WO2023079791A1

WO2023079791A1 - Control device and control method for internal combustion engine

Info

Publication number: WO2023079791A1
Application number: PCT/JP2022/027297
Authority: WO
Inventors: 章広小森; 邦彦鈴木
Original assignee: 日立Astemo株式会社
Priority date: 2021-11-04
Filing date: 2022-07-11
Publication date: 2023-05-11
Also published as: JPWO2023079791A1

Abstract

This control device is for controlling an internal combustion engine. The internal combustion engine is provided with: a three-way catalyst disposed in an exhaust flow passage; an upstream exhaust gas sensor disposed upstream of the three-way catalyst; and a downstream exhaust gas sensor disposed downstream of the three-way catalyst. The control device comprises: a frequency selection unit which extracts frequency components of output signals from the upstream exhaust gas sensor and the downstream exhaust gas sensor; a signal processing unit which estimates a condition of the three-way catalyst on the basis of the outputs from the frequency selection unit; and a calculation unit which calculates an air-fuel ratio from the condition of the three-way catalyst.

Description

CONTROL DEVICE AND CONTROL METHOD FOR INTERNAL COMBUSTION ENGINE

Import by reference

This application claims the priority of Japanese Patent Application No. 2021-180141, which was filed in Japan on November 4, 2021, and incorporates the contents thereof into the present application by reference.

The present invention relates to a control device for an internal combustion engine.

A control technology is known in which a three-way catalyst is provided in the exhaust pipe of an internal combustion engine, and the oxygen storage state in the three-way catalyst is detected by exhaust gas sensors installed in front and behind the catalyst, and the air-fuel ratio is corrected according to the result. there is

As background technologies in this technical field, there are the following prior arts. Patent Document 1 (Japanese Patent Laid-Open No. 2011-174426) describes a first air-fuel ratio sensor provided upstream of a catalyst mounted in an exhaust passage of an internal combustion engine, and a second air-fuel ratio sensor provided downstream of the catalyst. and an air-fuel ratio control unit that estimates the amount of oxygen released from the catalyst with reference to at least the output of the first air-fuel ratio sensor and performs feedback control of the amount of oxygen released to a target value; An air-fuel ratio control device is described that includes a learning unit that learns a learning parameter for reduction when a fuel cut occurs in an internal combustion engine.

Further, in Patent Document 2 (Japanese Patent Laid-Open No. 2020-133467), the mapping data stored in the storage device 76 is input with time-series data of variables corresponding to the upstream air-fuel ratio and the downstream air-fuel ratio. This is data that defines a neural network that outputs a deterioration degree variable, which is a variable that indicates the degree of deterioration of the side catalyst. The CPU inputs time-series data of variables corresponding to the upstream air-fuel ratio and the downstream air-fuel ratio to the mapping defined by the mapping data, thereby calculating the deterioration degree variable of the upstream catalyst. Are listed.

　The oxygen storage capacity of the catalyst changes due to the deterioration of the three-way catalyst. Since the change in oxygen storage capacity affects the determination of air-fuel ratio correction control, it is necessary to accurately diagnose the degree of deterioration of the catalyst. If the deterioration state of the catalyst is erroneously determined, the air-fuel ratio cannot be corrected to the optimum, and the oxygen storage capacity cannot be maintained within an appropriate range. In the prior art, the degree of deterioration of the catalyst is diagnosed from the correlation between the exhaust gas sensor signals before and after the catalyst. With a new catalyst, the amplitude of the downstream exhaust gas sensor signal is smaller than that of the upstream exhaust sensor. . The degree of deterioration of the catalyst is diagnosed using the correlation between the upstream and downstream sensor signals of the catalyst. However, the prior art is based on the assumption that the air-fuel ratio is fluctuating at a steady state, and there is a problem that the opportunities are limited.

In order to solve such a problem, the technique described in Patent Document 2 uses upstream average A/F, downstream average A/F, exhaust gas flow rate, rotation speed, A coefficient representing the degree of deterioration is obtained from a neural network with six inputs of filling efficiency and catalyst temperature. Then, the degree of deterioration of the catalyst is determined by comparing the coefficient with the threshold value. However, if the preconditions of the input data are changed due to disturbances such as the addition of devices or controls, the number of inputs will be increased, and the scale of the neural network will increase, which may make implementation difficult.

Therefore, an object of the present invention is to provide a technique for accurately determining the degree of deterioration of a catalyst under various conditions using the frequency component of the output of an exhaust gas sensor.

A representative example of the invention disclosed in the present application is as follows. That is, a control device for controlling an internal combustion engine, wherein the internal combustion engine includes a three-way catalyst arranged in an exhaust flow path, an upstream exhaust gas sensor arranged upstream of the three-way catalyst, and a three-way catalyst. a frequency selector for extracting frequency components of an output signal of the upstream exhaust gas sensor and an output signal of the downstream exhaust gas sensor; and a frequency selector based on the output of the frequency selector. A signal processing unit for estimating the state of the primary catalyst and a calculation unit for calculating the air-fuel ratio from the state of the three-way catalyst are provided.

According to one aspect of the present invention, the degree of deterioration of the catalyst can be accurately determined under various conditions. Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.

1 is a schematic configuration diagram of an entire system of an engine control device according to an embodiment of the present invention; FIG. 1 is a diagram showing the configuration of an aftertreatment system for purifying exhaust gas from an internal combustion engine; FIG. FIG. 4 is a diagram showing the tendency of the chemical species concentration of exhaust gas with respect to the equivalent ratio; FIG. 2 is a diagram showing the main reaction process of a three-way catalyst (ceria system) used in an aftertreatment system; FIG. 4 is a diagram showing the tendency of the purification efficiency of a three-way catalyst with respect to the exhaust gas equivalent ratio above the catalyst activation temperature. FIG. 4 is a diagram showing the air-fuel ratios upstream and downstream of the catalyst and the output of an oxygen sensor installed downstream of the catalyst; FIG. 4 is a diagram showing the hysteresis characteristic of an oxygen sensor; FIG. 4 is a diagram showing temporal changes in NOx concentrations downstream of an oxygen sensor and a catalyst; FIG. 4 is a diagram showing the relationship between the degree of catalyst deterioration and the oxygen storage capacity of a three-way catalyst; FIG. 4 is a diagram showing the relationship between the oxygen storage ratio and the NOx purification efficiency; FIG. 4 is a diagram showing the air-fuel ratios upstream and downstream of the catalyst and the output of an oxygen sensor installed downstream of the catalyst; FIG. 4 is a schematic diagram showing spectral changes for each low to high frequency of a band-pass filter in new, centrally deteriorated, and completely deteriorated states. FIG. 10 is a block diagram of a control model that performs catalyst deterioration diagnosis based on signals from exhaust gas sensors before and after the catalyst, and corrects and controls the air-fuel ratio in consideration of the amount of stored oxygen. FIG. 4 is a diagram showing an overview of catalyst deterioration diagnosis by a deterioration diagnosis unit; FIG. 10 is a diagram showing a method of realizing catalyst deterioration diagnosis based on the frequency of an exhaust gas sensor signal using a neural network; 14 is a flowchart of processing by the control model shown in FIG. 13; FIG. 14 is a diagram showing the effect of the control model shown in FIG. 13; It is a figure which shows the test|inspection procedure of ECU of a present Example. FIG. 4 is a block diagram of a control model that performs catalyst deterioration diagnosis from signals from exhaust gas sensors before and after the catalyst, and corrects and controls the air-fuel ratio in consideration of the amount of effective Ce. FIG. 20 is a flowchart of processing by the control model shown in FIG. 19; FIG.

A control device for an internal combustion engine according to a first embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic configuration diagram of the entire system of the engine control device.

The engine control device includes an internal combustion engine 1, a flow rate sensor 2, a turbocharger 3, an air bypass valve 4, an intercooler 5, a supercharging temperature sensor 6, a throttle valve 7, an intake manifold 8, a supercharging pressure sensor 9, and a flow enhancer. valve 10, intake valve 11, exhaust valve 13, fuel injection valve 15, spark plug 16, knock sensor 17, crank angle sensor 18, waste gate valve 19, air-fuel ratio sensor 20 which is an upstream exhaust gas sensor, exhaust purification catalyst (three-way catalyst) 21, an oxygen sensor 22 which is a downstream exhaust gas sensor, an EGR (Exhausted Gas Recirculation) pipe 23, an EGR cooler 24, an EGR valve 25, a temperature sensor 26, a differential pressure sensor 27 and an ECU (Electronic Control Unit) 28. Further, in addition to the oxygen sensor 22, an air-fuel ratio sensor may be provided as the downstream exhaust gas sensor arranged downstream of the exhaust purification catalyst 21.

An intake channel and an exhaust channel communicate with the cylinders of the internal combustion engine 1 . A flow rate sensor 2 and an intake temperature sensor incorporated in the flow rate sensor 2 are provided in the intake flow path. The turbocharger 3 is composed of a compressor 3a connected to an intake passage and a turbine 3b connected to an exhaust passage. The turbine 3b converts the energy of the exhaust gas from the internal combustion engine 1 into rotational energy of turbine blades. The compressor 3a compresses the intake air that has flowed in from the intake flow path by rotating the compressor blades that are connected to the turbine blades.

The intercooler 5 is provided downstream of the compressor 3a of the turbocharger 3, and cools the intake air whose temperature has been increased by being adiabatically compressed by the compressor 3a. The supercharging temperature sensor 6 is provided downstream of the intercooler 5 and measures the temperature of the intake air cooled by the intercooler 5 (supercharging temperature). The throttle valve 7 is provided downstream of the intercooler 5 and throttles the intake passage to control the amount of intake air flowing into the cylinders of the internal combustion engine 1 . The throttle valve 7 is composed of an electronically controlled butterfly valve whose valve opening can be controlled independently of the amount of depression of the accelerator pedal by the driver. An intake manifold 8 provided with a supercharging pressure sensor 9 communicates with an intake passage downstream of the throttle valve 7 .

The intake manifold 8 provided downstream of the throttle valve 7 and the intercooler 5 may be integrated. In this case, since the volume from the downstream of the compressor 3a to the cylinder can be reduced, the responsiveness of acceleration and deceleration can be improved, and the controllability can be improved.

The flow enhancement valve 10 is arranged downstream of the intake manifold 8 and enhances turbulence of the flow inside the cylinder by creating a bias in the intake air taken into the cylinder. Turbulent combustion is promoted and stabilized by closing the flow enhancing valve 10 when exhaust gas recirculation combustion, which will be described later, is performed. The internal combustion engine 1 has an intake valve 11 and an exhaust valve 13 . The intake valve 11 and the exhaust valve 13 each have a variable valve mechanism for continuously varying the valve opening/closing phase. The variable valve mechanisms of the intake valve 11 and the exhaust valve 13 are provided with

sensors

12 and 14, respectively, for detecting the opening/closing phases of the valves. A cylinder of the internal combustion engine 1 is provided with a direct fuel injection valve 15 that injects fuel directly into the cylinder. The fuel injection valve 15 may be of a port injection type that injects fuel into the intake port.

A cylinder of the internal combustion engine 1 is provided with a spark plug 16 that exposes an electrode portion inside the cylinder and ignites a combustible air-fuel mixture with a spark. The knock sensor 17 is provided in the cylinder block and detects the presence or absence of knock by detecting cylinder block vibration caused by combustion pressure vibration generated in the combustion chamber. The crank angle sensor 18 is provided on the crankshaft, and outputs a signal corresponding to the rotation angle of the crankshaft to the ECU 28, which will be described later, as a signal indicating the rotation speed.

The air-fuel ratio sensor 20 is provided in the exhaust passage downstream of the turbine 3b of the turbocharger 3, and outputs a signal indicating the detected exhaust gas composition, that is, the air-fuel ratio, to the ECU 28. The exhaust purification catalyst 21 is a three-way catalyst, is provided downstream of the air-fuel ratio sensor 20, and purifies harmful exhaust gas components such as carbon monoxide, nitrogen compounds, and unburned hydrocarbons in the exhaust gas by catalytic reaction. An oxygen sensor 22 is provided downstream of the exhaust purification catalyst 21 to detect the concentration of oxygen contained in the exhaust gas purified by the exhaust purification catalyst 21 .

The turbocharger 3 is provided with an air bypass valve 4 and a wastegate valve 19 . The air bypass valve 4 is arranged on a bypass flow path connecting upstream and downstream of the compressor 3a in order to prevent an excessive increase in pressure from the downstream of the compressor 3a to the upstream of the throttle valve 7. As shown in FIG. When the throttle valve 7 is abruptly closed in the supercharging state, the air bypass valve 4 is opened under the control of the ECU 28, so that the compressed intake air in the downstream portion of the compressor 3a passes through the bypass passage to the compressor 3a. flows back upstream of the As a result, the supercharging pressure is immediately lowered to prevent the surging phenomenon and appropriately prevent damage to the compressor 3a.

The wastegate valve 19 is arranged on a bypass flow path that connects upstream and downstream of the turbine 3b. The wastegate valve 19 is an electrically operated valve whose valve opening degree can be freely controlled with respect to the supercharging pressure under the control of the ECU 28 . When the opening degree of the waste gate valve 19 is adjusted by the ECU 28 based on the supercharging pressure detected by the supercharging pressure sensor 9, a part of the exhaust gas passes through the bypass flow path, thereby reducing the work given to the turbine 3b by the exhaust gas. can be reduced to maintain the boost pressure at the target pressure.

The EGR pipe 23 communicates with an exhaust passage downstream of the exhaust purification catalyst 21 and an intake passage upstream of the compressor 3a, and diverts the exhaust gas from the downstream of the exhaust purification catalyst 21 to the upstream of the compressor 3a. Reflux. An EGR cooler 24 provided in the EGR pipe 23 cools the exhaust gas. The EGR valve 25 is provided downstream of the EGR cooler 24 and controls the flow rate of exhaust gas. The EGR pipe 23 is provided with a temperature sensor 26 that detects the temperature of the exhaust gas upstream of the EGR valve 25 and a differential pressure sensor 27 that detects a differential pressure between upstream and downstream of the EGR valve 25 .

The ECU 28 is an arithmetic circuit that has a CPU, ROM, RAM, A/D converter, driver circuit, etc., and controls each component of the engine control device and executes various data processing. The various sensors and actuators described above are connected to the ECU 28 . The ECU 28 controls the operations of actuators such as the throttle valve 7, the intake and exhaust valves 11 and 13 with variable valve mechanisms, the fuel injection valve 15, the EGR valve 25, and the like. The ECU 28 also detects the operating state of the internal combustion engine 1 based on signals input from various sensors, and ignites the spark plug 16 at a timing determined according to the operating state.

FIG. 2(A) is a diagram showing the configuration of an aftertreatment system that purifies exhaust gas from an internal combustion engine. A three-way catalyst is used as the exhaust gas purification catalyst. For the purpose of maintaining the purification efficiency of the three-way catalyst at an optimum point, sensors for detecting the exhaust gas composition are provided at each of the upstream and downstream portions of the three-way catalyst. In the aftertreatment system, an air-fuel ratio sensor 20 is provided upstream, and an oxygen sensor 22 is provided downstream. According to this configuration, the air-fuel ratio of the exhaust gas flowing into the three-way catalyst can be measured, and the presence or absence of oxygen contained in the exhaust gas after catalyst purification can be detected.

FIG. 2(B) is a diagram showing the relationship between the equivalence ratio (=theoretical air-fuel ratio/air-fuel ratio) of the exhaust gas and the output of the air-fuel ratio sensor 20. FIG. As the equivalence ratio increases (richer), the sensor output tends to decrease. The air-fuel ratio sensor 20 is characterized in that it can accurately detect the equivalence ratio over a wide range from the exhaust gas lean state to the rich state by obtaining the relationship shown in FIG. 4 in advance.

FIG. 2(C) is a diagram showing the relationship between the equivalence ratio of the exhaust gas and the output of the oxygen sensor 22. FIG. The oxygen sensor 22 outputs a signal according to the electromotive force associated with the concentration difference between the oxygen concentration contained in the exhaust gas and the oxygen concentration in the air. The oxygen sensor 22 has a characteristic of outputting a generally minimum electromotive force under lean conditions, outputting a maximum electromotive force under rich conditions, and suddenly changing the output at a stoichiometric air-fuel ratio (equivalence ratio of 1.0). By capturing the change timing of the sensor output and feeding it back to the air-fuel ratio control, the exhaust gas equivalence ratio can be maintained near the theoretical air-fuel ratio.

FIG. 3 is a diagram showing the tendency of chemical species concentration in exhaust gas with respect to the equivalence ratio. Combustion gas compositions of hydrocarbon-based fuels tend to increase CO (carbon monoxide) and H2 (hydrogen) on the rich side and O2 (oxygen) on the lean side of the stoichiometric air-fuel ratio. Furthermore, NOx (nitrogen oxides) shows a maximum value slightly on the lean side of the stoichiometric air-fuel ratio, and tends to decrease on the lean and rich sides. HC (unburned hydrocarbons) is a component that is emitted without being burned, and excessive lean or enrichment tends to result in HC being emitted without leading to normal combustion. Even under the stoichiometric air-fuel ratio condition in which fuel and air (oxygen) are supplied in just the right amount, CO (carbon monoxide) and NOx (carbon monoxide) and NOx ( Nitrogen oxides) are emitted in a certain amount, so an appropriate purification treatment by an aftertreatment system is required.

Fig. 4 is a diagram showing the main reaction process of a three-way catalyst (ceria-based) used in the post-treatment system. The three-way catalytic reaction process mainly consists of an oxidation reaction, a NOx reduction reaction, and an oxygen storage/release reaction.

In the oxidation reaction, CO, H2, and HC generated under rich or high temperature conditions react with oxygen to generate harmless CO2 and H2O. Unburned hydrocarbons (HC) include hydrocarbon components such as methane, propane, ethylene, and butane, which react at different rates. The NOx reduction reaction is represented mainly by the reaction of CO and NO to produce harmless CO2 and N2. In the oxygen storage/release reaction, the storage/release of oxygen and the oxidation and reduction reactions of HC, CO, and NO proceed via Ce (cerium), which is a catalyst material. That is, the reaction of cerium dioxide (CeO2) with CO and HC produces harmless CO2 and H2O, and the reaction of cerium trioxide (Ce2O3) with NO produces harmless N2. At this time, the oxygen storage ratio of the three-way catalyst is defined by the balance between CeO2 and Ce2O3 that are produced at the same time. That is, when all of the Ce2O3 in the catalyst becomes CeO2, it cannot react with NO and cannot purify NO. In this way, in order to properly maintain the purification efficiency of the three-way catalyst, it is necessary to maintain the balance between CeO2 and Ce2O3, that is, the oxygen storage ratio, at a predetermined value. All of the reaction processes described above strongly depend on the catalyst temperature, and it is necessary to appropriately manage the catalyst temperature so that it reaches the activation temperature or higher in an early stage after starting. Although the system shown in this embodiment uses a ceria-based three-way catalyst, the present invention is not limited to this. Catalysts using other materials exhibiting similar effects can also achieve similar effects by adjusting the constants of the control model without changing the configuration of the invention. In addition to the reaction mechanism described in FIG. 4, the catalytic reaction may also use the water-gas shift reaction or the like. These reaction mechanisms can also be handled by adjusting the control model constants.

FIG. 5 is a diagram showing the tendency of the purification efficiency of the three-way catalyst with respect to the exhaust gas equivalence ratio above the catalyst activation temperature. The purification efficiency characteristics of the three-way catalyst change with the stoichiometric air-fuel ratio as a boundary. Under lean conditions, the CO and HC purification efficiency is generally maintained at 90% or more, and the NOx purification efficiency decreases as the equivalence ratio decreases. . On the other hand, on the rich side, the purification efficiency of HC and CO decreases as the equivalence ratio increases. In the vicinity of the theoretical air-fuel ratio, the purification efficiency of NOx, HC, and CO is 90% or more, and this point is called the ternary point. The three-way catalyst is controlled to keep the purification efficiency at a high level by maintaining the equivalence ratio at the stoichiometric air-fuel ratio, which is the three-way point.

FIG. 6 shows the air-fuel ratio downstream of the catalyst and the oxygen sensor 22 installed downstream of the catalyst when the air-fuel ratio is varied stepwise to the lean side and the rich side over time, centering on the equivalence ratio of 1.0. FIG. 10 is a diagram showing the output; Even when the equivalence ratio is the stoichiometric air-fuel ratio, it is maintained in an intermediate state by discharging a very small amount of oxygen downstream of the catalyst. When the ratio is changed stepwise to the lean side, the rear equivalence ratio gradually changes, and the output of the oxygen sensor 22 rapidly changes to the minimum value side after a delay time. On the other hand, when the air-fuel ratio is changed from the lean side to the rich side, the rear equivalence ratio gradually changes, and the output of the oxygen sensor 22 exhibits steresis characteristics in which it rapidly changes after a large delay. In this way, the change from lean to rich and the change from rich to lean tend to have different delay times. This is due to the difference in reaction rate between CeO2 and Ce2O3 in the three-way catalyst described in FIG. Since the reaction rate depends on the catalyst temperature and exhaust gas flow rate, the hysteresis described above also changes depending on the catalyst temperature and exhaust gas flow rate.

FIG. 7 is a diagram showing the hysteresis characteristic of the oxygen sensor 22. FIG. The static characteristics of the oxygen sensor 22 are shown in FIG. On the other hand, the oxygen sensor 22 also uses a catalytic material and has hysteresis due to detection delay. That is, when the lean state changes rapidly to the rich state, the change timing of the sensor output shifts to the rich side of the equivalence ratio, and when the rich state changes suddenly to the lean state, the change timing of the sensor output shifts to the equivalence ratio. Shift to the lean side of the ratio. Furthermore, the behavior described above is affected by deterioration of the material properties of the sensor and temperature. In the control model, it is necessary to consider changes in dynamic characteristics including sensor deterioration and temperature effects in addition to catalyst deterioration.

FIG. 8 shows the oxygen sensor 22 and the NOx concentration downstream of the catalyst when the firing operation is performed again in the stoichiometric air-fuel ratio state after the motoring operation (fuel cut) period from the state controlled at the stoichiometric air-fuel ratio. It is a figure which shows a change. After the fuel cut, when the firing operation with the stoichiometric air-fuel ratio is started again, the detected value of the oxygen sensor 22 increases with a delay as shown in FIG. The NOx concentration downstream of the catalyst shows a behavior in which NOx is discharged in a spike during the delay period until the oxygen sensor 22 is restored. On the other hand, when the firing operation is started again after the fuel cut, if the stoichiometric air-fuel ratio control is performed after performing the rich correction once, NOx emissions downstream of the catalyst can be prevented. Since the oxygen sensor 22 detects the oxygen state of the gas downstream of the catalyst, the internal state of the catalyst has already changed to the maximum or minimum oxygen storage state at the time the oxygen sensor 22 reacts. Therefore, in the control method of stopping the rich correction control after the reaction of the oxygen sensor 22, HC cannot be prevented appropriately because oxygen is insufficient for the catalyst during the control timing delay period. Therefore, in the air-fuel ratio control of the internal combustion engine, it is necessary to consider the state inside the catalyst, which cannot be directly observed from the outside, and control the rich correction during an appropriate period.

FIG. 9 is a diagram showing the relationship between the degree of catalyst deterioration and the oxygen storage capacity of the three-way catalyst. Catalyst deterioration is a state in which the catalytic action is lowered due to the influence of heat and poisoning by sulfur contained in the fuel. As the catalyst deterioration progresses, the oxygen storage capacity decreases. The effect of changes in oxygen storage capacity on the purification action of the three-way catalyst will be described below.

FIG. 10 is a diagram showing the relationship between the oxygen storage ratio and the NOx purification efficiency. When the oxygen storage ratio exceeds a predetermined value, NOx purification efficiency deteriorates significantly. This is because, as described in FIG. 4, Ce2O3 in the catalyst is important for purifying NOx, and if all Ce2O3 reacts and changes to CeO2, it cannot react with NO and cannot purify NO. be. For this reason, as described in FIG. 5, in order to keep the catalyst purification efficiency at a high level, it is necessary not only to keep the exhaust gas air-fuel ratio at the catalyst inlet at the three-way point, but also to keep the oxygen storage ratio within a predetermined range. In addition, it is necessary to appropriately correct and control the exhaust gas air-fuel ratio at the catalyst inlet.

FIG. 11 shows the equivalence ratio downstream of the catalyst and the equivalence ratio installed downstream of the catalyst when the air-fuel ratio is changed stepwise to the lean side and the rich side over time with the equivalence ratio of 1.0 as the center for the new catalyst and the deteriorated catalyst. FIG. 10 is a diagram showing an output of an oxygen sensor 22 (downstream exhaust gas sensor).

The deteriorated catalyst reduces the delay in change in the output of the oxygen sensor 22 with respect to changes in the air-fuel ratio to the lean and rich sides compared to the new catalyst. This can be explained by the temporal transition of the oxygen storage rate of the catalyst. That is, the oxygen storage capacity of the catalyst decreases due to deterioration, and the oxygen storage rate reaches the maximum and minimum values more quickly, which accelerates the release of oxygen downstream and reduces the delay in change in the output of the oxygen sensor 22. It is from. Therefore, the rich correction period after returning from fuel cut described in FIG. 8 must be set in consideration of the deterioration state of the catalyst.

FIG. 12 is a schematic diagram showing spectral changes for each low to high frequency band-pass filter in new, centrally deteriorated, and completely deteriorated states. The actual spectrum varies in various ways depending on the state of upstream exhaust gas.

From top to bottom, new, moderately deteriorated, and completely deteriorated are shown in that order. The left side shows the signal of the upstream exhaust gas sensor (air-fuel ratio sensor 20) after band-pass filtering, and the right side shows the signal of the downstream exhaust gas sensor (oxygen sensor 22) after band-pass filtering. Since the upstream exhaust gas sensor signal changes in synchronism with changes in the air-fuel ratio of the exhaust gas, it has the same spectrum regardless of the deterioration state of the catalyst. On the other hand, since the downstream exhaust gas sensor signal is affected by the oxygen storage capacity of the catalyst, the frequency spectrum differs for each deterioration state.

　In a new catalyst, the catalyst has sufficient oxygen storage capacity, so the downstream exhaust gas sensor signal shows a small and gradual change, resulting in a low frequency spectrum. In the moderately deteriorated catalyst, the downstream exhaust gas sensor signal shows a change that swings more than in the new state due to the decrease in the oxygen storage capacity due to the decrease in Ce of the catalyst. Therefore, the spectrum is from frequencies lower to higher than those in the new state. With a completely degraded catalyst, Ce decreases more than with a moderately degraded catalyst, so the change is similar to that of the upstream exhaust gas sensor signal. The frequency change is greater than for the moderately aged catalyst and approaches the spectrum of the upstream exhaust gas sensor signal.

By making the neural network 1301b learn in advance the relationship between various attenuation patterns for each catalyst deterioration state and the catalyst deterioration, the catalyst deterioration level can be determined. In this embodiment, three frequencies of high frequency, medium frequency, and low frequency are used in order to divide into three levels of complete deterioration, moderate deterioration, and new product as an example, but the number of frequency bands that can be used is not limited. , more than three or less than three can be applied. For example, the degradation state of the catalyst may be determined based on two signals obtained by frequency-decomposing the upstream exhaust gas sensor signal and two signals obtained by frequency-decomposing the downstream exhaust gas sensor signal using two band-pass filters of high frequency and low frequency.

Also, the frequency band set for each filter may not be a fixed value, but may be a variable that is adjusted according to the operating conditions of the internal combustion engine. For example, it is preferable to prepare a table that associates the pass frequency band of each bandpass filter with the rotational speed of the internal combustion engine, refer to the outer table, and increase the frequency as the rotational speed increases. Although the frequency characteristics of catalyst deterioration can be extracted from the frequency information of the exhaust gas sensor signal, the actual signal waveform does not have a constant shape, so signal processing is difficult. Therefore, in the present invention, the deterioration of the catalyst is diagnosed using the neural network 1301b that has been learned by associating the frequency information of the sensor with the deterioration of the catalyst.

Also, the model for diagnosing deterioration of the catalyst may be a model other than the neural network 1301b (for example, a regression model) for diagnosing deterioration of the catalyst.

FIG. 13 is a block diagram of a control model that performs catalyst deterioration diagnosis from exhaust gas sensor signals upstream and downstream of the catalyst, and corrects and controls the air-fuel ratio in consideration of the oxygen storage amount.

　The control model shown in FIG.

A deterioration diagnosis unit 1301 receives exhaust gas sensor signals upstream and downstream of the catalyst, diagnoses the degree of deterioration of the catalyst, and is composed of a frequency selection unit 1301a and a neural network 1301b. The frequency selector 1301a has a plurality of filters for extracting the frequency characteristics of the exhaust gas sensor signal as shown in FIG. In FIG. 13, as an example, three low-pass filters for each frequency are combined to form a band-pass filter.

The exhaust gas sensor signals upstream and downstream of the catalyst pass through low-pass filters with cutoff frequencies of high, medium, and low frequencies, are input to the calculation section to be decomposed into each frequency band, and are sent to the band-pass filter. filtered by For example, the band-pass filters are prepared for three frequency bands of high frequency, medium frequency, and low frequency, and each band-pass filter has 3 signals obtained by filtering the upstream exhaust gas sensor signal and 3 signals obtained by filtering the downstream exhaust gas sensor signal. Output a signal.
BPF3 (band pass filter 3)
Original waveform-Signal after passing through LPF3 BPF2 (band pass filter 2)
Signal after passing through LPF3 - Signal after passing through LPF2 BPF1 (band pass filter 1)
Signal after passing LPF2 - Signal after passing LPF1

The frequency selection unit 1301a configures a band-pass filter by combining low-pass filters. You can use three. Also, the frequency selection unit 1301a may extract the frequency component of the exhaust gas sensor signal by Fourier transform instead of the filter.

The neural network 1301b learns frequency information associated with the deterioration state of the catalyst, receives the signal after passing through the band-pass filter output from the frequency selection unit 1301a, and determines whether the input signal is new, moderately deteriorated, or perfect. A degradation diagnosis result (for example, the degree of certainty of each state) indicating which of the three states of degradation is present is output. For example, the neural network 1301b detects the relative relationship between the upstream and downstream exhaust gas sensor signals of the catalyst that have passed through the band-pass filter 1, the band-pass filter 2 , and the relative relationship between the upstream and downstream exhaust gas sensor signals of the catalyst that have passed through the band-pass filter 3 are learned as teacher data.

The OSC estimation model selection unit 1302 receives the deterioration diagnosis result and selects an OSC estimation model according to the deterioration diagnosis result. The OSC estimator 1303 calculates an estimated oxygen storage capacity (estimated OSC) using the selected oxygen storage capacity (OSC) estimation model. Catalyst control unit 1304 receives the estimated OSC and calculates an air-fuel ratio correction amount that does not lower the purification rate due to the estimated OSC, according to the operating conditions. The operating conditions are, for example, throttle opening, engine load, engine speed, temperature, atmospheric pressure, and the like. An air-fuel ratio correction amount calculation unit 1305 outputs a post-correction target air-fuel ratio based on the air-fuel ratio target value and the air-fuel ratio correction amount calculated by the catalyst control unit 1304 .

FIG. 14 is a diagram showing an overview of catalyst deterioration diagnosis by the deterioration diagnosis unit 1301. FIG. The exhaust gas sensor signals upstream and downstream of the catalyst input to the deterioration diagnosis unit 1301 are frequency-decomposed by a bandpass filter. The band-pass filters are three band-pass filters of high frequency, medium frequency, and low frequency, and each band-pass filter generates three signals obtained by frequency-resolving the upstream exhaust gas sensor signal and three signals obtained by frequency-resolving the downstream exhaust gas sensor signal. Input to neural network 1301b. The neural network 1301b learns by associating the intensity of each frequency with the degree of deterioration of the catalyst, classifies the deterioration state of the catalyst from the input signal into three levels of complete deterioration, moderate deterioration, and new condition, and outputs the deterioration diagnosis result. Output.

FIG. 15 is a diagram showing a method of realizing catalyst deterioration diagnosis based on the frequency of the exhaust gas sensor signal with the neural network 1301b. A neural network model is a mathematical model that imitates the mechanism of human brain neural circuits, and weights and biases are set for each neuron that constitutes the model. A function called an activation function is defined in a neuron. A logistic function, a ramp function, or the like may be appropriately set as the activation function. A plurality of neurons form one layer, and an intermediate layer is set between the input layer and the output layer. More complex input-output relationships can be approximated by increasing the number of neurons and the number of hidden layers. There is a trade-off relationship between approximation accuracy and model scale, and a compatible point that satisfies both requirements is selected. By setting the frequency of the exhaust gas sensor signal in the input layer and the catalyst deterioration diagnosis in the output layer, and performing (supervised) machine learning on the weight and bias of each neuron, the input/output relationship can be approximated. Backpropagation can be applied to machine learning algorithms.

FIG. 16 uses the control model shown in FIG. 13 to decompose the exhaust gas sensor signals upstream and downstream of the catalyst into three frequency bands using a band-pass filter, input them to the neural network 1301b, and output deterioration diagnosis results. 4 is a flowchart of processing for executing air-fuel ratio correction;

In step 101, the deterioration diagnosis unit 1301 receives input of exhaust gas sensor signals upstream and downstream of the catalyst. disassemble. In step 103, the neural network 1301b receives input of exhaust gas sensor signals upstream and downstream of the catalyst resolved into frequency bands. In step 104, the neural network 1301b, which has learned frequency information associated with the deterioration state of the catalyst, estimates the deterioration of the catalyst from the input exhaust gas sensor signals upstream and downstream of the catalyst. At step 105, the OSC estimation model selection unit 1302 selects an OSC model according to the deterioration state according to the deterioration diagnosis result. A plurality of OSC models are prepared according to the state of deterioration of the catalyst (for example, the coefficients to be multiplied by Ψ in the oxygen storage ratio formula in FIG. 4 are different), and an appropriate OSC model is selected according to the state of deterioration of the catalyst. selected. At step 106, the OSC estimator 1303 estimates the oxygen storage capacity (OSC) of the catalyst using the selected OSC model. At step 107, catalyst control unit 1304 calculates an air-fuel ratio correction amount based on the estimated oxygen storage amount. At step 108 , the air-fuel ratio correction amount calculator 1305 uses the air-fuel ratio target value and the air-fuel ratio correction amount calculated at step 107 to calculate the post-correction target air-fuel ratio.

FIG. 17 is a diagram showing the effects of the control model shown in FIG.

From the state controlled at the stoichiometric air-fuel ratio, after the motoring operation (fuel cut) period, the firing operation is performed again at the stoichiometric air-fuel ratio. When the firing operation is started again after the fuel cut, the rich correction is performed once, and then the stoichiometric air-fuel ratio control is performed to prevent NOx emissions downstream of the catalyst. According to this embodiment, the deterioration state of the catalyst can be diagnosed with high accuracy, so the rich correction can be performed in an appropriate period, and NOx and HC emissions downstream of the catalyst can be reduced.

Up to this point, the deterioration estimation from the upstream and downstream exhaust gas sensor signals has been explained. However, since the change in the output value of the upstream exhaust gas sensor is small, the deterioration of the catalyst is estimated only from the output of the downstream exhaust gas sensor (oxygen sensor 22). good too.

The control model of this embodiment may have an output sensor failure diagnosis function. For example, after the neural network 1301b learns the normal output signals of the upstream exhaust gas sensor and the downstream exhaust gas sensor, the output signal of the upstream exhaust gas sensor, the output signal of the downstream exhaust gas sensor, and the output signal of each exhaust gas sensor under normal conditions. An abnormality in the exhaust gas sensor may be diagnosed based on the degree of similarity between the two. Then, when it is determined that the exhaust gas sensor is abnormal, the neural network 1301b should stop estimating the state of the catalyst.

Furthermore, the control model of this embodiment may have a heater failure diagnosis function. The oxygen sensor 22, which is a downstream exhaust gas sensor, has a heater that maintains the temperature at which the material (zirconia) used for the sensor is activated. For example, the deterioration diagnosis unit 1301 monitors the heater temperature of the oxygen sensor 22, and after the neural network 1301b learns the heater energization state and normal heater temperature, You may diagnose the abnormality of a heater by the similarity of . Then, when it is determined that the heater is abnormal, the neural network 1301b should stop estimating the state of the catalyst.

FIG. 18 is a diagram showing the inspection procedure of the catalyst state estimation function of the ECU 28 of this embodiment.

When the learned neural network 1301b is under control with the stoichiometric air-fuel ratio, after a period of motoring operation (fuel cut), firing operation is performed again with the stoichiometric air-fuel ratio, the catalyst is new, moderate. The upstream and downstream exhaust gas sensor signals of the catalyst at the time of deterioration and at the time of complete deterioration are connected, and test data simulating the signal after passing through the band-pass filter is input. Then, output signals of exhaust gas sensors upstream and downstream of the catalyst before being passed through a band-pass filter are created from the signal. The generated signal is input to the ECU 28, and it is checked whether the amount of rich correction performed after the fuel cut changes in accordance with the deterioration state of the catalyst and whether the rich correction is performed within an assumed period.

Next, a configuration for estimating the effective Ce amount of the catalyst will be described with reference to FIGS. 19 and 20. FIG.

FIG. 19 is a block diagram of a control model that performs catalyst deterioration diagnosis from the signal of the exhaust gas sensor before and after the catalyst, and corrects and controls the air-fuel ratio in consideration of the effective Ce amount.

　The control model shown in FIG.

A deterioration diagnosis unit 1901 receives exhaust gas sensor signals upstream and downstream of the catalyst, estimates the amount of effective Ce, and is composed of a frequency selection unit 1901a and a neural network 1901b. The frequency selection unit 1901a has the same configuration as the frequency selection unit 1301a shown in FIG. , three signals obtained by filtering the upstream exhaust gas sensor signal and three signals obtained by filtering the downstream exhaust gas sensor signal.

The neural network 1901b learns frequency information associated with the effective Ce amount of the catalyst, receives the signal after passing through the bandpass filter output from the frequency selection unit 1901a, and determines the effective Ce amount corresponding to the input signal. to estimate The estimated effective Ce amount is input to OSC estimation section 1903 . The OSC estimator 1903 estimates the oxygen storage capacity (OSC) using the estimated available Ce amount.

After that, the catalyst control unit 1304 and the air-fuel ratio correction amount calculation unit 1305 operate and output the post-correction target air-fuel ratio in the same procedure as in FIG.

FIG. 20 uses the control model shown in FIG. 19 to decompose the exhaust gas sensor signals upstream and downstream of the catalyst into three frequency bands using a band-pass filter, input them to the neural network 1301b, and output the effective Ce amount. 4 is a flowchart of processing for executing air-fuel ratio correction;

In step 111, the deterioration diagnosis unit 1901 receives input of exhaust gas sensor signals upstream and downstream of the catalyst. disassemble. In step 113, the neural network 1301b receives input of exhaust gas sensor signals upstream and downstream of the catalyst resolved into frequency bands. In step 114, the neural network 1901b, which has learned frequency information associated with the effective Ce amount of the catalyst, estimates the effective Ce amount from the input upstream and downstream exhaust gas sensor signals of the catalyst. At step 115, the OSC estimator 1303 estimates the oxygen storage capacity (OSC) of the catalyst according to the estimated effective Ce amount. At step 116, catalyst control unit 1304 calculates an air-fuel ratio correction amount based on the estimated oxygen storage amount. At step 117 , the air-fuel ratio correction amount calculation unit 1305 calculates the post-correction target air-fuel ratio using the air-fuel ratio target value and the air-fuel ratio correction amount calculated at step 116 .

As described above, the control device (ECU 28) of this embodiment performs frequency selection for extracting the frequency components of the output signal of the upstream exhaust gas sensor (air-fuel ratio sensor 20) and the output signal of the downstream exhaust gas sensor (oxygen sensor 22). a signal processing unit (neural network 1301b) that estimates the state of the three-way catalyst (exhaust purification catalyst 21) based on the output of the frequency selection unit 1301a; and an air-fuel ratio that calculates the air-fuel ratio from the state of the three-way catalyst. Since the correction amount calculation unit 1305 is provided, the deterioration of the catalyst can be appropriately estimated using the frequency component of the output of the exhaust gas sensor, and appropriate engine control can be performed without increasing harmful components in the exhaust gas. In other words, it is possible to diagnose the deterioration of the catalyst based on the relationship between the relative relationship of each frequency after passing through the band-pass filter and the state of the catalyst. degree can be diagnosed.

Further, the frequency selection unit 1301a includes a first filter that passes a first frequency band corresponding to the new state of the three-way catalyst and a second filter that passes a second frequency band corresponding to the deteriorated state of the three-way catalyst. is used to filter the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor, extracting the feature amount related to the deterioration of the catalyst, it is possible to perform highly accurate learning of the neural network 1301b.

Further, the frequency selection unit 1301a includes a first filter that passes a first frequency band corresponding to the new state of the three-way catalyst and a second filter that passes a second frequency band corresponding to the deteriorated state of the three-way catalyst. and a third filter that passes a third frequency band between the first frequency and the second frequency to filter the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor. can be determined in three states, and the air-fuel ratio correction amount can be finely set.

In addition, since the model selector is provided for selecting an oxygen storage capacity estimation model according to the state of deterioration of the three-way catalyst based on the output of the signal processor, the oxygen storage capacity can be accurately estimated according to the state of deterioration. It is possible to improve the correction accuracy of the air-fuel ratio.

In addition, the neural network 1301b determines whether the exhaust gas sensor is abnormal based on the degree of similarity between the output signal of the upstream exhaust gas sensor, the output signal of the downstream exhaust gas sensor, and the output signal of each exhaust gas sensor during normal operation. Together, sensor failure can be determined.

In addition, the neural network 1301b monitors the heater of the oxygen sensor 22 and determines whether the heater is abnormal based on the degree of similarity between the temperature change of the heater and the normal state. Therefore, it is possible to determine the deterioration of the catalyst as well as the failure of the heater.

Since at least one of the pass frequency of the first filter, the pass frequency of the second filter, and the pass frequency of the third filter is changed according to the operating point of the internal combustion engine, the accuracy of air-fuel ratio correction can be improved.

It should be noted that the present invention is not limited to the above-described embodiments, and includes various modifications and equivalent configurations within the scope of the attached claims. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and the present invention is not necessarily limited to those having all the described configurations. Also, part of the configuration of one embodiment may be replaced with the configuration of another embodiment. Moreover, the configuration of another embodiment may be added to the configuration of one embodiment. Further, additions, deletions, and replacements of other configurations may be made for a part of the configuration of each embodiment.

In addition, each configuration, function, processing unit, processing means, etc. described above may be realized by hardware, for example, by designing a part or all of them with an integrated circuit, and the processor realizes each function. It may be realized by software by interpreting and executing a program to execute.

Information such as programs, tables, and files that implement each function can be stored in storage devices such as memory, hard disks, SSDs (Solid State Drives), or recording media such as IC cards, SD cards, and DVDs.

In addition, the control lines and information lines indicate those that are considered necessary for explanation, and do not necessarily indicate all the control lines and information lines necessary for implementation. In practice, it can be considered that almost all configurations are interconnected.

Claims

A control device for controlling an internal combustion engine,
The internal combustion engine has a three-way catalyst arranged in an exhaust passage, an upstream exhaust gas sensor arranged upstream of the three-way catalyst, and a downstream exhaust gas sensor arranged downstream of the three-way catalyst,
a frequency selection unit that extracts frequency components of the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor;
a signal processing unit that estimates the state of the three-way catalyst based on the output of the frequency selection unit;
and a computing unit that computes an air-fuel ratio from the state of the three-way catalyst.
The control device according to claim 1,
The frequency selection unit includes a first filter that passes a first frequency band corresponding to the new state of the three-way catalyst and a second filter that passes a second frequency band corresponding to the deteriorated state of the three-way catalyst. and filtering the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor.
The control device according to claim 2,
The frequency selection unit uses the first filter, the second filter, and a third filter that passes a third frequency band between the first frequency and the second frequency to select the upstream exhaust gas sensor. and the output signal of the downstream exhaust gas sensor.
The control device according to claim 1,
The signal processing unit is a control device for estimating the state of the three-way catalyst using a neural network model having amplitude information of the signal output from the frequency selection unit as an input variable and a deterioration state of the three-way catalyst as an output variable. .
The control device according to claim 3,
The signal processing unit outputs an output signal of the upstream exhaust gas sensor filtered by the first filter, an output signal of the downstream exhaust gas sensor filtered by the first filter, and an output signal filtered by the second filter. The output signal of the upstream exhaust gas sensor filtered by the second filter, the output signal of the downstream exhaust gas sensor filtered by the second filter, the output signal of the upstream exhaust gas sensor filtered by the third filter, and the third filter and the output signal of the downstream exhaust gas sensor filtered by a first value indicating that the three-way catalyst is in a new state and a second value indicating that the three-way catalyst is in a completely deteriorated state. and a third value indicating that the three-way catalyst is in a moderately deteriorated state as an estimated state of the three-way catalyst.
The control device according to claim 1,
The control device, wherein the calculation unit has a model selection unit that selects an oxygen storage capacity estimation model according to the deterioration state of the three-way catalyst based on the output of the signal processing unit.
A control device according to claim 6,
The computing unit computes a corrected air-fuel ratio based on the oxygen storage capacity estimated by the selected oxygen storage capacity estimation model.
The control device according to claim 1,
The signal processing unit detects an abnormality of the upstream exhaust gas sensor and the downstream exhaust gas sensor based on the degree of similarity between the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor and the output signal of each exhaust gas sensor during normal operation. Control device to judge.
A control device according to claim 8,
A control device that stops estimating the state of the three-way catalyst when it is determined that the upstream exhaust gas sensor or the downstream exhaust gas sensor is abnormal.
The control device according to claim 1,
The downstream exhaust gas sensor has a heater that maintains the temperature,
The signal processing unit monitors the heater and determines whether the heater is abnormal based on the degree of similarity between the temperature change of the heater and the normal state.
A control device according to claim 10,
A control device that stops estimating the state of the three-way catalyst when it is determined that the heater is abnormal.
The control device according to claim 1,
Test data simulating the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor when the three-way catalyst is in a new state, a moderately deteriorated state, and a completely deteriorated state are input to the signal processing unit, and an appropriate A control device that has a function to check whether the air-fuel ratio is output.
The control device according to claim 3,
A control device that changes at least one of the pass frequency of the first filter, the pass frequency of the second filter, and the pass frequency of the third filter according to the operating point of the internal combustion engine.
A control method for an internal combustion engine executed by a control device, comprising:
The internal combustion engine has a three-way catalyst arranged in an exhaust passage, an upstream exhaust gas sensor arranged upstream of the three-way catalyst, and a downstream exhaust gas sensor arranged downstream of the three-way catalyst,
The control method is
The control device extracts frequency components of the output signal of the upstream exhaust gas sensor and the output signal of the downstream exhaust gas sensor,
the control device estimating the state of the three-way catalyst based on the amplitude of the extracted frequency component;
A control method in which the control device calculates an air-fuel ratio from the state of the three-way catalyst.