JP2001003734A - METHOD FOR OPTIMIZING NOx TRAP REGENERATIVE CYCLE - Google Patents

Abstract

PROBLEM TO BE SOLVED: To minimize an amount of Co and NOx discharged in the atmosphere with the fuel economical efficiency of a vehicle maximized by continuously operating a trap on an optimum condition to NOx conversion efficiency. SOLUTION: This regenerative cycle of an NOx trap 34 is optimized whose trap is filled up to a given ratio of its present capacity and brought into a completely empty state during purging of the trap. When trap capacity is substantially decreased and an actual filling time is reduced to a value lower than a lowest filling time or rendered equal to the lowest filling time, desulfurization of the trap is executed to restored trap capacity. Based on the voltage amplitude of a switching type oxygen sensor 38 and the time response of the sensor, a programmed computer 10 controls a filling and a purge time. The frequency of a purge for which direct relation to a trap NOx absorption capacity decrease factor is ideal is controlled such that the trap is prevented from exceeding the upper limit of NOx occlusion capacity.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for controlling a regeneration cycle of a downstream NOx trap in an emission control system of a lean burn vehicle.

[0002]

BACKGROUND OF THE INVENTION Conventional lean burn engine control systems are proportional to the measured air mass to periodically obtain an air / fuel ratio leaner than stoichiometric, thereby improving fuel economy. Includes an air-fuel ratio controller that supplies fuel to the engine. A typical three-way catalytic converter located in the exhaust passage of an engine is engine-generated in the presence of excess oxygen, such as during lean-burn engine operation
Because of the inefficiency of NOx conversion, some prior art systems include lean NOx in the exhaust passage downstream of the three-way catalytic converter.
Some disclose placing x traps to chemically occlude the remaining NOx generated during lean operation.

[0003] It is important to note that the trap's ability to store NOx is limited and its air-fuel ratio is richer than the stoichiometric air-fuel ratio of about 14.65, such as exposing the trap to an air-fuel ratio of less than about 13. By exposing the trap to an air-fuel mixture, the trap is periodically "purged" for stored NOx. During this purge event, excess HC and
CO is passed through the three-way catalytic converter, and, HC and CO flowing through the trap is reduced by reacting with NOx stored into harmless N ₂ and O _2. The amount of excess fuel required to release the stored NOx, that is, the amount of the fuel for purging, is a period during which the fuel for purging is supplied in the state of a rich air-fuel mixture while the mass air amount is almost constant, and the `` purge time ''. In many cases.

[0004] NOx passes through the trap and exits at the outlet of the exhaust pipe.
Therefore, the timing of each purge event must be controlled so that the trap does not exceed its NOx storage capacity. Since the air-fuel mixture enriched due to the purge event is accompanied by a reduction in fuel economy, the frequency of the purge is controlled so as to avoid purging of only partially filled traps. Is preferred.

[0005] It has been recognized in the prior art that the NOx absorption capacity of a trap is itself a function of many variables. These variables include trap temperature, trap history, degree of sulfation, and the degree of thermal damage or damage to the NOx absorbing material of the trap due to excessive heat. For example, referring to U.S. Patent No. 5,437,153, it states that when a trap approaches its maximum capacity,
That the speed of absorbing NOx may start to fall
It has been described. Further, U.S. Pat.No. 5,437,153 uses a nominal NOx storage capacity that is significantly less than the actual NOx storage capacity of the trap, thereby providing a trap with full instantaneous NOx absorption efficiency, i.e., the amount of NOx stored. States that the trap will absorb any NOx produced by the engine as long as it remains below its nominal capacity. A purge event is planned so that the trap is regenerated whenever the accumulated NOx produced by the engine reaches the nominal trapping capacity.

[0006]

The actual amount of NOx stored in the trap will vary depending on the NOx concentration of the gas supplied to the engine, exhaust flow rate, atmospheric humidity, trap temperature, and other variables. Thus, both the capacity of the trap and the actual amount of NOx stored in the trap are complex functions of many variables. It is desirable to diagnose and control trap purge and fill events to ensure that the trap always operates at its optimal conditions.

[0007] When the engine is operated with a fuel containing sulfur, the sulfur is absorbed by the trap, which leads to a reduction in both the absolute NOx storage capacity of the trap and the instantaneous NOx absorption efficiency of the trap. If the sulfation of such traps exceeds a certain threshold, the absorbed SOx must be "burned off" or desorbed during the desulfurization event. The trap temperature is then raised to about 650 ° C. or higher in the presence of excess HC and CO. For example, U.S. Pat. No. 5,746,049 discloses that when the engine is operated with a rich air-fuel mixture, the exhaust gas upstream of the NOx trap is
At least trap temperature by introducing secondary air
A trap desulfurization method is described that includes increasing the temperature to 650 ° C. and using the resulting exothermic reaction to increase the trap temperature to the extent desired for purging the trap for SOx.

[0008]

According to the lean NOx control method of the present invention, the trap status, that is, "soundness" is diagnosed,
The NOx stored in the trap is then adjusted so that the parameters for purging the trap, the impairment rate of the NOx storage capacity, the purge time and the purge strength can be adjusted online during engine operation in the vehicle. Use information related to volume and maximum NOx storage capacity of the trap. Where NOx
The storage capacity impairment rate is the speed at which the NOx storage capacity of the trap decreases during the NOx filling process. In addition, the decision to purge the trap for sulfur, i.e., to schedule a desulfurization event, is determined by the NO of the diagnosed trap.
This is based on the decrease in x storage capacity and the associated increase in impairment of NOx storage capacity. Thus, the trap is operated continuously at optimal conditions for NOx conversion efficiency, thereby minimizing CO and NOx emissions to the atmosphere while maximizing vehicle fuel economy. Intelligent desulfurization of the trap ensures that the NOx conversion efficiency of the trap is always maintained above a given minimum.

More specifically, according to the present invention, the trap is monitored for loss of NOx storage capacity, and closed loop control of trap desulfurization as well as trap purge frequency and depth is performed. Based on the voltage amplitude of the switching oxygen sensor and the sensor response time, a programmed computer controls the trap fill and purge times. Deterioration of fuel economy and HC and
Ideally, it is directly related to the impairment rate of the trap's NOx absorption capacity, so that the trap is not filled beyond its NOx storage capacity limit, while avoiding incomplete filling of the trap with increased CO emissions Is controlled.

Further, according to the present invention, the trap comprises:
Based on the x capacity loss rate, it is filled to a predetermined percentage of its existing capacity, and then completely emptied during purging. If the capacity of the trap decreases due to deterioration, a closed loop
A purge optimization routine generates an adjustment multiplier that is used to adjust the NOx storage capacity impairment rate to obtain enough NOx storage capacity to fill the trap to the desired percentage of its capacity. When the trapping capacity is significantly reduced, as indicated by the actual NOx storage capacity impairment rate being equal to or higher than the predetermined maximum NOx storage capacity impairment rate, the trapping capacity is restored to its original value. Is performed. NO NO for predetermined number of trap desulfurization operations
If performed without any increase in the x-capacity impairment rate, the trap must be replaced, which is signaled to the driver by an indicator.

The default or initial value of the impairment rate of the NOx storage capacity of the trap is determined by a function of the engine system and the trap. The functionalization gives the values of the trap filling rate and the optimal NOx storage capacity impairment rate as a function of engine load or mass air volume. The operating air-fuel ratio where the trap is used does not change significantly, and changes in engine speed do not significantly affect the NOx storage capacity impairment rate of the trap. Thus, the primary variable for the trap NOx storage capacity impairment rate is the root of the engine (eg, the square root) where the index of engine load or mass flow is small.

The objects, features and advantages of the present invention, such as those set forth above, will be readily understood from the following detailed description of the best mode for carrying out the invention, when taken in conjunction with the accompanying drawings.

[0013]

DETAILED DESCRIPTION OF THE INVENTION Referring now to the drawings, and first to FIG. 1, a powertrain control module, generally designated PCM, indicated generally by the numeral 10 is shown.
Is an electronic engine controller, including ROM, RAM and CPU as shown. The PCM controls a set of injectors 12, 14, 16 and 18 that inject fuel into the four cylinders of the internal combustion engine 20. The fuel injector is of a general construction, and the controller 10
Are arranged to inject the correct amount of fuel into the corresponding cylinder as determined by The controller 10 sends a fuel injector signal to the injector in order to maintain the air-fuel ratio (AFR) determined by the controller 10. An air gauge or mass flow sensor 22 is located at the air inlet of the engine manifold 24 and provides a signal regarding the mass air flow determined by the position of the throttle valve 26. The air mass signal is used by controller 10 to calculate an air mass value indicative of the mass of air entering the intake system per unit time. Heated exhaust gas oxygen
A HEGO (abbreviated for short) sensor 28 detects the oxygen content of the exhaust generated by the engine and sends a signal to the controller 10. HE
The GO sensor 28 is used for controlling the engine air-fuel ratio particularly during the stoichiometric air-fuel ratio operation.

The exhaust system has one or more exhaust pipes, and converts exhaust gas generated by combustion of the air-fuel mixture in the engine into a general direct-coupled three-way catalytic converter (three-way catalyst 30). ). The three-way catalyst 30 contains a catalytic substance that chemically changes exhaust generated by the engine in order to generate catalytically reacted exhaust. The exhaust gas after the catalytic reaction passes through an exhaust pipe 32, is sent to a downstream NOx trap made of the above-mentioned substance, and is then discharged to the atmosphere through a tail pipe.

Downstream of the trap 34 is a second HEGO sensor 38, which supplies signals to the controller 10 for diagnosis and control according to the present invention. The second HEGO sensor 38 operates during the normal stoichiometric air-fuel ratio closed loop limit cycle operation.
It is used to monitor the HC conversion efficiency of the three-way catalyst 30 using a well-known method of comparing the signal amplitude of the second HEGO sensor 38 with that of the first HEGO sensor 28. The trap internal temperature sensor 42 generates an output signal representing the instantaneous temperature T of the trap 34. Further, other information regarding the engine performance, such as the camshaft position, the crankshaft position, the angular velocity, the throttle valve position, the temperature, and the like, are also sent to the controller from other sensors (not shown). Information from these sensors is used by controller 10 to control the operation of the engine.

A typical response of voltage to air / fuel ratio for a switching oxygen sensor, such as the second HEGO sensor 38, is shown in FIG. The output voltage of the second HEGO sensor 38 switches between a low level and a high level when the exhaust gas mixture switches from lean to rich at a stoichiometric air-fuel ratio of approximately 14.65. Since the air-fuel ratio is lean during the charging time, NOx generated by the engine is sent to the trap 34 through the three-way catalytic converter 30 and the exhaust pipe 32,
It is absorbed there.

A typical operation of a trap purge cycle
The movement is shown in FIG. The top waveform (Figure 3 (a))
Lean charge for three different purge times 1, 2 and 3
Filling time t _FAnd rich purge time t_PShows a relationship with
You. The second waveform (Fig. 3 (b)) shows three purge times.
The response of the second HEGO sensor 38 is shown. G
CO passing through the wrap and affecting the downstream sensor 38
The amount of HC shall indicate the efficiency of the trap purge event.
Used. Peak voltage level of the downstream oxygen sensor
Are trapped in NOx and O_TwoThe amount of
It is shown. For short purge time 1, trap is N
Oxygen sensor is not completely purged for Ox
Response is extremely small, resulting in CO emissions to the atmosphere
And closely related response spikes of the second HEGO sensor
Is small. In this case, the sensor peak voltage V_P
Is the reference voltage V_refHas not reached. Moderate, or optimal
In the simple purge time 2, the output V of the second HEGO sensor is_PIs the standard
Voltage V_refAnd this is an extremely small amount of air that is acceptable
Traps have been purged by the required amount
Indicates that the Long purge time 3
About the peak voltage V of the second HEGO sensor_PIs the reference voltage
V_refThe trap has been completely purged,
3 (d).
Which results in excessive CO emissions to the atmosphere as shown in
Absent.

Data on the output voltage of the second HEGO sensor
The data capture window shows the waveform shown in Fig. 3 (c).
It works like During this window, PCM is the second HEG
The response data of the O sensor 38 is obtained. FIG. 4 shows FIG.
Sensors 38 for three levels of indicated purge times
FIG. Δt_{twenty one} Is based on the sensor output voltage
Reference voltage V_refIs the length of time that was exceeded. Reference voltage
V_refSensor peak voltage V less than_Pabout,
PCM 10 is the sensor saturation time t_satTo t_sat = t_{sat ref}From
t_sat = 0 by linear extrapolation until
Make the values smoothly continuous. PCM 10 has the relationship shown in FIG.
Used, and as indicated there, the sensor saturation time t_satThe
Sensor peak voltage V_PProportional to

FIG. 5 shows the normalized oxygen sensor saturation time.
t_satAnd purge time t_PThe relationship is shown below. Sensor saturation time
t_satMeans that the signal of the second HEGO sensor is V_refIs over
Is the normalized amount of time and Δt_{twenty one}/ Δt
_{21 norm}(Where Δt_{21 norm}Is equal to the normalization factor).
Sensor saturation time t_satIs the desired value t_{sat desired}By
Normalized. Given filling time t_F And of the trap
For conditions, the optimal purge time value t_{P sat # desired}Exists
Exist. That value will result in excess HC and CO emissions to the atmosphere.
And maintain a positive NOx trap efficiency.
Normalized optimal saturation time t _sat・ = ・ 1 Sensor saturation
Time t_satIs t_sat・ If> 1, the purge time is too long
And need to be shortened. Sensor saturation time t_sat
・ Is t_sat・ If <1, the purge time is too short.
Need to be longer. In that way, the second HEGO
Based on the output of sensor 38, the trap purge
Control can be realized.

FIG. 7 shows the nominal relationship between purge time t _P and fill time t _F for a given operating condition of the engine and a given condition of the trap. The two suboptimal purge times t _Pusbopt1 and t _Pusbopt2 are fixed filling times t
_This corresponds to under-purging or over-purging of the trap 34 for the _FT . A purge time t _P for optimally purging the trap with respect to NOx stored in the trap is indicated as t _PT . This point is the target or desired purge time t _sat = t
Corresponds to _{sat desired} . This purge time minimizes CO emissions to the atmosphere during the fixed fill time _tFT . The result of this procedure is a determination of the stored oxygen purge time, t _Posc , which is related to the amount of oxygen stored directly in the trap. Oxygen can be stored directly, for example in the form of cerium oxide. Stored oxygen purge time t _{P osc} by performing optimization of t _P either two or or more optimal purge time than or extrapolate to a point t _F · = 0, or near a point t _F = 0 ,
You can decide. The operating point T2 is deliberately set to t _{F T2} <
It is obtained by finding t _{P T2} through optimization as t _FT .

[0021] Figure 7a illustrates the optimization of the fill time t _F. For a given filling time _tFT , an optimal purge time _tPT is determined as shown in FIG. Then fill time, initial value t slightly changing step to a smaller value t _FB than _FT, and the initial value t _FT than slightly larger value t _FA
It is shaken by changing the step to. by t _F
To determine the change in t _P, optimization of purge time its 3
Applies to all points T, A and B. from A to T of t _P,
Then, the change from B to T is evaluated. In FIG. 7a, the change from B to T is greater than the change from A to T. The absolute values of these differences are controlled to fall within a predetermined tolerance DELTA # MIN, as described more fully in connection with FIG. The absolute value of the difference is proportional to the slope of the curve showing the relationship between t _P and t _F. Optimization process defines the operating point T as the shoulder of the curve showing the relationship between t _P and t _F. Time t _{P sat}
Represents the saturation value of the purge time for an infinitely long filling time.

Four trucks having different NOx and oxygen storage amounts
Purge time t_PAnd filling time t_F Most
The result of the adaptation routine is shown in FIG. Purge time
t_PAnd filling time t_FBoth were shown in FIGS. 7 and 7a
Optimized using techniques. Points determined by FIG.
Is shown as the optimal operating point T1 and purge for that point
Time is t_{P T1}And the filling time is t_{F T1}It is. "1" is
Indicates that the trap has not deteriorated, and this is referred to as state A.
I do. Trap is sulfur poisoning, thermal damage, or other factors
, The trap state becomes B, C and
You will reach D. Nearly steady engine condition exists
The purge and fill optimization routine runs continuously.
Is performed. Optimal operation according to trap conditions B, C and D
The rolling states T2, T3 and T4 will be reached. t_{P T1},
 t_{P T2}, t_{P T3}And t_{P T4}NOx saturation level reflected in
And the oxygen storage related purge time t_{P osc T1}, t_{P osc T2}, t
_{P osc T Three}And t_{P osc T4}Are both trapped
Depending on the trap, and typically
Value will be reduced. About the NOx part of the purge
The purge fuel is t_{P NOx}= t_PT -t_{P osc}With a value equal to
is there. Purge fuel should be available for a given operating condition
It is expected to be equivalent to the time. The controller 10
The gin 20 is allowed to operate at a predetermined rich air-fuel ratio
Adjusting the time limits the actual purge fuel.
In order to simplify the explanation here, the purge time
To purge fuel at the operating conditions assumed for, etc.
Value. NOx thus occluded
And the purge time required for the stored oxygen is directly determined.
It will be used for diagnosis and control.

FIG. 9 shows the relationship between the NOx purge time t _{P NOx} and the absorption capacity of the NOx trap. In states A, B, and C, the NOx absorption efficiency, trapping capacity, and fuel economy are determined to be within the allowable range, and state D is determined to be outside the allowable range. Thus, as State D is approached, a trap desulfurization event is planned to regenerate the NOx trapping capacity and reduce fuel consumption associated with high NOx purge frequencies. Changes in t _{P osc} can provide additional information about the aging of the trap through changes in oxygen storage.

FIG. 10 shows the purge time t_PAbout optimization of
It is a flowchart. The purpose of this routine is given
Filling time t_FAbout rich purge of air-fuel ratio
Optimizing the spikes. This routine will be
The software for system optimization described with reference to FIG.
Software. At decision block 46, the password
The status of the storage flag is checked and the flag is set.
If so, lean as shown in block 48.
NOx purge is performed. Purge flag is trapped
Is set when the filling of is completed. For example, the hula
Is the time when that purge planning method is used
In block 136 of FIG.
You. In block 50, the peak voltage V_PAnd if you have
Is the transition time t₁And t_TwoOxygen sensor (E
GO) voltage is specified in the capture window
Is sampled through. As shown in FIG. 3 (c),
A window captures changes in the EGO sensor waveform. Judgment
It is V determined by block 52_P > V_refIf it is
Has a sensor saturation as shown in blocks 54 and 56
Time t_satIs Δt_{twenty one}, That is, the EGO sensor voltage is V_refOver
It is said that it is in proportion to the time that was required. V_P <V_refIn
The linear extrapolation function, as shown in block 58.
Using t_satIs determined. For this function, see FIG.
And t_satThe peak voltage V_PProportional to
By t_satIs required. This is,
 V_P > V_refFrom V_P <V_refSmooth transition to the case
Line and return control as shown in block 60
Suitable continuous positive and negative deviation function t_{sat error}(k)
I can. Where the deviation function t_{sat error}(k) is sensor saturation
Desired value of time t_{sat desired}From the actual sensor saturation time t_sat
Equivalent to subtracting And the deviation function t
_{sat error}(K) is a block 62 showing the desired sensor saturation.
Sum time t_{sat de sired}Is normalized by dividing by.

As a result, the normalized deviation function t
_{sat error # norm}(k) is like a PID (proportional-integral-derivative) controller
It is used as an input to a simple feedback controller. In block 64
As shown, the output of the PID controller is during the trap purge
Multiplicative correction factor PURGE # MUL for the interval. t
_{sat error # norm}Direct monotonic between (k) and PURGE # MUL
Have a relationship. t_{sat error # norm}If (k)> 0,
Pump is under purge and NOx purge
PURGE # MUL is its reference value to provide more CO
Must be increased from. t_{sat error # norm}(k) <
 If 0, the trap is overpurge and the NOx
PURGE # MUL to reduce CO supply for
It must be reduced from that reference value. This result
As a result, a new value for the purge time is
 Is t_P(K + 1) = t_P(K) × PURGE # MUL. At the time of purging
Optimization between the old and the new, as shown in blocks 68 and 70
The absolute value of the deviation of the purge time value is smaller than the allowable tolerance
It continues until it becomes. If | t_P(k + 1)-t_P(k) |> = ε
In some cases, the PID feedback control loop is_P
Is not within the allowable range ε. Therefore, as shown in block 70
The new par calculated in block 66
The delay time is delayed until the condition in block 68 is satisfied.
Used in the storage cycle. Filling time t_FIs the best par
Di time t_PT until_PDuring the optimization routine, Equation 2
And adjusted as needed. | t_P(k + 1)-t
_PIf (k) | <ε, the purge time optimization converges
The purge time, as shown in block 72.
The new value is stored, and the optimization process is performed as shown in FIG._FOptimal
It will be shifted to a routine for conversion. Purge time t_POnly
Instead of being used during the purge event (see Figure 3)
Adjust the relative richness of the air-fuel ratio in a similar manner.
You can also.

FIG. 11 is a flowchart for system optimization, including optimization of both purge and fill times. Optimization of the fill time is performed only when the engine is operating at approximately steady state, as indicated by block 74. In this regard, the substantially steady state means that the rate of change of certain engine operating variables, such as engine speed, load, intake air amount, ignition timing, EGR amount, is maintained below a predetermined level, Characterized. In block 76, one fill step increment FILL # ST
EP is equal to STEP #SIZE, if FILL # STEP> 0, increase the fill time. STEP_SIZE is adjusted with _respect to the storage capacity utilization rate Rij as shown in FIG.

At block 78, the purge time optimization routine described above in connection with FIG. 10 is executed. This will optimize the purge time t _P for a given fill time. PURGE # MUL at the end of the purge optimization routine executed in block 78 is CTRL # START
And the fill time multiplier FILL # MUL is incremented by FILL # STEP, as shown in block 80. The filling time step change FILL # MUL is multiplied t _F at block 82 is promoted. In block 84, the purge optimization routine of FIG. 10 is executed for a new filling time t _F (k + 1). The PURGE_MUL at the end of the purge optimization routine performed in FIG. 10 is stored at block 86 as CTRL_END. The amount of change in the purge multiplier CTRL # DIFF = ABS (CTRL # END CTRL # START) is also stored in block 86, and in block 88 the reference value DELTA #
Compared to MIN. DELTA # MIN corresponds to the tolerance stated in FIG. 7a, and CTRL # END and CTRL # START
corresponding to two values of t _P obtained at a of A and T or B and T. If the change in the purge multiplier is greater than DELTA # MIN, the sign of FILL # STEP will be
In order to be able to search for the optimal filling time in the opposite direction,
Can be changed. If the change in the purge multiplier is smaller than DELTA # MIN, search the optimum fill time t _F is, as indicated in block 92, is continued in the same direction. In block 94, FILL # MUL is incremented by the selected FILL # STEP. Then, at block 96, the filling time t _F (k
+1) is corrected by multiplying FILL # MUL. As a result, selection of the optimum t _P as the operating point and continuous destabilization at this point are performed. If the engine does not experience near steady state during this procedure, the fill time optimization process is aborted and the fill time from Equation 2 (described below) is used, as shown at block 74.

FIG. 12 shows the desulfurization of a trap according to the present invention.
Is shown. At block 100,
For a trap that does not degrade under given operating conditions
Reference value t representing purge time_{P NOxref}Looks up
Read from the table. t_{P NOxref}Is the intake air volume,
Can be a function of air-fuel ratio and other parameters.
come. In block 102, the latest purge time t_P(K)
Is called and t_{P NOxref}Minus the specified tolerance TOL
Is compared to So if t_P(k) <t_{P NOxref}−TOL
If so, a trap desulfurization event is planned. For desulfurization
Is slightly less than the theoretical value with an air-fuel ratio of, for example, 0.98λ.
To about 650 ° C for about 10 minutes.
Heating is included. The desulfurization counter D
Reset at block 104 and shown in block 106
Each time the desulfurization process is performed as
It is. After the desulfurization process has been completed,
As shown in FIG.
Is determined. New purge time t_P(k + 1) is block 110
At the reference time t_{P NOxref}From the specified tolerance TOL
Compared to the subtracted value, t_P(k + 1) <t_{P NOxref} − TOL
Is determined by the determination of block 112,
Then, at least two additional desulfurization events are performed. It
However, if the trap still does not fall within the acceptable range,
Fault indicator lamp (malfunction indicator la
mp for MIL) is illuminated and as shown in block 114
The trap should be replaced with a new one. t_P(k)
> = t_{P NO xref} − If TOL, trap 34 immediately
Has not deteriorated enough to require service and normal operation
Will be resumed.

A NOx purge event is scheduled when a given storage capacity that is lower than the actual storage capacity of the trap 34 is filled or used with the stored NOx. Oxygen is trapped in oxygen or NOx in the form of cerium oxide.
And the sum of those two substances is the amount of oxidant stored. FIG. 13 illustrates the relationship between the oxidant stored in the trap 34 and the time during which the trap 34 is exposed to the inflowing NOx. NOx storage occurs at a slower rate than oxygen. The optimal operating conditions for the NOx generation time correspond, for this figure, to the "shoulder" of the curve, ie a relative NOx generation time of about 60-70%. 100% of horizontal axis
Corresponds to the saturated NO _x storage capacity of the trap 34. Also shown are values for stored NOx and stored oxygen. The storage capacity utilization speed R _ij is the initial slope of this curve, and is a value obtained by dividing the ratio of the stored oxidant by the ratio of the NOx generation time.

[0030] Figure 14, except that a graph of the relative purged fuel amount with respect to the relative fill time t _F, is similar to FIG. The storage capacity utilization speed R _ij (purge fuel ratio / fill time ratio) is expressed as the initial slope of this curve.
For a given calibration value of air-fuel ratio, EGR and ignition timing (SPK) at a given speed and load point, the relative N
Ox generation amount is linearly related to the relative fill time t _F. Figure
FIG. 14 illustrates the relationship between the amount of purge fuel containing HC and CO added to the trap and the time the trap is exposed to the incoming NOx. Purge fuel is divided into those required for purging stored oxygen and those required for purging stored NOx.

The reduction in the NOx absorption capacity of the trap is as follows:
It is represented by Equations 1 and 2.

(Equation 1)

(Equation 2)The storage capacity utilization rate RS before modification, ie, before modification (%)
Is given by Equation 1. This is the engine speed and
As a function of load, all actions in which the trap filling operation occurs
For a moving cell, the cell filling rate R_ij (% / s) time
Represents the total found. Relative cell filling speed R_ij
(Purge fuel ratio / fill time ratio)
Filling time t corresponding to 100% filling_FChange in purge time
Obtained by dividing the quantity. Where Equation 1 is the reference value
Equation 2 is a real-acting formula with modifiers
It is. The modifier in Equation 2 is
M ₁M about air-fuel ratio_Two, M about EGR amount_Three, That
Then the ignition advance about M_FourIt is. Individual R_ijBut tiger
Yes, but not fully utilized
The points are summed up to a value less than 100%. this
For storage capacity, the sum of the time spent in all cells
Total t_FIs the trap filling time. The result of this calculation
Is the effective trap capacity utilization RSM given by Equation 2.
(%). Reference filling rate for a given area
Is the time t spent in that area_KMultiplied by_Two,
M_ThreeAnd M_Four, And continuously summed.
The sum is the trap temperature modifier M₁Modified by (T)
It is. Trap when modified sum RSM reaches 100%
Is almost completely filled with NOx, and a purge event is planned
Is done.

FIG. 15 shows a map of the data stored for the reference trap filling rate _Rij . The entire system, consisting of the engine and the exhaust gas purification system including the three-way catalyst and the trap, is functioned on a matrix map of engine speed and load. As typical calculations, typical calibration values for the air-fuel ratio (AFR), EGR, and spark advance are used. The trap temperature T _ij is recorded for each area represented by the rotation speed and the load. Figure 16a
To 16d are the air-fuel ratio, EGR, ignition advance SPK, and trap temperature T for which the trap filling speed _Rij was determined in FIG.
_{For ij} , typical functionalization conditions are listed.

If the actual driving condition of the vehicle is different from the functioning condition shown in FIG. 16, the modifiers M ₁ (T), M
Corrections are made for ₂ (air-fuel ratio), M ₃ (EGR amount), and M ₄ (ignition advance). The correction for M ₁ (T) is shown in FIG. Traps NOx absorption capacity reaches the maximum value under the optimum temperature T _o (350 ° C. in this embodiment), as shown, the trap temperature T is below or it exceeds the optimal temperature T _o Occasionally, corrections are made to reduce the NOx storage capacity of the trap.

Modifications to the modifiers M ₂ , M ₃ and M ₄ are shown in FIG.
To 18c. These corrections are applied when the actual air-fuel ratio, the actual EGR amount, and the actual ignition advance differ from the values used in the functionalization of FIG.

FIG. 19 is a flowchart showing a procedure for determining the reference filling time of the trap 34 when the trap 34 should be purged. If the purge event is complete (determined at block 120) and the engine is running lean (determined at block 122), block 1
As indicated by 24, the trap is filled. The charging time is based on the prediction of the NOx storage capacity impairment _Rij , which is appropriately modified for the air-fuel ratio, EGR amount, ignition advance and trap temperature. At block 126 the engine speed and load are read, and at block 128 the reference fill rate R _ij is obtained from a look-up table (FIG. 15) using engine speed and load as a starting point. Trap temperature, engine air-fuel ratio, EGR quantity, spark advance and the time t _K is the block 1
16a-16d and used at block 132 to calculate the time weighted sum RSM based on the time spent in the region determined by the given engine speed and load. If the RSM is close to 100%,
As shown in blocks 134 and 136, a purge event is scheduled. Otherwise, the trap filling process continues at block 122. The filling time determined in FIG. 19 is the reference filling time. This will change when the trap is sulphided or thermally damaged. However, in the foregoing procedure (FIGS. 7a, 8 and 11), the optimal filling time was determined by the destabilization process, the need for desulfurization was determined, and a determination was made whether the trap had been thermally damaged. .

The planned value of the purge time t _P is the oxygen purge time
Both components of t _{P osc} and NOx purge time t _{P NOx} must be included. Thus, t _P = t _Posc + t _{P NOx} .
Controller 10 has a look-up table that provides t _{P osc} , a function that is strongly dependent on temperature. For traps containing cerium oxide, t _{P osc} is Arrheniu
s) according to the formula t _{P osc} = C _exp (−E / kT). Where C is a constant determined by the type and state of the trap, E is the activation energy, and T is the absolute temperature.

Having described in detail the best mode for carrying out the invention, those skilled in the art to which the invention pertains will appreciate
Various alternatives and embodiments will be envisaged for implementing the invention as defined in the appended claims.

[0038]

According to the present invention, the trap is operated continuously at optimal conditions for NOx conversion efficiency, thereby maximizing vehicle fuel economy and reducing CO and atmospheric emissions.
NOx emissions can be minimized. Furthermore, it can be ensured that the high-performance desulfurization of the trap is constantly maintained above the given minimum value of the NOx conversion efficiency of the trap.

[Brief description of the drawings]

FIG. 1 is a schematic diagram of an engine control system embodying the present invention.

FIG. 2 is a graph showing a voltage response of an oxygen sensor with respect to an air-fuel ratio.

FIG. 3 shows (a) engine air-fuel ratio, (1) short purge time, (2) intermediate purge time, and (3) long purge time.
FIG. 6 is a graph comparing (b) the response of the downstream oxygen sensor, (c) the data capture window of the EGO sensor, and (d) the amount of CO emissions to the atmosphere.

FIG. 4 is a more detailed graph showing the response of the oxygen sensor to time for (1) short purge time, (2) intermediate purge time, and (3) long purge time.

FIG. 5 is a graph of normalized oxygen sensor saturation time t _sat as a function of purge time t _P.

[6] if the peak voltage V _P of the oxygen sensor is smaller than the reference voltage V _ref, the graphs of the normalized oxygen sensor saturation time t _sat as a function of the peak voltage V _P.

FIG. 7 shows the relationship between trap purge time t _P and trap fill time t _F, and shows the optimal purge time t _PT for a given fill time t _FT and two suboptimal purge points 1 and 2; It is a graph.

FIG. 7a shows the relationship between trap purge time t _P and fill time t _F when the purge time is optimized for all fill times, with the optimal purge time t _PT and fill time t _FT being the preferred system. Indicates the operating point T and is located on the response curve 2
The two suboptimal points A and B are also the graphs shown.

FIG. 8 shows the relationship between purge time t _P and fill time t _F for four different trap operating conditions as the degradation of NOx trap performance progresses gradually, and the total purge time
shows the outer挿化been purge time t _{P osc} for oxygen storage portion in t _P, a graph.

FIG. 9 is a graph showing the relationship between NOx trapping capability and purge time for four different trap operating condition states as degradation progressively progresses due to sulfation, thermal damage, or both.

FIG. 10 is a flowchart for optimizing a trap purge time t _P.

FIG. 11 is a flowchart for system optimization.

FIG. 12 is a flowchart for determining whether or not desulfurization of a trap is necessary.

FIG. 13 is a graph showing the relationship between the relative amount of oxidant stored in the trap and the relative time of exposure of the trap to NOx flowing into the trap.

FIG. 14 is a graph of relative purge fuel versus relative charge time.

FIG. 15 is a map diagram of a reference trap filling speed R _ij (NOx storage capacity impairment rate) for various rotation speeds and loads at given map values of temperature, air-fuel ratio, EGR amount, and ignition advance angle. is there.

FIG. 16a shows the relationship between the trap filling speed R _{ij and} FIG.
FIG. 7 is a map diagram showing a function state of an air-fuel ratio determined in FIG.

FIG. 16b shows the trap filling speed R _ij in FIG.
FIG. 4 is a map diagram showing a functioning state of an EGR amount determined in FIG.

FIG. 16c shows the trap filling speed R _ij in FIG.
FIG. 7 is a map diagram showing a function state of an ignition advance angle determined in FIG.

FIG. 16D shows the trap filling speed R _ij in FIG.
FIG. 6 is a map diagram showing a function state with respect to a trap temperature determined in FIG.

FIG. 17 is a diagram showing how the modifier of the trap capacity impairment ratio changes with temperature.

FIG. 18a is a diagram showing how the air-fuel ratio modifier changes when the value of the air-fuel ratio changes from the value functionalized in FIG. 16;

FIG. 18B is a diagram showing how the EGR amount modifier changes when the value of the EGR amount changes from the value functionalized in FIG. 16;

18c is a diagram showing how the ignition advance modifier changes when the value of the ignition advance changes from the value functionalized in FIG. 16; FIG.

FIG. 19 is a flowchart for determining a time for planning trap purging.

[Explanation of symbols]

 34 Lean NOx trap 38 Oxygen sensor

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) F02D 45/00 368 B01D 53/36 101A (72) Inventor Joseph Richard Eislalk Bloomfield Hills, Michigan 48304, United States of America Great Oaks Drive 776 (72) Inventor Garth Michael Meyer 48128, Dearborn N.S.

Claims (11)

[Claims] 1. A method for optimizing a purge time of a lean NOx trap disposed in an exhaust passage upstream of an oxygen sensor, comprising: detecting a sensor output voltage; and a purge time correction factor related to a characteristic of the sensor output voltage. And calculating the next purge time as a function of the product of the purge time correction factor and the current purge time. 2. The method of claim 1, wherein said purge time correction factor is based on a deviation between a desired saturation time value and a calculated saturation time value. 3. The method of claim 2, wherein the sensor output voltage is sampled in a window to determine a peak voltage. 4. The calculated saturation time value is based on a time when the peak voltage is greater than a reference voltage.
the method of. 5. The method of claim 3, wherein said calculated saturation time is based on the value of said peak voltage. 6. The calculated saturation time value is based on a value of the peak voltage when the peak voltage is equal to or less than a predetermined reference voltage, and otherwise, the peak voltage exceeds the reference voltage. 4. The method of claim 3, based on time. 7. The method of claim 6, wherein said deviation is input to a feedback controller that generates said correction factor. 8. The method of claim 7, wherein there is a direct monotonic relationship between said correction factor and said deviation. 9. The method of claim 7, wherein the optimization of the purge time is continued until the absolute value of the difference between the current and subsequent purge time values is within an acceptable range. 10. The method of claim 9, wherein said deviation is normalized to a reference saturation value. 11. The method of claim 10, wherein said sensor is a switching sensor.