Catalyst abnormality detection device

Abstract

To provide a catalyst abnormality detection device capable of promptly detecting an abnormality of an exhaust catalyst regardless of a driving state of a vehicle.SOLUTION: A catalyst abnormality detection device 20 includes: an input state amount detection section 23 detecting the state amount of exhaust gas flowing on an upstream side of an SCR 7 in an exhaust passage 4 and acquiring an input state amount actual measurement value of the exhaust gas; an output state amount detection section 24 detecting the state amount of exhaust gas flowing on a downstream side of the SCR 7 in the exhaust passage 4 and acquiring an output state amount actual measurement value of the exhaust gas; an output state amount prediction section 27 predicting the state amount of exhaust gas flowing on the downstream side of the SCR 7 in the exhaust passage 4 by using the input state amount actual measurement value of the exhaust gas and acquiring an output state amount prediction value of the exhaust gas; and an NOx elimination ratio calculation section 28 and an abnormality determination section 29 determining whether or not the SCR 7 is abnormal on the basis of the output state amount actual measurement value of the exhaust gas and the output state amount prediction value of the exhaust gas.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a catalyst abnormality detection device. [Background technology]

[0002] Known as a conventional catalyst abnormality detection device is, for example, the technology described in Patent Document 1. The catalyst abnormality detection device described in Patent Document 1 calculates the NOx emission amount based on sample values ​​of the engine speed and torque when monitoring conditions for detecting the engine operating state are met and the engine is in a steady operating state, calculates the NOx purification rate from this NOx emission amount and the NOx concentration detected by the NOX sensor, and determines that an abnormality has occurred when the NOx purification rate is smaller than a threshold value. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2013-181427 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the above-mentioned conventional technology, an abnormality is not judged unless all predetermined conditions for stabilizing the catalyst purification rate are met. During a transient state in which the vehicle's operating state is more unstable than during a steady state, the conditions for stabilizing the catalyst purification rate are less likely to be met. For this reason, during a transient state, a time lag occurs between when an abnormality actually occurs in the exhaust catalyst and when it is judged to be abnormal.

[0005] An object of the present invention is to provide a catalyst abnormality detection device that can quickly detect an abnormality in an exhaust catalyst regardless of the driving state of a vehicle. [Means for solving the problem]

[0006] One aspect of the present invention is a catalyst abnormality detection device that detects an abnormality in an exhaust catalyst arranged in an exhaust passage connected to an internal combustion engine of a vehicle, and includes an input state quantity detection unit that detects a state quantity of exhaust gas flowing upstream of the exhaust catalyst in the exhaust passage and obtains an actual measured value of an input state quantity of the exhaust gas, an output state quantity detection unit that detects a state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage and obtains an actual measured value of an output state quantity of the exhaust gas, an output state quantity prediction unit that predicts the state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage using the actual measured value of the input state quantity of the exhaust gas obtained by the input state quantity detection unit and obtains a predicted value of the output state quantity of the exhaust gas, and a judgment unit that judges whether the exhaust catalyst is abnormal based on the actual measured value of the output state quantity of the exhaust gas obtained by the output state quantity detection unit and the predicted value of the output state quantity of the exhaust gas obtained by the output state quantity prediction unit.

[0007] In such a catalyst abnormality detection device, the state quantities of the exhaust gas flowing upstream and downstream of the exhaust catalyst in the exhaust passage are detected, and an actual measured value of the input state quantity and an actual measured value of the output state quantity of the exhaust gas are obtained. Then, the actual measured value of the input state quantity of the exhaust gas is used to predict the state quantity of the exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage, and a predicted value of the output state quantity of the exhaust gas is obtained. Then, based on the actual measured value of the output state quantity of the exhaust gas and the predicted value of the output state quantity of the exhaust gas, it is determined whether the exhaust catalyst is abnormal. Here, by predicting the state quantity of the exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage so as to include not only the steady state of the vehicle but also the transient state of the vehicle, it is possible to determine whether the exhaust catalyst is abnormal even before the operating state of the vehicle is sufficiently stabilized. As a result, an abnormality in the exhaust catalyst is quickly detected regardless of the operating state of the vehicle.

[0008] The determination unit may calculate a deterioration index serving as an indicator of exhaust catalyst deterioration based on the actual measured value of the output state quantity of the exhaust gas and the predicted value of the output state quantity of the exhaust gas, and determine that the exhaust catalyst is abnormal when the deterioration index is equal to or greater than a threshold value. In this configuration, by calculating the deterioration index serving as an indicator of exhaust catalyst deterioration, it is possible to realize an abnormality determination of the exhaust catalyst with simple processing and high accuracy.

[0009] The determination unit may calculate an actual purification rate of harmful substances contained in the exhaust gas based on the actual measured input state quantity of the exhaust gas and the actual measured output state quantity of the exhaust gas, and may calculate a predicted purification rate of harmful substances based on the actual measured input state quantity of the exhaust gas and the predicted output state quantity of the exhaust gas, and may compare the actual purification rate with the predicted purification rate to calculate a deterioration index. In this configuration, by calculating the actual purification rate and the predicted purification rate of harmful substances, a deterioration index that is used as a guide to the deterioration of the exhaust catalyst can be easily obtained. Also, before the actual purification rate of harmful substances becomes stable, it is determined whether the exhaust catalyst is abnormal.

[0010] The catalyst abnormality detection device may further include a learning data determination unit that determines learning data used to predict the state quantities of the exhaust gas based on the actual input state quantity values ​​of the exhaust gas acquired by the input state quantity detection unit and the actual output state quantity values ​​of the exhaust gas acquired by the output state quantity detection unit during a specified period after the vehicle has been used, and the output state quantity prediction unit may predict the state quantities of the exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage based on the actual input state quantity values ​​of the exhaust gas and the learning data determined by the learning data determination unit. In this configuration, the learning data used to predict the state quantities of the exhaust gas are determined when the internal combustion engine, exhaust catalyst, etc. are brand new, so that more accurate output state quantity prediction values ​​can be obtained. Therefore, the accuracy of detecting an abnormality in the exhaust catalyst is improved. Effect of the Invention

[0011] According to the present invention, an abnormality in an exhaust catalyst can be detected promptly regardless of the driving state of a vehicle. [Brief description of the drawings]

[0012] [Figure 1] 1 is a schematic diagram showing an exhaust purification system equipped with a catalyst abnormality detection device according to an embodiment of the present invention; [Diagram 2] 1 is a block diagram showing a configuration of a catalyst abnormality detection device according to an embodiment of the present invention; [Diagram 3] 3 is a conceptual diagram showing how the state quantity of the exhaust gas is predicted using machine learning in the output state quantity prediction unit shown in FIG. 2, together with the input state quantity actual measurement values ​​and output state quantity actual measurement values ​​of the exhaust gas. FIG. [Figure 4] 3 is a flowchart showing details of a procedure of an abnormality determination process executed by an abnormality determination unit shown in FIG. 2. [Diagram 5] 3 is a timing diagram showing an operation of detecting an abnormality in an SCR by the catalyst abnormality detection device shown in FIG. 2. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0014] Fig. 1 is a schematic diagram showing an exhaust purification system equipped with a catalyst abnormality detection device according to one embodiment of the present invention. In Fig. 1, the exhaust purification system 1 is mounted on a vehicle 2 such as an automobile.

[0015] The exhaust purification system 1 includes an exhaust passage 4 connected to an engine 3, which is an internal combustion engine, and a DOC 5 (Diesel Oxidation Catalyst), a DPF 6 (Diesel Particulate Filter), and an SCR 7 (Selective Catalytic Reduction catalyst) arranged in this exhaust passage 4 in order from upstream to downstream.

[0016] The exhaust passage 4 is a passage through which exhaust gas generated inside the engine 3 flows. The DOC 5 oxidizes and purifies carbon monoxide (CO), hydrocarbons (HC), and other substances contained in the exhaust gas. The DPF 6 collects particulate matter (PM) contained in the exhaust gas. The SCR 7 reduces and purifies nitrogen oxides (NOx) contained in the exhaust gas using urea water. The DOC 5, DPF 6, and SCR 7 are exhaust catalysts that remove harmful substances contained in the exhaust gas.

[0017] The exhaust purification system 1 also includes a fuel addition valve 8 disposed between the engine 3 and the DOC 5, and a urea water addition valve 9 disposed between the DPF 6 and the SCR 7. The fuel addition valve 8 injects liquid fuel toward the DOC 5 to raise the temperature of the exhaust gas when regenerating the DPF 6. Regeneration of the DPF 6 means burning PM accumulated in the DPF 6. The urea water addition valve 9 injects urea water (described above) toward the SCR 7.

[0018] The exhaust purification system 1 also includes temperature sensors 11 and 12, a differential pressure sensor 13, a flow rate sensor 14, a temperature sensor 15, a NOx sensor 16, a temperature sensor 17, and a NOx sensor 18.

[0019] The temperature sensor 11 detects the temperature of exhaust gas flowing upstream of the DOC 5 in the exhaust passage 4. The temperature sensor 12 detects the temperature of exhaust gas flowing downstream of the DOC 5 and upstream of the DPF 6 in the exhaust passage 4.

[0020] The differential pressure sensor 13 detects the pressure difference (differential pressure) between the upstream side and the downstream side of the DPF 6. That is, the differential pressure sensor 13 detects the difference between the pressure upstream of the DPF 6 and the pressure downstream of the DPF 6 in the exhaust passage 4.

[0021] The flow rate sensor 14 detects the flow rate of exhaust gas flowing downstream of the DPF 6 and upstream of the SCR 7 in the exhaust passage 4. The temperature sensor 15 detects the temperature of exhaust gas flowing downstream of the DPF 6 and upstream of the SCR 7 in the exhaust passage 4. The NOx sensor 16 detects the concentration of NOx contained in exhaust gas flowing downstream of the DPF 6 and upstream of the SCR 7 in the exhaust passage 4.

[0022] The temperature sensor 17 detects the temperature of the exhaust gas flowing downstream of the SCR in the exhaust passage 4. The NOx sensor detects the concentration of NOx contained in the exhaust gas flowing downstream of the SCR in the exhaust passage 4.

[0023] 2 is a block diagram showing the configuration of a catalyst abnormality detection device according to one embodiment of the present invention. In FIG. 2, a catalyst abnormality detection device 20 of this embodiment is a device that detects an abnormality in the SCR 7.

[0024] The catalyst abnormality detection device 20 includes the flow rate sensor 14, the temperature sensor 15, the NOx sensor 16, the temperature sensor 17, and the NOx sensor 18, a controller 21, the urea water addition valve 9, and an alarm device 22.

[0025] The flow rate sensor 14, the temperature sensor 15, and the NOx sensor 16 constitute an input state quantity detection unit 23 that detects state quantities of the exhaust gas flowing upstream of the SCR 7 in the exhaust passage 4 and acquires actual measured values ​​of input state quantities of the exhaust gas. As shown in Fig. 3(a), the actual measured values ​​of the input state quantities of the exhaust gas are the actual measured input exhaust gas flow rate, the actual measured input exhaust gas temperature, and the actual measured input NOx concentration.

[0026] The temperature sensor 17 and the NOx sensor 18 constitute an output state quantity detection unit 24 that detects state quantities of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4 and acquires actual measured values ​​of the output state quantities of the exhaust gas. The actual measured values ​​of the output state quantities of the exhaust gas are the actual measured output exhaust gas temperature value and the actual measured output NOx concentration value, as shown in Fig. 3(a).

[0027] When an abnormality occurs in the SCR 7, the alarm device 22 issues an abnormality warning by a warning display, a warning sound, or the like.

[0028] The controller 21 has a learning data determination unit 25 , an addition valve control unit 26 , an output state quantity prediction unit 27 , a NOx purification rate calculation unit 28 , and an abnormality determination unit 29 .

[0029] The learning data determination unit 25 determines learning data to be used for predicting the state quantities of the exhaust gas based on the actual measured values ​​of the input state quantities of the exhaust gas acquired by the input state quantity detection unit 23 and the actual measured values ​​of the output state quantities of the exhaust gas acquired by the output state quantity detection unit 24 during a specified period after the use of the vehicle 2 has begun.

[0030] The learning data determination unit 25 determines learning data for a period from when a brand new vehicle 2 is shipped until the vehicle 2 travels a preset distance (for example, 100 km). The learning data includes not only data in a steady state in which the driving state of the vehicle 2 is stable, but also data in a transient state in which the driving state of the vehicle 2 becomes unstable. A transient state is a state in which the state quantities of the exhaust gas (flow rate, temperature, NOx concentration, etc.) fluctuate (see Figs. 5(a) to 5(c)). The learning data also includes individual differences of the engine 3 and the SCR 7.

[0031] The addition valve control section 26 controls the urea water addition valve 9 so as to change the already set injection amount of urea water based on, for example, the input state quantity actual measurement value and the output state quantity actual measurement value of the exhaust gas.

[0032] The output state quantity prediction unit 27 predicts the state quantity of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4 using the input state quantity actual measurement value of the exhaust gas, and acquires an output state quantity predicted value of the exhaust gas. The output state quantity prediction unit 27 predicts the state quantity of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4, based on the input state quantity actual measurement value of the exhaust gas and the learning data determined by the learning data determination unit 25.

[0033] As shown in FIG. 3(b), the output state quantity prediction unit 27 predicts the state quantities of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4 by using machine learning. Machine learning is one type of AI (artificial intelligence). For example, a neural network (NN) or the like is used as the machine learning.

[0034] The predicted values ​​of the output state quantities of the exhaust gas include a predicted output exhaust gas temperature value and a predicted output NOx concentration value. The predicted values ​​of the output state quantities of the exhaust gas include a stable state and a transient state as the operating state of the vehicle 2.

[0035] The NOx purification rate calculation unit 28 calculates an actual purification rate of NOx contained in exhaust gas based on the actual input exhaust gas flow rate, the actual input NOx concentration, and the actual output NOx concentration. The actual NOx purification rate is an actual purification rate of harmful substances contained in exhaust gas. The actual NOx purification rate is calculated by the following formula. The actual input NOx amount is calculated from the actual input exhaust gas flow rate and the actual input NOx concentration. The actual output NOx amount is calculated from the actual input exhaust gas flow rate and the actual output NOx concentration. Actual NOx purification rate = (actual input NOx amount - actual output NOx amount) / actual input NOx amount

[0036] Furthermore, the NOx purification rate calculation unit 28 calculates a predicted purification rate of NOx contained in the exhaust gas based on the actual input exhaust gas flow rate, the actual input NOx concentration, and the predicted output NOx concentration. The predicted NOx purification rate is a predicted purification rate of harmful substances contained in the exhaust gas. The predicted NOx purification rate is calculated by the following formula. The predicted output NOx amount is calculated from the actual input exhaust gas flow rate and the predicted output NOx concentration. Predicted NOx purification rate = (actual input NOx amount measured value - predicted output NOx amount measured value) / actual input NOx amount measured value

[0037] The abnormality determination unit 29 calculates a deterioration index, which is used as a guideline for deterioration of the SCR 7, based on the actual NOx purification rate value and the predicted NOx purification rate value, and determines that the SCR 7 is abnormal when this deterioration index is equal to or greater than a threshold value.

[0038] Fig. 4 is a flowchart showing details of the procedure of the abnormality determination process executed by the abnormality determination unit 29. In Fig. 4, the abnormality determination unit 29 first inputs the NOx purification rate actual measurement value and the NOx purification rate predicted value calculated by the NOx purification rate calculation unit 28 (step S101).

[0039] Next, the abnormality determination unit 29 calculates a deterioration index of the SCR 7 based on the actual NOx purification rate value and the predicted NOx purification rate value (step S102). For example, if the SCR 7 is overheated or harmful substances that affect the performance of the SCR 7 enter the SCR 7, the NOx purification rate of the SCR 7 decreases, and the SCR 7 deteriorates. For this reason, the actual NOx purification rate value is more likely to be lower than the predicted NOx purification rate value (see FIG. 5(e)). Therefore, the deterioration index of the SCR 7 is calculated by the following formula. Deterioration index = predicted NOx purification rate - actual NOx purification rate

[0040] Next, the abnormality determination unit 29 compares the deterioration index of the SCR 7 with a predetermined threshold value to determine whether the deterioration index is equal to or greater than the threshold value (step S103). When the abnormality determination unit 29 determines that the deterioration index is smaller than the threshold value, it determines that the SCR 7 is normal (step S104).

[0041] When the abnormality determination unit 29 determines that the deterioration index is equal to or greater than the threshold value, it determines that the SCR 7 is abnormal (step S105). Then, the abnormality determination unit 29 outputs a warning control signal to the alarm device 22 (step S106). Then, the alarm device 22 issues a warning. Also, the abnormality determination unit 29 outputs injection amount data suitable for the deterioration index of the SCR 7 to the addition valve control unit 26 (step S107). Then, an amount of urea water according to the deterioration index of the SCR 7 is injected from the urea water addition valve 9.

[0042] In the catalyst abnormality detection device 20 as described above, when the operating state of the vehicle 2 enters a transient state A as shown in Figures 5(a) to 5(c), the flow rate and NOx concentration of exhaust gas flowing upstream of the SCR 7 in the exhaust passage 4 fluctuate, and the temperature of exhaust gas flowing upstream of the SCR 7 in the exhaust passage 4 rises. Then, as shown in Figure 5(d), the amount of urea water injected from the urea water addition valve 9 also fluctuates. After that, when the operating state of the vehicle 2 enters a steady state B, the flow rate, NOx concentration, and temperature of exhaust gas flowing upstream of the SCR 7 in the exhaust passage 4 and the amount of urea water injected from the urea water addition valve 9 stabilize.

[0043] However, as shown in FIG. 5(e), when the operating state of the vehicle 2 enters a transient state A, the NOx purification rate actual measurement value Q and the NOx purification rate predicted value P change, and the difference between the NOx purification rate actual measurement value Q and the NOx purification rate predicted value P becomes large.

[0044] Therefore, when determining whether or not an abnormality has occurred in the SCR7 based on the NOx purification rate actual value Q after the time t1 at which the NOx purification rate actual value Q stabilizes, a time lag occurs before it is determined that the SCR7 is abnormal.

[0045] 5(e) and (f), a deterioration index of the SCR 7 is calculated based on the actual NOx purification rate value Q and the predicted NOx purification rate value P, and at time t2 when this deterioration index reaches threshold value S, it is determined that the SCR 7 is abnormal. Time t2 is before time t1. Therefore, even if the operating state of the vehicle 2 enters transient state A, the abnormality determination of the SCR 7 is performed early.

[0046] As described above, in this embodiment, the state quantities of the exhaust gas flowing upstream and downstream of the SCR 7 in the exhaust passage 4 are detected, and the input state quantity actual measurement value and the output state quantity actual measurement value of the exhaust gas are obtained. Then, the state quantity of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4 is predicted using the input state quantity actual measurement value of the exhaust gas, and the output state quantity predicted value of the exhaust gas is obtained. Then, based on the output state quantity actual measurement value of the exhaust gas and the output state quantity predicted value of the exhaust gas, it is determined whether the SCR 7 is abnormal. Here, by predicting the state quantity of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4 so as to include not only the steady state of the vehicle 2 but also the transient state of the vehicle 2, it is possible to determine whether the SCR 7 is abnormal even before the operating state of the vehicle 2 is sufficiently stabilized. As a result, abnormality detection of the SCR 7 is performed quickly regardless of the operating state of the vehicle 2. As a result, an abnormality warning can be immediately issued to the driver of the vehicle 2.

[0047] In this embodiment, a deterioration index serving as a guide to deterioration of the SCR 7 is calculated based on an actual measurement value of the output state quantity of the exhaust gas and a predicted value of the output state quantity of the exhaust gas, and when the deterioration index is equal to or greater than a threshold value, it is determined that the SCR 7 is abnormal. By calculating the deterioration index serving as a guide to deterioration of the SCR 7 in this manner, it is possible to accurately determine an abnormality of the SCR 7 by simple processing.

[0048] In this embodiment, the actual NOx purification rate value is calculated based on the actual measured value of the input state quantity of the exhaust gas and the actual measured value of the output state quantity of the exhaust gas, and the predicted NOx purification rate value is calculated based on the actual measured value of the input state quantity of the exhaust gas and the predicted value of the output state quantity of the exhaust gas, and the actual NOx purification rate value and the predicted NOx purification rate value are compared to calculate a deterioration index. By calculating the actual NOx purification rate value and the predicted NOx purification rate value in this manner, a deterioration index that is used as a guideline for deterioration of the SCR 7 can be easily obtained. Also, before the actual NOx purification rate value Q becomes stable, it is determined whether the SCR 7 is abnormal.

[0049] Furthermore, in this embodiment, learning data used to predict the state quantities of the exhaust gas are determined based on the actual measured values ​​of the input state quantities of the exhaust gas and the actual measured values ​​of the output state quantities of the exhaust gas during a specified period after the start of use of the vehicle 2, and the state quantities of the exhaust gas flowing downstream of the SCR 7 in the exhaust passage 4 are predicted based on the actual measured values ​​of the input state quantities of the exhaust gas and the learning data. In this case, the learning data used to predict the state quantities of the exhaust gas are determined when the engine 3, SCR 7, etc. are brand new, so that more accurate output state quantity prediction values ​​can be obtained. Therefore, the accuracy of detecting an abnormality in the SCR 7 is improved.

[0050] The present invention is not limited to the above embodiment. For example, in the above embodiment, the learning data used to predict the state quantity of exhaust gas is determined during a specified period after the start of use of the vehicle 2, but the present invention is not particularly limited to this form, and for example, the process of determining the learning data may be performed periodically.

[0051] Further, in the above embodiment, the deterioration index of the SCR 7 is calculated as the difference between the predicted NOx purification rate value and the actual measured NOx purification rate value. However, the method of calculating the deterioration index of the SCR 7 is not particularly limited to this form, and may be, for example, the ratio between the predicted NOx purification rate value and the actual measured NOx purification rate value.

[0052] In the above embodiment, an abnormality in the SCR 7 is detected, but the present invention is not limited to this form, and an abnormality in the DOC 5 or the DPF 6 may be detected. Even in this case, the abnormality in the DOC 5 or the DPF 6 is promptly detected regardless of the operating state of the vehicle 2. In addition, for example, when regenerating the DPF 6, by injecting an amount of liquid fuel according to the deterioration index of the DPF 6 from the fuel addition valve 8, it is possible to prevent the regeneration time of the DPF 6 from becoming longer and prevent deterioration of fuel economy.

[0053] In addition, in addition to the SCR7, DOC5, and DPF6 described above, the present invention can also be applied to detecting abnormalities in exhaust catalysts such as NSR (NOx Storage-Reduction) that purifies NOx contained in exhaust gas, and three-way catalysts that purify CO, HC, and NOx contained in exhaust gas. [Explanation of symbols]

[0054] 2...vehicle, 3...engine (internal combustion engine), 4...exhaust passage, 7...SCR (exhaust catalyst), 20...catalyst abnormality detection device, 23...input state quantity detection unit, 24...output state quantity detection unit, 25...learning data determination unit, 27...output state quantity prediction unit, 28...NOx purification rate calculation unit (judgment unit), 29...abnormality judgment unit (judgment unit), P...NOx purification rate predicted value, Q...NOx purification rate actual measured value, S...threshold value.

Claims

1. A catalyst abnormality detection device for detecting an abnormality in an exhaust catalyst disposed in an exhaust passage connected to an internal combustion engine of a vehicle, an input state quantity detection unit that detects a state quantity of the exhaust gas flowing upstream of the exhaust catalyst in the exhaust passage and acquires an actual measured input state quantity value of the exhaust gas; an output state quantity detection unit that detects a state quantity of the exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage and obtains an actual measured value of an output state quantity of the exhaust gas; an output state quantity prediction unit that predicts a state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage using an input state quantity actual measurement value of the exhaust gas acquired by the input state quantity detection unit, and acquires an output state quantity prediction value of the exhaust gas; a determination unit that determines whether the exhaust catalyst is abnormal based on an actual output state quantity measurement value of the exhaust gas acquired by the output state quantity detection unit and an output state quantity prediction value of the exhaust gas acquired by the output state quantity prediction unit, the determination unit calculates a deterioration index serving as an indicator of deterioration of the exhaust catalyst based on the actual measured value of the output state quantity of the exhaust gas and the predicted value of the output state quantity of the exhaust gas, and determines that the exhaust catalyst is abnormal when the deterioration index is equal to or greater than a threshold value; a catalytic converter that detects a deterioration index by comparing the actual purification rate with the predicted purification rate, and a determination unit that determines a deterioration index based on an actual measurement of an input state quantity of the exhaust gas and an actual measurement of an output state quantity of the exhaust gas, and calculates a predicted purification rate of the harmful substances based on the actual measurement of the input state quantity of the exhaust gas and a predicted output state quantity of the exhaust gas, and a determination unit that determines a deterioration index by comparing the actual purification rate with the predicted purification rate,

2. A catalyst abnormality detection device for detecting an abnormality in an exhaust catalyst disposed in an exhaust passage connected to an internal combustion engine of a vehicle, comprising: an input state quantity detection unit that detects a state quantity of the exhaust gas flowing upstream of the exhaust catalyst in the exhaust passage and acquires an actual measured input state quantity value of the exhaust gas; an output state quantity detection unit that detects a state quantity of the exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage and obtains an actual measured value of an output state quantity of the exhaust gas; an output state quantity prediction unit that predicts a state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage using an input state quantity actual measurement value of the exhaust gas acquired by the input state quantity detection unit, and acquires an output state quantity prediction value of the exhaust gas; a determination unit that determines whether the exhaust catalyst is abnormal based on an actual output state quantity measurement value of the exhaust gas acquired by the output state quantity detection unit and an output state quantity prediction value of the exhaust gas acquired by the output state quantity prediction unit; a learning data determination unit that determines learning data to be used in predicting the state quantity of the exhaust gas, based on the input state quantity actual measurement value of the exhaust gas acquired by the input state quantity detection unit and the output state quantity actual measurement value of the exhaust gas acquired by the output state quantity detection unit, during a specified period after the use of the vehicle has started, The output state quantity prediction unit predicts a state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage based on an actual measured input state quantity value of the exhaust gas and learning data determined by the learning data determination unit.

3. A catalyst abnormality detection device for detecting an abnormality in an exhaust catalyst disposed in an exhaust passage connected to an internal combustion engine of a vehicle, comprising: an input state quantity detection unit that detects a state quantity of the exhaust gas flowing upstream of the exhaust catalyst in the exhaust passage and acquires an actual measured input state quantity value of the exhaust gas; an output state quantity detection unit that detects a state quantity of the exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage and obtains an actual measured value of an output state quantity of the exhaust gas; an output state quantity prediction unit that predicts a state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage using an input state quantity actual measurement value of the exhaust gas acquired by the input state quantity detection unit, and acquires an output state quantity prediction value of the exhaust gas; a determination unit that determines whether the exhaust catalyst is abnormal based on an actual output state quantity measurement value of the exhaust gas acquired by the output state quantity detection unit and an output state quantity prediction value of the exhaust gas acquired by the output state quantity prediction unit; a learning data determination unit that determines learning data used to predict the state quantity of the exhaust gas, based on the input state quantity actual measurement value of the exhaust gas acquired by the input state quantity detection unit and the output state quantity actual measurement value of the exhaust gas acquired by the output state quantity detection unit, The learning data includes data during a steady state in which the driving state of the vehicle is stable, and data during a transient state in which the driving state of the vehicle becomes unstable due to fluctuations in the state quantity of the exhaust gas, The output state quantity prediction unit predicts a state quantity of exhaust gas flowing downstream of the exhaust catalyst in the exhaust passage based on an actual measured input state quantity value of the exhaust gas and learning data determined by the learning data determination unit.

Images