an exhaust sensor 1 of the resistive type positioned downstream of the particulate filter (PF) to be diagnosed is used in the context of the invention. As shown diagrammatically in FIG. 1, this sensor 1 is composed of a substrate 2 made of insulating material, on which is disposed: on a first face A, two electrodes 3, 4 in conducting material that can withstand the temperatures reached in the exhaust (900 ° C), two tracks 5, 6 for postponing the contacts to make a measurement and a dielectric material protecting the measuring electrodes 3, 4 on their major part and leaving a measuring window 7 for the collection of soot, and on a second face B or in the substrate 2 (for example made by a planar lambda probe type multilayer technique) an electrical resistance 8 allowing the regeneration of the sensor as well as the measurement of the temperature of the exhaust gases, two tracks 9 , 10 for postponing the contacts to make a measurement and a dielectric material to protect the resistor.

When soot contained in the exhaust gas stream of the vehicle engine comes into contact with the substrate 2 of the exhaust sensor 1 in the collection window 7, soot bridges are created between the electrodes 3, 4, causing a fall the resistance R across the sensor. The cyclic measurement of this resistance R and its evolution over time makes it possible to evaluate the state of the particulate filter (FAP) and the level of leakage thereof. The measuring circuit of this resistor R across the exhaust sensor 1 can for example be effected via a simple voltage divider bridge coupled with an operational amplifier mounted follower. FIG. 2 represents the shape of a resistance curve R measured at the terminals of such an exhaust sensor 1 with respect to time. This curve is broken down into three distinct sections corresponding to three states of the sensor. In a first section S1, the resistance R across the exhaust sensor remains substantially constant over time. This section corresponds to the percolation phase of the sensor, that is to say the duration tp during which no bridge, and therefore no short circuit exists between the electrodes 3, 4 of the sensor 1 and where the resistance R is Max. Then as soon as a bridge of soot has formed between the electrodes 3, 4 the resistance R across the sensor 1 begins to fall and decreases as the accumulation of soot between the electrodes 3,4 of the sensor . This is the so-called sensor loading phase, of duration tc, identified by the decreasing curve section S2 in FIG. 2. Finally, in a third section S3, the curve tends towards a horizontal asymptote and a substantially constant minimum resistance R . This third section S3 corresponds to the saturation phase of the sensor 1 by soot deposited thereon. When the sensor 1 is saturated, it is necessary to regenerate it in order to proceed to a new measuring cycle. This regeneration is carried out by combustion of the soot between the electrodes 3, 4 by heating of the electrical resistance 8 located on the face B of the substrate 2 of the sensor 1. In order to allow the deposit of soot on the substrate 2 of the sensor 1 during the resistance measuring cycles R, the electrodes 3, 4 of the sensor 1 are polarized by applying a bias voltage Vp. The inventors have noticed that the bias voltage Vp of the electrodes 3, 4 of the sensor 1 has an influence on the percolation time tp of the sensor and on the loading speed of the sensor, represented by the slope of the resistance curve R in the section S2 corresponding to the loading phase of the sensor. Indeed the electric field between the electrodes 3, 4 loads the soot particles, which facilitates their attachment to the ceramic substrate. The curves of FIG. 3 clearly show the influence of the bias voltage Vp on these percolation times and the loading speeds of the sensor. Thus, it can be seen that the slope (in absolute value) during the loading phase increases with the increase of the bias voltage and that the percolation and saturation times decrease with the increase of the bias voltage Vp.

Physically, it appears that with a higher bias voltage Vp and equivalent saturation resistance, the particles forming the soot bridges between the electrodes 3, 4 of the exhaust sensor 1 are more numerous, better distributed on the surface of the collecting the substrate 2, of smaller size and better bonded to the substrate than with a low bias voltage, which explains these times of percolation and saturation shorter. In addition, the inventors have also found that when the soot particles present in an exhaust gas stream are loaded with a higher bias voltage, because of their size, their better distribution and their proximity to the substrate, the Sensor regeneration is facilitated once saturation is reached. Indeed, a smaller size and a better distribution improves the regeneration time because there are fewer agglomerates on the substrate. Thus, the energy density delivered by the electrical resistance over the entire collecting surface of the sensor substrate is taken advantage of and the finer particles are consumed faster than the larger particles. Finally, the fact that they are better bonded allows a better heat transfer between the substrate and the particles, which also accelerates their combustion. These various findings have led the inventors to design a new method for measuring the level of leakage of a particulate filter via a resistive exhaust sensor 1, this method taking into account the particular importance of the voltage. Vp polarization of the electrodes 3, 4 of the resistive type of exhaust sensor 1. This process is schematically represented in FIG. 4. According to the method of the invention, a cycle i is carried out, i being an integer, comprising at least the following consecutive steps: a. Regeneration of the sensor to destroy agglomerated soot between the sensor electrodes, 9870 74 9 b. Applying a maximum bias voltage Vp greater than or equal to 12V to the exhaust sensor for a time t, c. Measuring the resistance variation across the sensor 5 during the duration t, of applying the bias voltage Vp, said duration t, covering phases of percolation and loading of the sensor to saturation of its resistance, d. Estimating the soot concentration in the exhaust gas stream and the leakage level of the particulate filter from the measured resistance variation during the sensor saturation time t, e. Modification of the maximum bias voltage Vp as a function of the first soot concentration estimate to obtain, during the measurement cycle i + 1, a signal / noise ratio of greater than or equal to 100 and a saturation time t, ± 1 substantially constant . Thus, before any resistance measuring cycle, the regeneration of the exhaust sensor 1 is first carried out in step a) by heating the resistor 8 to burn the soot present on the sensor between its electrodes 3 4. Then, in order to promote rapid percolation times tp and loading tc, a high bias voltage Vp, greater than or equal to 12V, is applied by default during the first cycle to measurement step b) in order to promote maximum collection of soot on the substrate of the sensor 1 25 during the first measuring cycle and thus to obtain in step d), after measuring the resistance R across the sensor 1 in step c) a rough first estimate the leakage level of the FAP. From this first rough estimate of the leakage level of the FAP during the measuring cycle No. 1, the biasing voltage Vp is modified and adjusted for the measurement cycle No. 2 (i + 1) before proceeding to a measurement cycle. new regeneration at the beginning of this cycle n ° 2.

The first strong and original idea of the method of the invention is therefore to use a variable bias voltage as a function of the soot concentration detected in the exhaust estimated in step d) at each measurement cycle.

Also, according to the method of the invention, there is carried out for each resistance measurement cycle i across the exhaust sensor, an adjustment of the bias voltage Vp of the sensor electrodes as a function of the concentration of soot, and therefore resistance across the sensor, measured in the previous cycle i-1.

In addition, before each new measuring cycle is carried out a complete regeneration of the sensor, thereby resetting the measurement signal. Then, during subsequent cycles, the biasing voltage Vp is readjusted by a downwardly decreasing key so as to obtain loading times of the sensor that are substantially constant, or of the same order of magnitude, which then gives a very good fidelity. measuring resistance and therefore soot concentrations, and therefore the level of leakage of the particulate filter in the exhaust system of the vehicle under test. Still according to the method of the invention, it is also advantageous to consolidate the leak level measurements made after a certain number of cycles. This is the reason why, in a preferred embodiment, the concentration and leak level estimates computed after a number n of measurement cycles are computed, where n is an integer, and establishes a consolidated estimate of the soot concentration in the exhaust stream and the leakage level of the particulate filter. Advantageously, the duration ti of application of the bias voltage Vp as a function of the modification of the maximum value of the bias voltage Vp in step e) of the preceding cycle is also adjusted to each measurement cycle i. . As previously indicated, the regeneration is a very important step because it guarantees a reset of the sensor and therefore the reproducibility of the measurements made using it. The regeneration time and the power required for regeneration depend on the conditions, in particular the soot concentration in the gas, the temperature and the speed of the gas flow.

However, it is important to minimize the regeneration time in order to maximize the measurement time and thus alert as soon as possible of a failure of the FAP. This is the reason why, according to the method of the invention, the bias voltage Vp during the regeneration steps of the sensor is kept minimum but not zero, in order to minimize the collection of the particles and thus to accelerate the regeneration, while while monitoring the evolution of the resistance R. Also, during the regeneration phases of the sensor, the bias voltage Vp applied to the sensor electrodes is preferably less than 12V. The effective regeneration of the sensor after each regeneration phase is thus verified by measuring the resistance across the sensor when the bias voltage Vp is at a minimum value. According to another advantageous characteristic of the method of the invention, the bias voltage Vp is varied during the duration to obtain different voltage values Vp during the percolation and loading phases of the sensor. Preferably, the value of the bias voltage Vp applied during the percolation phase of the sensor is also greater than the bias voltage value Vp during the loading phase of the sensor. Indeed, for the percolation phase to be as fast as possible, it is preferable to apply a high voltage Vp in order to quickly collect soot on the sensor to estimate the leakage level of the FAP. On the other hand, during the loading phase, it is necessary to maximize the signal-to-noise ratio and thus to modify the polarization by reducing the appropriate voltage Vp accordingly.

Indeed, if the bias voltage is too high during the charging phase, the drop in the resistance R is very fast and it proves very difficult to measure the slope (AR / A time), even if the measurement is very much modified by external disturbances. On the contrary, if the bias voltage is too low, the slope (AR / A time) will be very small and the slightest drop of particles on the measurement surface (which may be due to a sudden change in flow rate for example ) or any particle whose size is greater than the average by a factor of 2 or 3, will cause the measurement signal to change abruptly, resulting in noise. With an intermediate bias voltage, related to the leakage of the particulate filter, a large measurement signal can be maintained while minimizing the noise generated by the disturbances. The signal-to-noise ratio is optimized, that is, the signal-to-noise ratio is greater than or equal to 100. In preferred embodiments, the inventors have observed that during the percolation phase the bias voltage Vp must be maintained at a value between 30V and 750V continuously and 30V and 500V AC. Similarly, during the loading phase until the resistance of the sensor is saturated, the value of the bias voltage Vp is preferably less than or equal to the bias voltage of the percolation phase and greater than 12V. According to a preferred embodiment of the method of the invention, the bias voltage Vp is applied to the electrodes of the exhaust sensor during the duration by modulating the pulse width with a duty ratio of between 0 and 100% and a frequency between 100Hz 25 and 100Khz. For this, one can use a high output voltage and generate polarization peaks of the order of 100V of a given duration (a few hundred milliseconds) at a low frequency (of the order of 100 Hz to 100Khz) according to the desired bias voltage Vp. We can imagine all the possible electronic assemblies in order to generate these different polarization voltages and therefore to use this strategy.

Claims (10)

REVENDICATIONS1. Method of measuring the level of leakage of a CLAIMS1 filter. Method for measuring the leakage level of a particulate filter placed in an exhaust stream, in particular an internal combustion engine, in which the soot concentration in the exhaust gas stream is determined as a function of the resistance measured at the electrode terminals of a resistive-type exhaust sensor positioned in the exhaust stream, characterized in that it consists in carrying out a cycle i, i being an integer, at least the following consecutive steps : at. Regeneration of the sensor to destroy soot agglomerated between the sensor electrodes, b. Application of a maximum bias voltage Vp greater than or equal to 12V to the exhaust sensor for a period of time c. Measuring the resistance variation across the sensor during the period of application of the bias voltage Vp, said duration ti covering phases of percolation and loading of the sensor to saturation of its resistance, d. Estimating the soot concentration in the exhaust gas stream and the leakage level of the particulate filter from the measured resistance variation during the saturation time of the sensor, e. Modification of the maximum bias voltage Vp as a function of the first soot concentration estimate to obtain, during the measurement cycle i + 1, a signal / noise ratio greater than or equal to 100 and a saturation time ti ± i that is substantially constant. Measuring method according to Claim 1, characterized in that the duration of the application of the bias voltage Vp is adjusted as a function of the change in the maximum value of the bias voltage Vp in step 30. ). 3. Measuring method according to claim 1 or 2, characterized in that the bias voltage Vp during the regeneration steps of the sensor is kept minimum but not zero. 4. Measuring method according to one of claims 1 to 3, characterized in that the concentration estimates are compiled after a number n of measurement cycles carried out, n being an integer, and establishes a consolidated estimate of the soot concentration in the exhaust stream and the leakage level of the particulate filter. 5. Measuring method according to one of claims 1 to 4, characterized in that the bias voltage Vp is varied during the duration ti to obtain different voltage values Vp during the percolation and charging phases. of the sensor. 6. Measuring method according to claim 5, characterized in that the value of the bias voltage Vp during the percolation phase of the sensor is greater than the bias voltage value Vp during the loading phase of the sensor. 7. Measuring method according to one of claims 5 or 6, characterized in that during the percolation phase the bias voltage Vp is maintained at a value between 30V and 750V continuously and 30V and 20 500V AC and during the loading phase until saturation of the resistance of the sensor, the value of the bias voltage Vp is less than or equal to the bias voltage of the percolation phase and greater than 12V. 8. Measuring method according to one of claims 1 to 7, characterized in that during the regeneration phases of the sensor the bias voltage Vp applied to the sensor electrodes is less than 12V. 9. Measuring method according to one of claims 1 to 8, characterized in that the bias voltage Vp is applied to the electrodes of the exhaust sensor during the duration by modulating the pulse width with a duty cycle included between 0 and 100% and a frequency between 100 Hz and 100 Khz. 10. Measuring method according to one of claims 1 to 9, characterized in that it verifies the effective regeneration of the sensor after each regeneration phase by measuring the resistance across the sensor when the bias voltage Vp is at a minimum value.