JP2017083288A - Filter failure detection device and particulate matter detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a failure detection device of a filter capable of suppressing lowering of the accuracy of the failure determination of a filter for collecting the particulate matter and a particulate matter detection device capable of suppressing lowering of the detection accuracy of a particulate matter quantity due to the influence of a combustion supporting gas in the exhaust gas of an internal combustion engine.SOLUTION: The downstream side of a filter 12 is provided with a PM sensor 13 for detecting a PM quantity (particulate matter quantity). An ECU17 obtains each density of NO, NO, and Oin the downstream side of the filter 12 on the basis of operating conditions of an engine 20 or from a sensor 22, and obtains the temperature of an exhaust gas from an exhaust gas temperature sensor 21. A memory 18 is stored with relation data with a type of NO, NO, and O, and a density and a gas temperature as well as a correction value equivalent to an output reducing degree of the PM sensor 13. The ECU 17 corrects the output of the PM sensor 13 in a direction to increase the PM quantity on the basis of the type, the density, and the temperature of the obtained gas and the relation data, and performs a failure determination of the filter 12 on the basis of a sensor output after correction.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a filter failure detection device that collects particulate matter in exhaust gas discharged from an internal combustion engine and a particulate matter detection device that detects the amount of particulate matter in exhaust gas.

  Conventionally, there has been proposed a device for detecting the amount of particulate matter (particulate matter, PM) in exhaust gas discharged from an internal combustion engine (see, for example, Patent Document 1). Patent Document 1 discloses correcting an output value of an electrical resistance type sensor that outputs a value corresponding to the amount of particulate matter in exhaust gas using an exhaust temperature, a sensor temperature, and an exhaust flow rate. Yes. According to this, the amount of the particulate matter can be detected with high accuracy without being affected by the temperature and the exhaust flow rate.

  Here, the electric resistance type sensor has a portion to be attached to which particulate matter in the exhaust gas is attached, and a pair of counter electrodes provided at a distance from the portion to be attached. Since the particulate matter (Soot) has conductivity, the resistance value between the pair of counter electrodes decreases as the amount of the particulate matter attached to the adherend increases. The sensor outputs a value corresponding to the resistance value between the counter electrodes, that is, a value corresponding to the amount of particulate matter adhering to the adherend.

Japanese Patent No. 5240679

  By the way, according to the results of the investigations by the present inventors, in the presence of a specific combustion-supporting gas in the exhaust gas, a part of the particulate matter collected by the sensor (attached part) is burned and removed. It was found that the sensor output greatly changes to the value side indicating that the amount of particulate matter is small. In addition, when a sensor is arranged downstream of a filter that collects particulate matter and a filter failure determination is made based on the output value of the sensor, the sensor output changes due to the influence of the combustion-supporting gas. As a result, the accuracy of the filter failure determination is reduced.

  The present invention has been made in view of the above problems, and it is possible to suppress the deterioration of the filter failure determination accuracy due to the influence of the combustion-supporting gas, and the particulate matter due to the influence of the combustion-supporting gas. It is an object of the present invention to provide a particulate matter detection device capable of suppressing a decrease in the amount detection accuracy.

In order to solve the above-mentioned problem, the first invention is provided downstream of a filter (12) for collecting particulate matter in exhaust gas provided in an exhaust passage (23) of an internal combustion engine (20). A sensor (13) having an adherent portion (52) for adhering particulate matter therein, and outputting a value corresponding to the amount of particulate matter attached to the adherent portion;
Concentration acquisition means (S2, S14, S34, S54, S74, 17) for acquiring the concentration of the combustion-supporting gas coexisting with the particulate matter in the exhaust gas;
Temperature acquisition means (S3, S15, S35, S55, S75, 17) for acquiring the temperature of the exhaust gas;
Correction means (S5, S76, S77) for correcting the output value of the sensor in the direction of increasing the amount of particulate matter based on the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means. , S81, 17),
Failure determination means (S6 to S8, S82 to S84) for determining failure of the filter based on the output value corrected by the correction means;
It is characterized by providing.

  According to the investigation results of the present inventors, particulate matter adhering to the adherend is burned and removed by an amount corresponding to the concentration and gas temperature of the combustion-supporting gas coexisting in the exhaust gas, and the sensor output changes. I found out. The present invention has been made on the basis of the results of this investigation. The concentration of the combustion-supporting gas and the temperature of the exhaust gas are acquired, and the output value of the sensor is calculated based on the concentration and the temperature, and the amount of particulate matter. Correct in the direction to increase. Thereby, the output value of a sensor can be brought close to the value before combustion by combustion-supporting gas is performed. And since the failure determination of a filter is performed based on the output value after correction | amendment, the failure determination can be performed in the form which suppressed the influence of the combustion by combustion-supporting gas, and the precision fall of failure determination can be suppressed.

The second invention is provided downstream of the filter (12) for collecting particulate matter in the exhaust gas provided in the exhaust passage (23) of the internal combustion engine (20), and attaches particulate matter in the exhaust gas. A sensor (13) having an adherent part (52) to be output and outputting a value corresponding to the amount of particulate matter adhering to the adherent part;
Estimating means (S17, S38, S58, S78, 17) for estimating the output value of the sensor when the filter is a filter serving as a criterion for failure determination;
Failure determination means (S22 to S24, S42 to S44, S62 to S64, S82 to S84, 17) for determining a failure of the filter based on a comparison between the value estimated by the estimation means and the actual output value of the sensor. When,
Concentration acquisition means (S14, S34, S54, S74, 17) for acquiring the concentration of the combustion-supporting gas coexisting with the particulate matter in the exhaust gas;
Temperature acquisition means (S15, S35, S55, S75, 17) for acquiring the temperature of the exhaust gas;
Based on the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means, an output value correction for correcting the output value of the sensor in a direction to increase the amount of particulate matter, and the estimation means Correction means (S16, S18, S19, S76, S77, S81, 17) for performing any one of the estimated value correction for correcting the estimated value in the direction of reducing the amount of particulate matter,
The failure determination unit performs failure determination of the filter using the value corrected by the correction unit.

  According to the second aspect of the invention, the sensor output value is estimated when the filter is a filter serving as a failure determination reference, and the filter failure determination is performed based on a comparison between the estimated value and the actual sensor output value. . At this time, based on the concentration of the combustion-supporting gas and the temperature of the exhaust gas, the output value correction for correcting the output value of the sensor in the direction of increasing the amount of the particulate matter, and the estimated value of the sensor output are the particulate matter. One of the estimated value corrections for correcting in the direction of decreasing the amount is performed. When the output value is corrected, the sensor output value can be brought close to the value before combustion with the combustion-supporting gas, and the sensor output value and the estimated value can be calculated in the state before combustion with the combustion-supporting gas. You can compare. On the other hand, when the estimated value correction is performed, the estimated value can be brought close to the value after combustion by the combustion-supporting gas, and the output value of the sensor can be compared with the estimated value in the state after combustion. Thereby, the failure determination can be performed in a form in which the influence of combustion by the combustion-supporting gas is suppressed, and a decrease in accuracy of the failure determination can be suppressed.

The third invention is provided in the exhaust passage (23) of the internal combustion engine (20) and has an adherent portion (52) for adhering particulate matter in the exhaust gas, and the particulate matter attached to the adherent portion. A sensor (13) that outputs a value corresponding to the amount of
Concentration acquisition means (S2, S14, S34, S54, S74, 17) for acquiring the concentration of the combustion-supporting gas coexisting with the particulate matter in the exhaust gas;
Temperature acquisition means (S3, S15, S35, S55, S75, 17) for acquiring the temperature of the exhaust gas;
Correction means (S5, S76, S77) for correcting the output value of the sensor in the direction of increasing the amount of particulate matter based on the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means. , S81, 17),
It is characterized by providing.

  According to the third aspect of the invention, the sensor output value is corrected in the direction of increasing the amount of particulate matter based on the concentration of the combustion-supporting gas and the temperature of the exhaust gas. It is possible to approach the value before the combustion with the sex gas. Therefore, it can suppress that the precision of the amount of particulate matter detected with a sensor falls under the influence of combustion supporting gas.

It is a block diagram of an engine system. It is the figure which showed the structure of PM sensor typically. FIG. 5 is a diagram illustrating a state in the vicinity of a pair of counter electrodes in a sensor element, and illustrating the principle of PM amount detection by a PM sensor. It is the figure which showed gas concentration and the output decreasing rate of PM sensor for every kind of combustion-supporting gas. _NO, indicating for each O 2, NO ₂ gas species, for the gas temperature is a diagram showing a change in the CO ₂ generation amount corresponding to the combustion amount of PM trapped in the PM sensor. It is the figure which showed the relationship data of gas temperature and NO density | concentration, and output correction value _ANO . Is a diagram showing the relationship data between the gas temperature and the NO ₂ concentration and the output correction value A _NO2. Is a diagram showing the relationship data between the gas temperature and the O ₂ concentration and the output correction value A _O2. It is the flowchart which showed the failure determination process of DPF in 1st Embodiment. It is the figure which showed the change of the sensor output with respect to collection time, and showed the mode of correction | amendment in 1st Embodiment. It is the figure which showed the change of the output of PM sensor with respect to collection time, and showed the actual sensor output and the estimated output of PM sensor in case the DPF is the reference | standard failure DPF. It is the flowchart which showed the failure determination process of DPF in 2nd Embodiment. It is the figure which showed the change of the sensor output with respect to collection time, and showed the mode of correction | amendment in 2nd Embodiment. It is the flowchart which showed the failure determination process of DPF in 3rd Embodiment. It is the figure which showed the change of the sensor output with respect to collection time, and showed the mode of correction | amendment in 3rd Embodiment. It is the flowchart which showed the failure determination process of DPF in 4th Embodiment. It is the figure which showed the change of the sensor output with respect to collection time, and showed the mode of correction | amendment in 4th Embodiment. It is the flowchart which showed the failure determination process of DPF in 5th Embodiment. It is the figure which showed the change of the sensor output with respect to collection time, and showed the mode of correction | amendment in 5th Embodiment.

(First embodiment)
Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram of a vehicle engine system 100 to which the present invention is applied. The engine system 100 includes a multi-cylinder type (four cylinders in FIG. 1) diesel engine 20 (hereinafter simply referred to as an engine) as an internal combustion engine. The engine 20 is provided with an injector 6 for injecting fuel into the cylinder 7. The engine 20 generates power for driving the vehicle by the fuel injected from the injector 6 self-igniting in the cylinder 7.

  The intake passage 10 of the engine 20 is provided with, from the upstream side, a supercharger 1 that compresses air, an intercooler 2 that cools air compressed by the supercharger 1, and a throttle valve 3 that adjusts the amount of air. . The intake passage 10 is connected to an intake manifold 5 having a branch passage connected to each cylinder 7.

  Each cylinder 7 is connected to an exhaust manifold 8 for collectively passing the exhaust gas discharged from each cylinder 7 to the exhaust passage 23. The exhaust passage 23 is provided with the turbine of the supercharger 1, and an EGR (Exhaust Gas Recirculation) passage 19 for returning a part of the exhaust gas to the intake system is connected to the upstream side of the turbine. . The EGR passage 19 is connected to the intake passage 10 at a position between the throttle valve 3 and the air flow meter 4. The EGR passage 19 is provided with an EGR valve 9 that adjusts the amount of exhaust gas recirculated to the intake system (EGR amount).

  The exhaust passage 23 is provided with an oxidation catalyst (DOC: Diesel Oxidation Catalyst) 11 that oxidizes and removes CO, HC, etc. in the exhaust gas downstream from the position where the EGR passage 19 is connected. A diesel particulate filter (DPF) 12 corresponding to the filter of the present invention is provided downstream of the oxidation catalyst 11. The DPF 12 is a ceramic filter having a known structure. For example, a heat-resistant ceramic such as cordierite is formed into a honeycomb structure so that a large number of cells serving as gas flow paths are staggered on the inlet side or the outlet side. Contained and configured. Exhaust gas discharged from the engine 20 flows downstream while passing through the porous partition walls of the DPF 12, and PM (particulate matter, particulate matter) contained in the exhaust gas is collected and gradually accumulated.

  An electric resistance PM sensor 13 as a sensor of the present invention for detecting the amount of PM in the exhaust gas is provided downstream of the DPF 12 in the exhaust passage 23. Here, FIG. 2 is a diagram schematically showing the structure of the PM sensor 13. As shown in FIG. 2, the PM sensor 13 includes, for example, a metal cover 51 whose inside is hollow, and a sensor element 52 arranged in the cover 51. A number of holes 511 are formed in the cover 51, and a part of the exhaust gas flowing through the exhaust passage 23 is introduced into the cover 51 from these holes 511. Further, the cover 51 is formed with a discharge hole 512 for discharging the exhaust gas introduced into the cover 51. FIG. 2 shows an example in which the discharge hole 512 is formed at the tip of the cover 51.

  The sensor element 52 is composed of an insulating substrate such as ceramics. On one surface of the sensor element 52 (insulator substrate), a pair of opposed electrodes 53 spaced apart from each other and opposed to each other are provided. Here, FIG. 3 is a diagram for explaining the principle of detection of the PM amount by the PM sensor 13 and shows the state of PM adhesion in the vicinity of the pair of counter electrodes 53. As shown in FIG. 3, the sensor element 52 is connected to a voltage application circuit 55 that applies a predetermined DC voltage between a pair of counter electrodes 53. Part of the PM in the exhaust gas introduced into the cover 51 is collected (attached) to the sensor element 52 due to its own adhesiveness.

  Further, when a voltage is applied between the counter electrodes 53 by the voltage application circuit 55, each counter electrode 53 is charged positively and negatively, respectively. As a result, PM passing through the vicinity of the counter electrode 53 is charged, and collection into the sensor element 52 is promoted. Hereinafter, PM collection on the sensor element 52 by applying a voltage between the counter electrodes 53 is referred to as electrostatic collection.

  Explaining the output characteristics of the PM sensor 13, the PM sensor 13 utilizes the fact that the resistance between the counter electrodes 53 changes due to the PM collected by the sensor element 52, and the amount of PM collected by the sensor element 52 ( An output corresponding to the accumulated amount of PM adhesion at each time point is generated. That is, the PM sensor 13 outputs a value corresponding to the resistance value between the counter electrodes 53 as the PM amount. Specifically, the sensor output does not occur while the amount of PM trapped in the sensor element 52 is small (strictly, only an output smaller than a threshold output that can be regarded as the sensor output rising) is generated. Since the soot component contained in the PM is composed of carbon particles and has conductivity, when the amount of trapped PM becomes a certain amount or more, the pair of counter electrodes 53 are electrically connected and the sensor output rises (threshold value). Output more than output).

  After the sensor output rises, the resistance between the pair of counter electrodes 53 decreases as the amount of collected PM increases, so that the current flowing between the counter electrodes 53, that is, the sensor output increases. The engine system 1 includes a detection unit 56 (see FIG. 3) that detects the voltage by converting the current flowing between the counter electrodes 53 or the current into a voltage, and the measurement value of the detection unit 56 is a PM sensor. 13 outputs. Note that, for example, a resistance value between the pair of counter electrodes 53 may be measured as a value correlated with the current flowing between the counter electrodes 53, and the resistance value may be used as the output of the PM sensor 13.

  The sensor element 52 is provided with a heater 54 that heats the sensor element 52. The heater 54 is used to regenerate the PM sensor 13 by burning and removing PM collected by the sensor element 52, for example. The heater 54 is provided, for example, on the surface of the sensor element 52 (insulator substrate) where the counter electrode 53 is not provided or inside the sensor element 52. The heater 54 is composed of a heating wire such as platinum (Pt). In the regeneration of the PM sensor 13, a temperature at which all of the components (Soot component, SOF component, etc.) constituting the PM can be burned and removed, specifically, a temperature of, for example, 600 ° C. or higher (eg, 700 ° C.) The heater 54 is controlled. The heater 54 is connected to the SCU 14. The sensor element 52 corresponds to the adherend portion in the present invention.

  Returning to the description of FIG. 1, the engine system 100 is provided with various sensors necessary for the operation of the engine 20 in addition to the PM sensor 13. Specifically, for example, an air flow meter 4 that detects the amount of air sucked into the cylinder 7, a rotation speed sensor 15 that detects the rotation speed of the engine 20 (for example, a crank angle sensor that detects a crank angle), a vehicle driver's An accelerator pedal sensor 24 for detecting an operation amount (depression amount) of an accelerator pedal for notifying the required torque to the vehicle side is provided. Further, an exhaust temperature sensor 21 that detects the temperature of the exhaust gas is provided in the exhaust passage 23 downstream of the DPF 12. Detection signals from these sensors are input to the ECU 17.

  Further, the engine system 100 is provided with an EDU 16 (electronic drive unit) that controls the injector 6. The EDU 16 drives the injector 6 so that fuel is injected under the injection conditions (injection timing, injection amount, etc.) commanded from the ECU 17.

  The engine system 100 is provided with an SCU (Sensor Control Unit) 14. The SCU 14 has the voltage application circuit 55 and the detection unit 56 of FIG. 3, and controls the voltage application between the counter electrodes 53 by the voltage application circuit 55, that is, controls the execution of electrostatic collection, and energizes the heater 54. The operation of the PM sensor 13 is controlled such as controlling. When the SCU 14 controls the heater 54, the SCU 14 adjusts the current (energization amount) flowing through the heater 54 and the energization time. Further, the SCU 14 detects the current flowing between the counter electrodes 53 or a value correlated therewith by the detection unit 56. The SCU 14 and the ECU 17 are connected by a communication line such as a CAN (Controller Area Network), and can communicate bidirectionally.

  The engine system 100 includes an ECU (Electronic Control Unit) 17 that controls the overall control of the engine system 100. The ECU 17 has a normal computer structure, and includes a CPU (not shown) that performs various operations and a memory 18 such as a ROM, RAM, and flash memory that stores various information. The ECU 17 detects operating conditions of the engine 20 based on detection signals from the various sensors, for example, calculates an optimal fuel injection amount, injection timing, injection pressure, etc. according to the operating conditions, and injects fuel into the engine 20. To control. Further, the ECU 17 executes a failure determination process for determining whether or not the DPF 12 has failed based on the detection value of the PM sensor 13. Details of this failure determination processing will be described later. The memory 18 stores various information such as a control program for processing executed by the ECU 17 (CPU) and related data shown in FIGS.

By the way, when PM coexists with a specific gas, the PM is burned and removed by the gas, and the output of the PM sensor 13 decreases. Here, FIG. 4 and FIG. 5 are collected by the PM sensor in the coexistence of NO, O ₂ and NO ₂ contained in the exhaust gas among the oxidizing gas (combustible gas) containing oxygen atoms. The experiment result that a part of the generated PM burns (oxidizes) and the output of the PM sensor decreases is shown. FIG. 4 shows the output reduction rate of the PM sensor in the presence of NO, O ₂ or NO ₂ . Explaining the experimental conditions in FIG. 4, when a PM sensor is installed in an engine exhaust gas environment and the output of the PM sensor reaches a predetermined value, the PM sensor is brought to an electric furnace. After waiting for PM sensor output to stabilize in an electric furnace maintained at 400 ° C., supply any one of NO, O ₂ , or NO ₂ as a combustion-supporting gas, and start supplying it The output reduction rate of the PM sensor after a predetermined time has elapsed is confirmed. This output decrease rate is a decrease rate with respect to the output value of the PM sensor at the start of gas supply. The concentration of the supply gas is 2000 ppm for NO, 20% for O ₂ , and 500 ppm for NO ₂ . As shown in FIG. 4, the sensor output decreases in any case of NO, O ₂ , and NO ₂ . Further, the rate of decrease in sensor output varies depending on the gas type.

FIG. 5 shows another experimental result different from FIG. 4, and CO ₂ generation corresponding to the combustion amount of PM collected by the PM sensor with respect to temperature in the presence of NO, O ₂ , or NO _2. Shows the change in quantity. Explaining the experimental conditions in FIG. 5, the PM sensor that has collected a certain amount of PM is removed from the actual machine and taken to the electric furnace. Then, NO in an electric furnace, O _2, or to supply any one gas of NO _2, and when the temperature of the electric furnace was changed in the range of 0 ° C. to 800 ° C., at each temperature The amount of CO ₂ production is detected. The CO ₂ generation amount, the feed gas (NO, _{O 2,} or NO ₂₎ and a CO ₂ generation amount generated by the PM is oxidized reaction, the feed gas of the PM trapped in the PM sensor This corresponds to the amount removed by combustion. The concentration of the supply gas is 1% for NO, 10% for O ₂ , and 1% for NO ₂ .

As shown in FIG. 5, the tendency of the change in the CO ₂ generation amount (PM combustion amount) with respect to the temperature differs depending on the gas type. Specifically, when the feed gas is NO, in the range of 0 ° C. to 400 ° C. it has become less CO ₂ production amount, is gently CO ₂ generation amount as the temperature increases in the range of above 400 ° C. It will increase.

When the supply gas is NO _2, the amount of CO ₂ generated is small in the range of 0 ° C. to 200 ° C., and the amount of CO ₂ generated gradually increases as the temperature increases from around 200 ° C. The CO ₂ generation amount takes a peak value at a temperature around 400 ° C., and the CO ₂ generation amount gradually decreases as the temperature increases in a temperature range higher than that. In addition, even around 800 ° C., the amount of CO ₂ produced slightly increases.

Also, when coexistence gas is _{O 2} has less CO ₂ production amount in the range of 0 ° C. to 400 ° C., rapidly CO ₂ generation amount is increased from around beyond 400 ° C., CO ₂ at 650 ° C. Atari When the peak value of the production amount is taken and the temperature at which the peak value is taken is exceeded, the CO ₂ production amount is rapidly reduced.

  In addition, the present inventor has confirmed that the higher the concentration of the combustion-supporting gas, the more the amount of PM removed by combustion and the larger the output decrease rate of the PM sensor. Furthermore, even when the same type of gas, concentration, and temperature coexist, the present inventor increases the amount of PM that is burned and removed as the coexistence time increases, and the output decrease rate of the PM sensor increases. I have confirmed that.

As a result of these experiments, a part of the PM collected by the PM sensor is burned and removed by the combustion-supporting gas coexisting with the PM, and the output of the PM sensor decreases. It can be seen that it varies depending on the type, concentration and temperature of the flammable gas. Based on this knowledge, the memory 18 of the ECU 17 stores the relationship data between the types, concentrations and temperatures of the combustion-supporting gas and the output correction value of the PM sensor shown in FIGS. The relation data may be stored in the memory of the SCU 14. The output correction value is a decrease in the sensor output O2 when a predetermined time has elapsed from the reference time with respect to the sensor output O1 at the reference time when the PM sensor in which PM is collected is placed in the presence of each combustion-supporting gas. It is a value corresponding to the ratio (= (O1-O2) / O1). In other words, the output correction value is a reduction rate of the sensor output when each gas coexists at each concentration and each temperature with respect to the sensor output when each gas of NO, NO ₂ and O ₂ does not coexist. Is a value corresponding to. Each of the output correction values in FIGS. 6 to 8 is a value smaller than 1.

FIG. 6 shows relationship data between the temperature and NO concentration and the output correction value A _NO when NO coexists. The decrease in the output of the PM sensor when the combustion-supporting gas coexists correlates with the amount of PM collected by the PM sensor burned and removed by the combustion-supporting gas. Therefore, the relationship between the output correction value A _NO and the temperature in FIG. 6 is the same as the relationship between the temperature and the amount of CO ₂ produced when NO coexists in FIG. Further, at the same temperature, the higher the NO concentration, the larger the output correction value _ANO .

FIG. 7 shows relationship data between the temperature and NO ₂ concentration and the output correction value A _NO2 when NO ₂ coexists. The relationship between the output correction value A _NO2 and the temperature in FIG. 7 is the same as the relationship between the temperature in FIG. 5 and the CO ₂ generation amount when NO ₂ coexists. Further, at the same temperature, the output correction value A _NO2 becomes larger as the NO ₂ concentration is higher.

FIG. 8 shows relationship data between the temperature and O ₂ concentration and the output correction value A _O2 when O ₂ coexists. The relationship between the output correction value A _O2 and the temperature in FIG. 8 is the same as the relationship between the temperature and the CO ₂ generation amount when O ₂ coexists in FIG. Also, the output correction value A _O2 as the O ₂ concentration is high at the same temperature has a large value. The relational data in FIGS. 6 to 8 can be obtained by the same method as the experiment in FIGS.

  Next, failure determination processing executed by the CPU of the ECU 17 will be described. FIG. 9 shows a flowchart of the failure determination process. In the process of FIG. 9, after the engine 20 is started, the heater 54 (FIG. 2, FIG. 2) is determined after it is determined whether or not the exhaust pipe has been dried to the extent that the PM sensor 13 (particularly the sensor element 52) is not wetted. 3) is started and the sensor regeneration for burning and removing the PM collected by the sensor element 52 is performed, and then the process is started. In the drying determination, for example, it is determined whether or not the temperature of the exhaust gas detected by the exhaust temperature sensor 21 is equal to or higher than a predetermined temperature (for example, 100 ° C.) at which the condensed water disappears due to evaporation.

  When the processing of FIG. 9 is started, the ECU 17 causes the SCU 14 to perform electrostatic collection of the PM sensor 13 (S1). Thereby, PM collection to the PM sensor 13 is started.

Next, the concentrations of NO, NO ₂ and O ₂ in the exhaust gas that has passed through the DPF 12 are detected (S2). Since the composition of the exhaust gas varies depending on the operating conditions of the engine 20, for example, the concentrations of NO, NO ₂ and O ₂ are estimated based on the operating conditions of the engine 20. Specifically, the relationship data between the engine operating conditions and the gas concentration downstream of the DPF is examined in advance and stored in the memory 18 for each gas type under the same conditions as the actual machine. Then, the current engine operating condition is acquired, and the gas concentration corresponding to the acquired operating condition is obtained from the related data. Specifically, the engine operating conditions include engine speed, fuel injection conditions (fuel injection amount, injection timing, etc.) into the cylinder, intake air amount, EGR rate (a ratio of EGR amount to intake air amount), and the like. The engine speed is obtained by the speed sensor 15. The fuel injection condition may be a command value set by the ECU 17 based on the engine speed and the detected value (engine load) of the accelerator pedal sensor 24. The intake air amount is obtained by the air flow meter 4. The EGR rate may be the target value set by the ECU 17 itself or the opening of the EGR valve 9 based on the engine speed, the engine load, and the like.

As shown in FIG. 1, a sensor 22 that detects the concentration of the coexisting gas may be provided around the PM sensor 13 downstream of the DPF 12, and the gas concentration may be acquired from the sensor 22. As the sensor 22, in particular it can be used and NOx sensor for detecting the NOx concentration, the _{O 2} sensor for detecting the _{O 2} concentration (e.g., air-fuel ratio sensor). Further, the concentration of some of NO, NO ₂ and O ₂ may be estimated based on the engine operating conditions, and the concentration of the remaining gas may be acquired from a sensor. At this time, when the NOx sensor cannot detect the NO concentration and the NO ₂ concentration separately, the NO concentration and the NO ₂ concentration may be estimated from the engine operating conditions, and the O ₂ concentration may be acquired from the O ₂ sensor. .

  Next, the temperature of the exhaust gas downstream of the DPF 12 is detected (S3). This temperature may be acquired from the exhaust temperature sensor 21 or may be estimated based on the operating conditions of the engine 20. When the gas temperature is estimated from the engine operating condition, the relationship data between the engine operating condition (engine speed, engine load, etc.) and the gas temperature is stored in the memory 18, and the gas temperature corresponding to the current engine operating condition is stored. Is obtained from the relational data. Further, the temperature of the sensor element 52 when heating by the heater 54 is not performed becomes a value similar to the temperature of the exhaust gas. Therefore, when the PM sensor 13 is provided with a temperature sensor that detects the temperature of the sensor element 52, the temperature of the exhaust gas may be estimated based on the detection value of the temperature sensor.

Next, the output E ₁ of the PM sensor 13 is detected by the SCU 14, and the output E ₁ is acquired from the SCU 14 (S4).

Then, S2, S3 detects the gas species, based on the gas concentration and gas temperature, the sensor output E ₁ detected in S4, the correction in a direction to increase the PM amount (S5). Specifically, the output correction value A is calculated based on the gas type, gas concentration, and gas temperature detected in S2 and S3 and the relationship data shown in FIGS. 6 to 8 show the output correction values A _NO , A _{NO 2} , and A _{O 2} for each gas type. In S 5, one of the output correction values A _NO , A _{NO 2} , and A _{O 2} is reflected. A total correction value A is obtained. For this purpose, for example, the memory 18 obtains the relationship data (A = f (gas species, concentration, etc.) between the gas type, gas concentration and gas temperature, and the total correction value A obtained based on the relationship data of FIGS. Temperature)) is stored, and an overall correction value A is obtained based on the relation data. The total correction value A can be, for example, a value obtained by adding the output correction values A _NO , A _NO2 , and A _O2 in FIGS.

Further, based on the relational data of FIGS. 6 to 8, first, output correction values A _NO , A _{NO 2} , A _{O 2} for each gas type are individually obtained, and these output correction values A _NO , A _{NO 2} , A _{O 2} are obtained. The total correction value A may be obtained by performing a predetermined calculation (for example, addition).

Total correction value A obtained is, NO, with respect to the sensor output when no coexist respective gases NO ₂ and _{O 2,} NO, if NO ₂ and _{O 2} coexist with the detected concentration and temperature at S2, S3 This is a value corresponding to the decrease rate of the sensor output. Therefore, in S5, with the total correction value A, to correct the sensor output E ₁ so as to offset this reduction ratio. Specifically, E _1R = (1 + A) × E ₁ is calculated with the corrected sensor output as E _1R .

  Here, FIG. 10 is a diagram showing the state of correction in S5, and shows sensor output with respect to the elapsed time (collection time) from the start of electrostatic collection before correction (thin solid line) and after correction (thick). It is the figure shown by the solid line. FIG. 10 shows three different sensor outputs (1), (2), and (3) before correction, and three sensor outputs after correction. Thus, by performing the correction of S5, the sensor output after the correction becomes a large value before the correction.

Then, on the basis of the sensor output _{E 1R} corrected, the failure determination of the DPF 12 (S6). Specifically, for example, the elapsed time from the start of the electrostatic collection is the sensor output E _1R corrected at a predetermined time to determine whether greater than a predetermined threshold value K. Further, for example, it may be determined whether or not the degree of change with time of the sensor output _1R after the generation of the output, that is, the inclination of the corrected sensor output line (thick solid line) in FIG. 11 is greater than a predetermined threshold value. .

When the sensor output _1R is larger than the threshold value K (S6: YES), it is determined that the DPF 12 is not functioning normally, that is, the DPF 12 is faulty, because the amount of PM flowing out from the DPF 12 is larger than expected (S7). On the other hand, when the sensor output _1R is equal to or less than the threshold value K (S6: NO), it is determined that the DPF 12 is normal because the amount of PM flowing out from the DPF 12 is small (S8). In the example of FIG. 10, for example, the corrected sensor outputs (1) and (2) are determined to be DPF failure, and the sensor output (3) is determined to be DPF normal. After S7 and S8, the process of FIG. 9 is terminated.

As described above, in this embodiment, the sensor output is corrected in the direction of increasing the PM amount according to the type, concentration, and temperature of the combustion-supporting gas (NO, NO ₂ , O ₂ ). The output can be brought close to the value before PM combustion by the coexisting gas is performed, and it can be suppressed that the detection accuracy of the PM amount is lowered due to the influence of the combustion-supporting gas. Further, since the DPF failure determination is performed based on the corrected sensor output, it is possible to suppress the failure determination accuracy from being deteriorated due to the influence of the combustion-supporting gas.

(Second Embodiment)
Next, a second embodiment of the present invention will be described focusing on differences from the above embodiment. The configuration of this embodiment is the same as the configuration of FIG. The failure determination process executed by the ECU 17 is different from the above embodiment.

  First, the basic concept of the failure determination process of this embodiment will be described with reference to FIG. FIG. 11 is a diagram showing changes in the output of the PM sensor 13 with respect to the time (collection time) after the start of electrostatic collection. Specifically, the one-dot chain line in FIG. 11 indicates the estimated output value Ee of the PM sensor 13 when the DPF 12 is a filter that serves as a reference for failure determination (hereinafter referred to as a reference failure filter). Lines (1), (2), and (3) show the actual output values of the PM sensor 13.

  In this embodiment, in order to determine the failure of the DPF 12, the output value Ee of the PM sensor 13 when the DPF 12 is a reference failure filter is estimated. Based on the comparison between the estimated output value Ee and the actual output value of the PM sensor 13, the presence or absence of a failure of the DPF 12 is determined. Specifically, the timing at which the estimated output value Ee reaches a predetermined threshold value K is determined as a failure determination timing, and if the actual sensor output value at the failure determination timing is greater than the threshold value K, it is determined that a DPF failure has occurred. It is determined that the DPF is normal. This means that if the actual sensor output value is larger than the estimated output value Ee, it is determined that the DPF has failed, and if it is smaller than the estimated output value Ee, it is determined that the DPF is normal. In the example of FIG. 11, when the actual output value is the line (1) or (2), it is determined that the DPF has failed, and when the actual output value is the line (3), it is determined that the DPF is normal.

  The failure determination method described in FIG. 11 is the same as the method described in Japanese Patent No. 5115873. That is, the failure determination method of this embodiment estimates the time (reference time) when the output of the PM sensor 13 rises when the DPF 12 is a reference failure filter (corresponding to the failure determination timing in FIG. 11). Then, when the time when the output of the PM sensor 13 actually rises (actual time) is earlier than the reference time, it is determined that the DPF 12 has failed, and in the latter case, it is determined that the DPF 12 is normal. means.

The memory 18 stores the relational data of FIGS. 6 to 8 as in the first embodiment. The output correction values A _NO , A _NO2 , and A _O2 of this embodiment are sensor output units. It is set to a value corresponding to a reduction rate (a reduction degree) per time Δt.

  Next, details of the failure determination process of the present embodiment will be described. FIG. 12 shows a flowchart of the failure determination process of this embodiment. The ECU 17 executes the process of FIG. 12 instead of the process of FIG. The process of FIG. 12 is started after the sensor regeneration is performed after the exhaust pipe drying determination is established, as in the process of FIG. 9.

When the process of FIG. 12 is started, first, an elapsed time t from the start of electrostatic collection is set to an initial value (= 0) (S11), and then the SCU 14 is instructed to perform electrostatic collection (S12). ). Next, the elapsed time t is updated to a time (t = t + Δt) that has progressed by a predetermined unit time Δt seconds (S13). Next, similarly to S2 and S3 of FIG. 9, the concentrations of NO, NO ₂ and O ₂ in the exhaust gas that has passed through the DPF 12 are detected (S14), and the temperature of the exhaust gas downstream of the DPF 12 is detected (S14). S15). The gas type, gas concentration, and gas temperature detected here are the gas type, gas concentration, and gas temperature in unit time Δt with reference to the current time t.

Next, based on the gas type, gas concentration, and gas temperature detected in S14 and S15, and the relationship data of FIGS. 6 to 8 stored in the memory 18, the environment of the detected gas type, gas concentration, and gas temperature is detected. A correction value A (Δt) corresponding to the output reduction degree per unit time Δt of the PM sensor 13 is calculated (S16). Here, as in the first embodiment, one total correction value A (= f (gas type, concentration, temperature)) reflecting all the output correction values A _NO , A _{NO 2} , A _{O 2} for each gas type is obtained. Ask. The total correction value A (Δt) can be, for example, a value obtained by adding the output correction values A _NO , A _NO2 , and A _O2 in FIGS.

Then, based on the operating conditions of the engine 20, DPF 12 is output change amount _{E 2} per unit of time Delta] t of the PM sensor 13 when the reference fault _DPF, estimates the _{Delta] t} (S17) (see also FIG. 11). Here, the reference failure DPF in the present embodiment specifically means that the collection rate of the DPF 12 is remarkably reduced due to the failure, and the amount of PM passing through the DPF 12 is determined by the self-failure diagnosis (OBD: On-board-diagnostics). The DPF is an amount equivalent to the regulation value. The OBD regulation value is set to a value larger than the EM regulation value (exhaust gas regulation value) such as EURO6. For example, in a specific travel mode, when the PM amount in the EM regulation value = 4.5 mg / km, the OBD regulation value is set to, for example, about 2.67 times the PM amount = 12.0 mg / km.

  Specifically, in S17, first, based on the operating condition of the engine 20, the amount f per unit time Δt is estimated based on the current time t of PM passing through the DPF 12 when the DPF 12 is the reference failure DPF. To do. Specifically, similarly to the method of Japanese Patent No. 5115873, the PM amount discharged from the engine 20 based on the operating conditions of the engine 20 such as the rotational speed and load (fuel injection amount) of the engine 20 and the like. In other words, the PM amount per unit time (inflow PM amount) flowing into the reference failure DPF is estimated. For example, a map of the inflow PM amount per unit time with respect to the operating conditions (rotation speed, load, etc.) of the engine 20 is stored in the memory 18 in advance. Then, the inflow PM amount corresponding to the current operating condition of the engine 20 may be read from the map.

  In addition, the PM collection rate of the reference failure DPF is estimated. Specifically, for example, a predetermined value α is used as the PM collection rate of the reference failure DPF. Further, since the PM collection rate of the DPF varies depending on the amount of PM deposited in the DPF (PM deposition amount) and the exhaust flow rate, the PM collection rate α depends on the PM deposition amount and the exhaust flow rate. May be corrected. The PM accumulation amount may be estimated based on, for example, the differential pressure across the DPF 12. The exhaust flow rate may be estimated based on the intake air amount detected by the air flow meter 4, for example.

  Based on the estimated inflow PM amount and the PM collection rate of the reference failure DPF, a PM amount f (outflow PM amount) per unit time flowing out from the reference failure DPF is obtained.

Next, the PM amount collected by the PM sensor 13 in the obtained outflow PM amount f is estimated. Specifically, for example, how much of the PM flowing outside the PM sensor 13 enters the cover 51 from the hole 511 (see FIG. 2), or how much PM of the PM that has entered the cover 51 The PM collection rate β to the PM sensor 13 is estimated in consideration of whether or not the sensor element 52 is attached to the sensor element 52. As the PM collection rate β, a constant predetermined value may be used regardless of various states such as the exhaust gas flow rate, λ (excess air ratio), the exhaust temperature, the temperature of the sensor element 52, and the like. A value corrected accordingly may be used. For example, the larger the exhaust gas flow rate, the more difficult it is for PM to enter the cover 51, and PM that has entered the cover 51 is less likely to adhere to the sensor element 52, and even if it adheres, it is likely to be detached from the sensor element 52. Further, the smaller λ is, that is, the richer the PM concentration becomes, the higher the ratio of PM not collected by the PM sensor 13 becomes. Therefore, for example, the PM collection rate β is estimated such that the larger the exhaust gas flow rate or the smaller λ, the smaller the value. In addition, since the permanent thermal power acting on the sensor element 52 changes according to the exhaust temperature or the temperature of the sensor element 52, the PM collection rate β changes. Then, the PM amount collected by the PM sensor 13 is obtained based on the outflow PM amount f and the PM collection rate β. Since the output of the PM sensor 13 increases as the PM amount increases, the relationship between the PM amount and the output of the PM sensor 13 is examined in advance and stored in the memory 18. Based on this relationship and the PM amount obtained this time, estimated values E _{2 and Δt} of the output change amount per unit time of the PM sensor 13 when the DPF 12 is the reference failure DPF are obtained.

Next, based on the correction value A (Δt) obtained in S16, the estimated values E2 _{and Δt} obtained in S17 are corrected in the direction of decreasing the PM amount (S18). Specifically, E _{2R, Δt} = (1−A (Δt)) × E _{2, Δt} is calculated by assuming that the corrected output change amount is E _{2R, Δt} .

Next, the output estimated value _{E2R, t} of the PM sensor 13 when the DPF 12 is the reference failure DPF at the current time t is obtained (S19). Specifically, the output estimate _{E 2R} obtained in the previous _S19, the _t, the estimated value _{E 2R} after correction calculated in this _S18, the addition of _{Delta] t,} and updates the output estimate _{E 2R,} a _t. That is, E _{2R, t} = E _{2R, t} + E _{2R, Δt} is calculated.

Next, it is determined whether or not the timing for determining the DPF failure (failure determination timing) has been reached based on the estimated output value _{E2R, t} obtained in S19 (S20). Specifically, it is determined whether or not the estimated output value E _{2R, t} is greater than a predetermined threshold value K. For example, the threshold value K is set to a value that can be regarded as the output of the PM sensor 13 rising. In S20, it is synonymous with determining whether or not the timing at which the output of the PM sensor 13 rises when the DPF 12 is the reference failure DPF has been reached.

If the failure determination timing has not yet been reached in S20, that is, if the output estimated value _{E2R, t} is equal to or less than the threshold value K (S20: NO), the process returns to S12. As described above, while the failure determination timing is not reached, the output change amount per unit time in each time of the PM sensor 13 is corrected based on the gas type, gas concentration, and gas temperature in each time, and the output change after correction is corrected. The estimated output value is corrected by integrating the amount over the elapsed time from the start of electrostatic collection.

As a result, as shown in FIG. 13, the estimated output value _E2R after the correction becomes a smaller value than before the correction, and as a result, the arrival of the failure determination timing is delayed compared with before the correction.

If the failure determination timing has reached in S20, that is, when the output estimation value _{E 2R, t} exceeds the threshold value K (S20: YES), acquires the actual output _{E 1} of the PM sensor 13 from SCU14 (S21) . As the sensor output E ₁ determines whether larger than the threshold K that is set to the same value as the threshold value K of S20 (S22). Then, when the sensor output _{E 1} is greater than the threshold value K (S22: YES), DPF12 determines that the PM trapping ability than a reference fault DPF is faulty DPF was reduced (S23). In contrast, if the sensor output _{E 1} is less than the threshold value K (S22: NO), DPF12 determines that the PM trapping ability than a reference fault DPF is a good normal DPF (S24). In the example of FIG. 13, at the failure determination timing (after correction), in the case of (1) and (2), the sensor output is larger than the threshold value K, so it is determined that the DPF has failed, and in the case of (3), the sensor output is the threshold value K. Since it is smaller, it is determined that the DPF is normal. After S23 and S24, the process of FIG.

In this way, in this embodiment, the sensor output in the case of the reference failure DPF is estimated, and the failure determination of the DPF is performed based on the comparison between the estimated value and the actual sensor output. Whether it is high or low can be accurately determined. Moreover, since the estimated value of the sensor output is corrected in the direction of decreasing the PM amount according to the type, concentration and temperature of the combustion-supporting gas (NO, NO ₂ , O ₂ ), the estimated value is corrected to the combustion-supporting property It is possible to approximate the value after PM combustion with gas, and the actual output value of the sensor can be compared with the estimated value in the state after combustion. As a result, the failure determination of the DPF can be performed in a manner that suppresses the influence of combustion due to the combustion-supporting gas, and the deterioration in accuracy of the failure determination can be suppressed. For example, in the example of FIG. 13, when the sensor output is (2), when carrying out the failure determination at the failure determination timing based on the estimated output value E ₂ before correction (pre-correction), the sensor output and the threshold value K is There is a possibility that it becomes a close value and it is erroneously determined that it is normal despite the DPF failure. On the other hand, when the failure determination is performed at the failure determination timing (after correction) based on the corrected estimated output value _E2R , the sensor output is clearly larger than the threshold value K, so that the DPF failure can be determined with high accuracy.

  Furthermore, as correction of the estimated value of the sensor output, each time correction based on the concentration and temperature for each gas type at each time is performed over the elapsed time from the start of electrostatic collection, so A highly accurate estimated value reflecting the gas type, gas concentration, and gas temperature at each time and the elapsed time from the start of collection can be obtained. Accordingly, it is possible to further suppress a decrease in accuracy of the DPF failure determination.

(Third embodiment)
Next, a third embodiment of the present invention will be described focusing on differences from the above embodiment. The configuration of this embodiment is the same as the configuration of FIG. The failure determination process executed by the ECU 17 is different from the above embodiment. In the second embodiment, the failure determination timing is corrected to a later timing by correcting the estimated output value of the PM sensor, but in this embodiment, by correcting the threshold value to be compared with the estimated output value, As a result, the failure determination timing is corrected to a later timing.

  Also in the present embodiment, as in the second embodiment, the output correction value A in the relational data of FIGS. 6 to 8 stored in the memory 18 corresponds to the reduction rate of the sensor output per unit time Δt seconds. Is set to a value.

  ECU17 performs the process of FIG. 14 as a failure determination process of DPF. The processing in FIG. 14 is started after the sensor regeneration is performed after the exhaust pipe drying determination is established as in the above embodiment. When the processing of FIG. 14 is started, processing similar to the processing of S11 to S16 of FIG. 12 is executed (S31 to S36).

  Next, an integrated correction value A (t) obtained by integrating the correction value A (Δt) corresponding to the output decrease degree per unit time Δt of the PM sensor 13 obtained in the process of S36 for each time up to the current time t. Is calculated (S37). Specifically, the correction value A (t) obtained in S36 of this time is added to the correction value A (t) obtained in the previous S37 to update the integration correction value A (t). That is, A (t) = A (t) + A (Δt) is calculated.

Then, based on the operating conditions of the engine 20, estimates the output _{E 2} of the PM sensor 13 when DPF12 is the reference fault DPF (S38). In S17 in FIG. 12, the output change amount _{E 2} per unit of time Delta] t of the PM sensor 13 _to estimate the _{Delta] t.} Here, as in S17, the output change amounts E2 _{and Δt} are obtained at each time, and the integrated value obtained by integrating the obtained output change amounts E2 _{and Δt} over the current time t from the start of electrostatic collection is obtained as the current value. It estimated as the output _{E 2} of the PM sensor 13 at time t.

Next, based on the integrated correction value A (t) obtained in S37, the threshold value K (corresponding to the first threshold value of the present invention) for determining the failure determination timing is corrected in the direction of increasing the PM amount ( This corresponds to the first threshold correction of the present invention) (S39). Specifically, K _R = (1 + A (t)) × K is calculated with the corrected threshold as K _R.

Next, the estimated output _{E 2} obtained in S38, it is determined whether the threshold value _{K R} is greater than or not the corrected calculated in S39 (S40). Estimated output _{E 2} is the threshold value _K in the case of _R below (S40: NO), as yet not reached the failure determination timing, the flow returns to S32. Thus, while it does not reach the failure determination timing, gradually cumulative correction value A (t) increases with time, as a result, the threshold K _R gradually becomes larger. Also, the estimated value of the sensor output E ₂ also gradually increases with the passage of time.

If the estimated output _{E 2} exceeds the threshold value _{K R} in S40 (S40: YES), the failure determination timing has reached, and acquires the actual output _{E 1} of the PM sensor 13 from SCU14 (S41). Next, the sensor output _{E 1} determines whether greater than a predetermined threshold value K (S42). This threshold value K is set to the same value as the threshold value K for determining the failure determination timing before the correction of S39 is performed. If the sensor output _{E 1} is greater than the threshold value K (S42: YES) determines DPF failure and (S43), in the case of less than the threshold value K (S42: NO) DPF normal and determining (S44). After S43 and S44, the process of FIG.

The failure determination process of the present embodiment will be described with reference to FIG. 15. Based on the type, concentration, and temperature of the combustion-supporting gas, the threshold value K for determining the failure determination timing is corrected so as to increase the PM amount. The timing at which the estimated output value Ee exceeds the corrected threshold value K _R (failure diagnosis execution determination threshold value) is determined as the failure determination timing. Therefore, the failure determination timing can be set later than the failure determination timing when the threshold value K is not corrected. By delaying the failure determination timing, it is possible to increase the sensor output at the failure determination timing, and it is possible to suppress the decrease in the sensor output due to the influence of the combustion-supporting gas from affecting the DPF failure determination. For example, in the example of FIG. 15, when the sensor output is (2), if the failure determination is performed at the failure determination timing (before correction) based on the threshold value K before correction, the sensor output and the threshold value K are close to each other. Thus, there is a possibility that it is erroneously determined to be normal despite the DPF failure. On the other hand, when carrying out the failure determination at the failure determination timing based on the threshold K _R after correction (corrected), the sensor output is greater than the obviously threshold K, it can determine DPF failure with high accuracy.

  Further, as correction of the threshold value K, a correction value per unit time based on the concentration and temperature for each gas type at each time is calculated based on an integrated correction value obtained by integrating the elapsed time from the start of electrostatic collection. Since the threshold is corrected, the gas type, gas concentration, and gas temperature at each time and the elapsed time from the start of collection can be reflected in the corrected threshold. By determining the failure determination timing based on the corrected threshold value and performing the failure determination at this failure determination timing, the failure determination includes the gas type, gas concentration, and gas temperature at each time, and the start of collection. The elapsed time can be reflected. Therefore, a highly accurate determination result that suppresses the influence of the combustion-supporting gas can be obtained.

(Fourth embodiment)
Next, a fourth embodiment of the present invention will be described focusing on the differences from the above embodiment. The configuration of this embodiment is the same as the configuration of FIG. The failure determination process executed by the ECU 17 is different from the above embodiment.

The ECU 17 executes the process of FIG. 16 as the DPF failure determination process. The process of FIG. 16 is started after the sensor regeneration is performed after the exhaust pipe drying determination is established as in the above embodiment. Comparing the process of FIG. 16 with the process of FIG. 14, the process of FIG. 16 omits the process corresponding to the process of S39 of FIG. 14, and the process of FIG. 14 is added in that the process of S61 is added. Is different. Further, in the processing of S59, the output estimate E ₂ are compared with the threshold value K of the correction unexecuted. Further, in the processing of S62 is compared with the threshold value K _R and the corrected sensor output E _1. Otherwise, the process is the same as in FIG. That is, the processing of S51 to S58 in FIG. 16 is the same as the processing of S31 to S38 in FIG. 14, the processing of S60 is the same as the processing of S41, and the processing of S63 and S64 is the same as the processing of S43 and S44. It is.

When the process is started in FIG. 16, similarly to the third embodiment calculates the cumulative correction value A (S51~S57), estimates the output _{E 2} of the PM sensor 13 in the case DPF12 is a reference fault DPF (S58), the estimated output value E ₂ determines whether exceeds a threshold value K (S59). If not exceeded (S59: NO), it returns to S52, and updates the cumulative correction value A and the estimated output value _{E 2 (S52~S58).}

If the estimated output value _{E 2} exceeds the threshold value K, that is, when it reaches the failure determination timing (S59: YES), then obtains the actual output _{E 1} of the PM sensor 13 from SCU14 (S60). Next, based on the integrated correction value A (t) obtained in S57, the failure determination threshold value K (corresponding to the second threshold value of the present invention) is corrected in the direction of decreasing the PM amount (the second threshold value of the present invention). Equivalent to correction) (S61). Specifically, K _R = (1−A (t)) × K is calculated with the corrected threshold as K _R.

Next, the sensor output _{E 1} to determine threshold _K or _R is larger than the corrected (S62). If the sensor output _{E 1} is greater than the threshold value _{K R} (S62: YES) determines DPF failure and (S63), in the case of less than the threshold value _{K R} (S62: NO) DPF normal and determining (S64). Thereafter, the process of FIG. 16 is terminated.

In the third embodiment, the threshold value for determining the failure determination timing (threshold value to be compared with the estimated output value Ee) is corrected, while in the present embodiment, as shown in FIG. 17, it is compared with the actual sensor output. The failure determination threshold value K is corrected so as to be a value that is smaller by the decrease in the output of the PM sensor due to the influence of the combustion-supporting gas. As a result, it is possible to compare the sensor output and the failure determination threshold value in a form that offsets the decrease in the output, and it is possible to suppress a decrease in accuracy of the DPF failure determination. For example, in the example of FIG. 17, when the sensor output is (2), if the failure determination is performed with the threshold value K before correction, the sensor output and the threshold value K are close to each other. Regardless, there is a possibility of misjudging that it is normal. On the other hand, when carrying out the failure determination at the threshold K _R after the correction, the sensor output is greater than the obviously threshold K _R, it can determine DPF failure with high accuracy.

(Fifth embodiment)
Next, a fifth embodiment of the present invention will be described focusing on the differences from the above embodiment. The configuration of this embodiment is the same as the configuration of FIG. The failure determination process executed by the ECU 17 is different from the above embodiment.

  The ECU 17 executes the process of FIG. 18 as the DPF failure determination process. The processing in FIG. 18 is started after the sensor regeneration is performed after the exhaust pipe drying determination is established as in the above embodiment. Comparing the processing of FIG. 18 with the processing of FIG. 16, the processing of FIG. 18 differs from the processing of FIG. 16 in the processing of S81 and S82, and the other processing (S71 to S80, S83, S84) is the processing of FIG. The same as (S51 to S60, S63, S64).

In the process of FIG. 18, when the failure determination timing has reached (S79: YES), after detecting the actual sensor output _{E 1} (S80), cumulative correction value A (t found the sensor output _{E 1} in S77 ) To increase the PM amount (S81). Specifically, E _1R = (1 + A (t)) × E ₁ is calculated with the corrected sensor output as E _1R . Next, it is determined whether or not the corrected sensor output _E1R is larger than the threshold value K (S82). If it is larger (S82: YES), it is determined that the DPF has failed (S83). (S82: NO) It is determined that the DPF is normal (S84). Thereafter, the process of FIG. 18 is terminated.

  Thus, in this embodiment, since the sensor output is corrected in the same manner as in the first embodiment (see FIG. 19), the sensor output can be brought close to the value before PM combustion by the combustion-supporting gas, and before PM combustion. In this state, the sensor output and the failure determination threshold can be compared. Thereby, the failure determination can be performed in a form in which the influence of PM combustion by the combustion-supporting gas is suppressed, and a decrease in the accuracy of the failure determination can be suppressed. FIG. 19 shows a state in which only the sensor output of (3) is corrected.

(Other embodiments)
In addition, this invention is not limited to the said embodiment, A various change is possible to the limit which does not deviate from description of a claim. For example, in the above embodiment, as the combustion-supporting gas NO, have been carried out correction of the sensor output, such as on the basis of each concentration of NO ₂ and O _2, NO, any one of NO ₂ and O ₂ or two Correction may be performed based on the gas concentration. This also makes it possible to perform DPF failure determination in a manner that suppresses the influence of the combustion-supporting gas, as compared with the case where correction is not performed, and to suppress deterioration in accuracy of failure determination. Further, the correction may be performed in consideration of the concentration of a combustion-supporting gas (for example, sulfur oxide (SO ₂ or the like)) other than NO, NO ₂ and O ₂ .

  Moreover, in the said embodiment, although PM sensor was used for the use of the failure detection of DPF, you may use PM sensor for uses other than failure detection. For example, a PM sensor may be disposed upstream of the DPF, and this PM sensor may be used for detecting the amount of PM discharged from the engine (the amount of PM flowing into the DPF). At this time, by correcting the sensor output according to the present invention, the amount of PM before PM combustion by the combustion-supporting gas can be detected with high accuracy.

  In the above embodiment, the ECU executes the failure determination processing. However, the SCU may execute the failure determination processing, and some processing of the failure determination processing is executed by the ECU. This process may be executed by the SCU.

  In the above embodiment, the PM sensor 13, the SCU 14, and the ECU 17 correspond to the filter failure detection device and the particulate matter detection device of the present invention. The ECU 17 that executes the processes of FIGS. 9, 12, 14, 16, and 18 performs concentration acquisition means, temperature acquisition means, correction means, failure determination means, estimation means, timing determination means, calculation means, and integration according to the present invention. This corresponds to the correction means.

12 DPF (filter)
13 PM sensor 14 SCU
17 ECU
20 engine (internal combustion engine)
23 Exhaust passage 52 Sensor element (attachment)

Claims (10)

A portion to be attached (adhered to the particulate matter in the exhaust gas) provided downstream of the filter (12) for collecting particulate matter in the exhaust gas provided in the exhaust passage (23) of the internal combustion engine (20). 52), and a sensor (13) that outputs a value corresponding to the amount of particulate matter adhered to the adherend portion;
Concentration acquisition means (S2, S14, S34, S54, S74, 17) for acquiring the concentration of the combustion-supporting gas coexisting with the particulate matter in the exhaust gas;
Temperature acquisition means (S3, S15, S35, S55, S75, 17) for acquiring the temperature of the exhaust gas;
Correction means (S5, S76, S77) for correcting the output value of the sensor in the direction of increasing the amount of particulate matter based on the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means. , S81, 17),
Failure determination means (S6 to S8, S82 to S84, 17) for determining failure of the filter based on the output value corrected by the correction means;
A failure detection device for a filter, comprising: A portion to be attached (adhered to the particulate matter in the exhaust gas) provided downstream of the filter (12) for collecting particulate matter in the exhaust gas provided in the exhaust passage (23) of the internal combustion engine (20). 52), and a sensor (13) that outputs a value corresponding to the amount of particulate matter adhered to the adherend portion;
Estimating means (S17, S38, S58, S78, 17) for estimating the output value of the sensor when the filter is a filter serving as a criterion for failure determination;
Failure determination means (S22 to S24, S42 to S44, S62 to S64, S82 to S84, 17) for determining a failure of the filter based on a comparison between the value estimated by the estimation means and the actual output value of the sensor. When,
Concentration acquisition means (S14, S34, S54, S74, 17) for acquiring the concentration of the combustion-supporting gas coexisting with the particulate matter in the exhaust gas;
Temperature acquisition means (S15, S35, S55, S75, 17) for acquiring the temperature of the exhaust gas;
Based on the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means, an output value correction for correcting the output value of the sensor in a direction to increase the amount of particulate matter, and the estimation means Correction means (S16, S18, S19, S76, S77, S81, 17) for performing any one of the estimated value correction for correcting the estimated value in the direction of reducing the amount of particulate matter,
The filter failure detection device, wherein the failure determination unit performs failure determination of the filter using a value corrected by the correction unit. Timing determination means (S20, S40, S59, S79, 17) for determining the timing at which the estimated value of the estimation means reaches a predetermined first threshold;
The failure determination means performs a failure determination of the filter based on a comparison between an output value of the sensor at the timing determined by the timing determination means and a second threshold having the same value as the first threshold,
The correction means (S16, S18, S19, S36, S37, S39, S56, S57, S61, S76, S77, S81, 17) include the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means. Based on the output value correction, the estimated value correction, the first threshold value correction for correcting the first threshold value in the direction of increasing the amount of the particulate matter, and the second threshold value of the particulate matter. The filter failure detection device according to claim 2, wherein the filter failure detection device performs any one of second threshold correction that corrects the amount in a direction of decreasing the amount. 4. The filter failure detection apparatus according to claim 1, wherein the concentration acquisition unit acquires at least one concentration of NO, NO _2, and O ₂ . The concentration acquisition means (S14, S34, S54, S74, 17) and the temperature acquisition means (S15, S35, S55, S75, 17) acquire the concentration and the temperature at each time,
The correction means (S16, S18, S19, S36, S37, S39, S56, S57, S61, S76, S77, S81, 17) are the concentration and temperature at each time, and the particles on the adherend. The filter failure detection device according to any one of claims 1 to 4, wherein correction is performed that reflects an elapsed time from the start of collection of the particulate matter. The correction means includes
Based on the concentration and the temperature, calculation means (S16) calculates a correction value corresponding to the degree of change per unit time of the output value of the sensor due to combustion of particulate matter by the combustion-supporting gas at each time. , S36, S56, S76, 17),
Based on the correction value, the amount of change per unit time of the correction target at each time is corrected, and the amount of change after the correction is integrated over the elapsed time from the start of collection on the adherend, or Integration correction means (S18, S19, S37, S39, S57, S61, S77, S81, 17) that performs integration correction for correcting the correction target value based on the integration correction value obtained by integrating the correction value over the elapsed time. The filter failure detection apparatus according to claim 5, further comprising: The concentration acquisition means acquires the concentration by specifying the type of the combustion-supporting gas,
The correction means obtains relational data between the type of the combustion-supporting gas, the concentration and the temperature, and a correction value corresponding to the degree of change in the output value of the sensor due to combustion of particulate matter by the combustion-supporting gas. The filter failure detection device according to claim 1, wherein correction is performed based on the relationship data.   The said concentration acquisition means has the relationship data of the driving | running condition of the said internal combustion engine, and the density | concentration of combustion-supporting gas, The said concentration is acquired based on the relationship data. The filter failure detection device according to any one of the above.   The filter failure detection according to any one of claims 1 to 8, wherein the concentration acquisition means acquires the concentration from a sensor (22) that detects the concentration provided in the exhaust passage. apparatus. It is provided in the exhaust passage (23) of the internal combustion engine (20) and has an adherent part (52) for adhering particulate matter in the exhaust gas, and according to the amount of particulate matter adhering to the adherent part. A sensor (13) for outputting a value;
Concentration acquisition means (S2, S14, S34, S54, S74, 17) for acquiring the concentration of the combustion-supporting gas coexisting with the particulate matter in the exhaust gas;
Temperature acquisition means (S3, S15, S35, S55, S75, 17) for acquiring the temperature of the exhaust gas;
Correction means (S5, S76, S77) for correcting the output value of the sensor in the direction of increasing the amount of particulate matter based on the concentration acquired by the concentration acquisition means and the temperature acquired by the temperature acquisition means. , S81, 17),
A particulate matter detection device comprising:

Images