JPH0794808B2 - Lean burn engine control device and control method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an internal combustion engine, and more particularly to a means for suitably controlling an air-fuel ratio and an ignition timing of a lean burn engine in a target region and means therefor. It is a thing.

[Conventional technology]

A conventional lean burn control device detects an air-fuel ratio controlled by a lean sensor that produces an output signal corresponding to the oxygen concentration in exhaust gas, as described in JP-A-61-279747, and this is the target. It was designed to be closed-loop controlled to be the same as the lean air-fuel ratio.

This target lean air-fuel ratio is, as shown in FIG. 3, necessary for clearing the engine's misfire limit air-fuel ratio and exhaust emission control value.
The air-fuel ratio narrowed by the air-fuel ratio determined from the NOx limit, that is, the specific air-fuel ratio within the control target range was set.

[Problems to be solved by the invention]

In the above-mentioned conventional technique, the change over time of the engine is not sufficiently taken into consideration, and the set lean target air-fuel ratio is a constant value regardless of change over time. It is predicted that the misfire limit will shift to the rich air-fuel ratio side if there is a change over time, a change in fuel properties, a change in atmospheric conditions, etc. Therefore, in the past, the lean target air-fuel ratio was set in advance by taking this into account. It was set from the left side of the area. Therefore, there are problems that fuel efficiency is sacrificed, NOx emissions increase as the vehicle weight increases, and it becomes difficult to meet the regulation value.

An object of the present invention is to provide a lean burn control device capable of always suitably determining a control target region and controlling the target target region regardless of the above-described change with time.

[Means for solving problems]

The above objective is accomplished by detecting the misfire limit and the NOx limit. Both of these have characteristics as shown in FIG. 4 with respect to the air-fuel ratio and the ignition timing. The misfire limit () and the advance side misfire limit () of the ignition timing constitute a misfire limit, and the NOx limit becomes leaner as the ignition timing advances. As the lean burn reduces the engine torque with respect to the vehicle weight, the NOx limit shifts to the right in FIG. 4, and the control target area surrounded by both becomes smaller, making control difficult.

There are two misfire cycles: a poor ignition cycle in which flame kernels are not generated even if sparks fly (Fig. 4) and a poor flame propagation cycle in which flame does not grow even after ignition (Fig. 4). Is a cycle in which almost no combustion flame is produced. Therefore, the misfire cycle can be detected from the intensity of combustion light or the combustion temperature. On the other hand, NOx is known to be a parameter of combustion temperature. Also NOx
Has a wavelength of 5.3 μm, and the light intensity at this wavelength is said to be a function of NOx concentration.

The features of the present invention include (a). Misfire state determination means for determining the misfire state of the engine; (b). NOx concentration determination means for determining the concentration of NOx generated in the engine; (c). (1) A function of determining whether or not the current combustion state is in an allowable combustion region defined by the misfire limit and the NOx concentration limit, based on both information from the misfire state determination means and the NOx concentration determination means; A function to calculate a fuel control signal and an ignition control signal for maintaining the current combustion state in the allowable combustion range; (3) Correct the fuel control signal to increase or decrease the current fuel in the allowable combustion range. Function of calculating; (4) Having a function of correcting and calculating the ignition control signal so that the ignition timing is advanced or retarded even if the fuel control signal is corrected and calculated but is not within the allowable combustion range. Digital arithmetic means that is repeatedly executed to control the fuel and ignition timing so as to enter the allowable combustion region; (d). A lean burn engine control device having an output means for supplying a fuel control signal and an ignition control signal calculated by the digital calculation means to a fuel injection valve and an ignition coil; and (a). A step of obtaining a misfire state of the engine; (b). Determining the concentration of NOx generated by the engine; (c). Determining whether or not the current combustion state is in an allowable combustion region defined by the misfire limit and the NOx concentration limit, based on both the misfire state and NOx concentration information; (d). In the allowable combustion region, a step of calculating a fuel control signal and an ignition control signal for maintaining the current combustion state; (e). Correcting the fuel control signal so as to increase or decrease the current fuel if it is not within the allowable combustion range; (f). Even if the fuel control signal is corrected, if it is not within the allowable combustion range, the step of correcting the ignition control signal to advance or retard the ignition timing is repeatedly executed to maintain the combustion state of the engine within the allowable combustion range. The lean burn engine control method is characterized by controlling the fuel and the ignition timing so as to be able to do so.

[Action]

Therefore, it is determined whether the current combustion state is within the allowable combustion range defined by the misfire limit and the NOx concentration limit based on both the misfire state information and the NOx concentration information, and if it is within the allowable combustion range, the current combustion state is maintained. Maintains the engine combustion state in the allowable combustion range by controlling fuel and ignition, increasing or decreasing the current fuel if it is not in the allowable combustion range, and advancing or retarding the ignition timing if it is not in the allowable combustion range. can do.

〔Example〕

Hereinafter, details of embodiments of the present invention will be described. FIG. 1 is a schematic diagram of a control flow of the present invention. Step 1, Step 2
Input the signals of air-fuel ratio A / F, ignition timing Adv, misfire and NOx to a digital arithmetic unit consisting of a microcomputer, and input the above misfire signal and NOx signal to the control target range determined by the misfire limit and NOx limit at step 3. Determine if it is connected. If it is outside the control target range, the correction operation mode of step 4 is entered to control A / F and Adv so that the control target range is entered. If it is within the control target range, the block 5 is shifted to the normal operation mode and the normal control is executed.

FIG. 2 is a conceptual diagram of the present invention. A misfire detector 7 for detecting the misfire state of the engine 6, a NOx detector 8 for outputting a signal corresponding to the concentration of NOx generated by the engine 6 or discharged from the engine, and the signals of both detectors are inputted. A digital arithmetic unit 9 composed of a micro computer,
In addition to this, the digital arithmetic unit 9 has an engine 6
Air-fuel ratio A / F signal and ignition timing signal Adv of the mixture supplied to
Is entered. Based on these information, the digital arithmetic unit 9 selects the normal operation mode 10 or the correction operation mode 11 according to the control flow shown in FIG. 1, executes either mode, and the ignition timing control means 12, respectively. The present invention is configured such that a control signal is output to the fuel amount or air amount control means 13 to control Adv and A / F.

The control method of the present invention will be further described with reference to FIG. Considering the misfire limit, NOx is located to the left of the misfire limit line in Fig. 4.
Considering the limit, it is necessary to be on the right side of the NOx limit line shown in the figure, and in the end, when both are considered, it is necessary to control to the control target area surrounded by these both. Here, the misfire detector detects that it is in a misfire state, and it is shown from the A / F and Adv values at that time. Assuming that it is at the point of P ₁ , the A / F is controlled in the direction of the latch to the point of P ₂ , and the misfire state is eliminated. Then, if the value detected by the NOx detector is equal to or less than the allowable value of the NOx limit, it is determined that the control is performed in the control target range. On the other hand, at a position such as shown P ₃ points, when the detected value of the NOx detector has been filed with the allowable value or more, the P _4-point by controlling the A / F in the lean direction, the NOx within the allowable value Control. In this case, if it is determined that the signal detected by the misfire detector is not in the misfire state, it is determined that the control target range has been controlled accordingly, and the subsequent correction control is interrupted.

Next, at the position like the point P _{5 in} the figure, when the value detected by the NOx detector is above the allowable value, the A / F is controlled in the lean direction in the same manner as described above to make it the P ₆ point, and NOx Is within the allowable value. However, in this case, the misfire detector determines that the misfire condition exists, so that both conditions are not satisfied. In this case, Dotsute to Itsutan P ₅ points in the direction of delaying the ignition timing Adv, A / F
It is necessary to control so as to reach the point of P ₇ by controlling in the lean direction, or control to move from the point of P _{6 to} the point of P ₇ .

Then in a position such as a point shown P _8, although NOx is within the allowable value, if it is determined that exceeds the misfire limit, the P ₉ by Ritsuchi the A / F, to escape from the misfire limit. However, in this case, NOx exceeds the allowable value at P ₉ point, so P ₈
Also the throat connexion ignition timing Adv proceed predetermined value, to reach a point P ₁₀ to Ritsuchi the A / F at the same time, misfire, it is necessary to perform control that satisfies both of NOx. As described above, the control that shifts from P ₉ to P ₁₀ may also be used.

FIG. 5 is an overall configuration diagram of the lean burn control device according to the present invention. The engine misfire state, the combustion sensor having a function of detecting the NOx concentration, the combustion detection end 21, which is attached to the combustion chamber 23 of the engine 22 so as to open, between the combustion detection end 21 and the optical signal processing circuit 25. It is composed of an optical fiber cable 24 for controlling optical transmission and an optical signal processing circuit 25.

The combustion detecting end 21 may be of a type integrated with an ignition plug as shown in FIG. 6 or a stand-alone type as shown in FIG. In FIG. 5, the case where the former detection end is attached is taken as an example. In this case, the combustion detection end
Ignition pulses are given to the 21 from the controller 27 with a built-in microcomputer computer through the ignition device 26, so that it also functions as an ignition plug.

The combustion light signal or the combustion temperature signal, which has been photoelectrically converted and signal-processed by the optical signal processing circuit 25, is guided to the controller 27. In addition to the controller 27, the air-fuel ratio signal detected by the air-fuel ratio sensor 28 (oxygen sensor may be used), throttle valve opening sensor 29 or throttle valve opening information signal from the throttle valve opening switch 29, air flow sensor 30 The detected air flow rate signal, the engine cooling water temperature signal detected by the water temperature sensor 31, the engine speed signal detected by the rotation sensor 32, the crank angle signal, and the like are input. The controller 27 determines the optimum fuel amount, the optimum ignition timing, etc. by the arithmetic processing of these plural signals, and outputs control signals to the injector 33 and the ignition device 26, respectively. The injector 33 injects the optimum amount of fuel according to this control signal and supplies it to the engine. On the other hand, in the ignition device 26, spark discharge is performed to the spark plug 21 at the optimum ignition timing by the control signal.

In the case of FIG. 5, the multipoint fuel injection system is taken as an example, but the present invention is not limited to this, and any of a carburetor, a single-point fuel injection system and the like may be used. In addition, although the air flow rate direct measurement method using the air flow sensor 30 is taken as an example, a speed density method that calculates the air flow rate from the engine speed and the suction negative pressure,
Any air measuring method such as a method of calculating the air flow rate from the engine speed and the throttle valve opening may be used.

FIG. 6 is a cross-sectional view of essential parts of a fuel detection end integrated with a spark plug. A quartz glass fiber 36 having a diameter of about 1.0 to 1.5 mm, which is a light guide, is arranged so as to penetrate the center electrode 34 of the ignition plug and the center axis of the high-voltage terminal 35. The three members and the electrical insulator 37 are fixed to each other at the seal portion 38 by heating and melting and sealing the conductive glass sealing material. The high voltage for the spark discharge is guided to the center electrode 34 through the high voltage terminal 35 and the conductive glass seal portion 38, and causes the spark discharge with the ground electrode 39. The quartz glass fiber tip portion 36a on the combustion chamber side may have various shapes such as a projecting lens shape, a flat shape, and a taper shape, but an optimum one may be appropriately selected from the shape of the combustion chamber and the installation angle of the combustion detection end 21. .

FIG. 7 is a sectional view of a main part of a stand-alone combustion detection end. A quartz glass fiber 36 is provided so as to penetrate a metal housing 41 having a screw portion 40 for mounting on an engine and a central shaft portion of an optical fiber guide terminal 42. As in the case of FIG. 6, the three members are fixed to each other by heating, melting and sealing at the seal portion 38 with a glass sealing material. Although omitted in the explanation of FIG. 4, this sealing material has a melting point of about 600 to 800 ° C., and even if this sealing part becomes high temperature due to heat transfer from the engine, radiation, etc. ,
It is designed to ensure the adhesiveness. Further, the quartz glass fiber tip portion 36a may be constructed in the same manner as in the case of FIG. 6, and therefore its explanation is omitted here.

As a method of detecting a misfire, conventionally, it has been generally obtained by estimating from the magnitude of fluctuation in engine rotation, but in the present invention, the intensity of combustion flame light associated with combustion or the magnitude of combustion temperature associated with combustion is used. The method of detecting misfire is adopted.

As a method of detecting the concentration of NOx, in the present invention, paying attention to the fact that the combustion temperature in the combustion chamber and the NOx concentration have a correspondence relationship,
The method of finding NOx from the combustion temperature is used. Further, as another method, focusing on the fact that the emission spectrum of NOx is 5.3 μm, a method of obtaining the NOx concentration from the combustion light intensity in this wavelength band is also used. As another method, focusing on the fact that the exhaust gas temperature immediately after the combustion chamber almost corresponds to the combustion temperature, a method of inferring the NOx concentration from this exhaust gas temperature is also effective.

The above misfire and NOx detection methods can be classified into (1) a method for detecting misfire and NOx from the intensity of combustion light, and (2) a method for obtaining misfire and NOx by detecting combustion temperature.

First, the method of detecting misfire and NOx from the intensity of combustion light in (1) will be described. In the case of this method, the tip portion 36a of the combustion detecting end 21 is constructed so that its surface is subjected to polishing or the like so as to ensure sufficient translucency, as shown in FIGS.

FIG. 8 is a conceptual configuration diagram of optical signal transmission and signal processing in this case. The combustion light detecting end 21 is attached to each combustion chamber 23 of the engine so as to open. The present invention may be a system in which only one specific combustion chamber is detected. Each combustion light detection end 21a, 21b,
The optical signals from 21c and 21d are optical fiber cables 24a, 24b and 24
It is guided to the optical signal processing circuit 25 by c and 24d. Here, as the optical fiber cable 24, a highly heat resistant plastic fiber having a heat resistant temperature of 140 to 150 ° C. or higher is desirable. In the case of a plastic fiber, the single core integrated fiber 43 in which each fiber is melted as shown in FIG. The combustion stroke of a four-cycle engine is one cycle out of four cycles, and in the case of a four-cylinder (combustion chamber) engine, the combustion stroke of each combustion chamber is sequentially repeated about every 180 °. Even with a 6-cylinder engine, the intervals are 120 °, and the overlap of the combustion stroke between the combustion chambers is only 60 °. Therefore, even with the single-core integrated fiber 43, the optical signal from each combustion chamber hardly overlaps and is led to the optical signal processing circuit 25 intermittently in time, so the optical signal of which cylinder (combustion chamber) Can be easily identified. Needless to say, a method of individually transmitting light for each cylinder without integrating the fibers may be used. FIG. 9 is a block diagram of the optical signal processing circuit 25. The combustion light guided by the single-core integrated fiber 43 is branched into two again, and is guided to the NOx concentration detection 44 and the misfire cycle detection fiber 45. NO
The x-density detection light is guided to the first photoelectric conversion element 47 via the optical filter 46 that allows only the wavelength band of 5.3 μm to pass through. The light intensity corresponding to the NOx concentration is detected and converted into electricity.
On the other hand, the light for detecting the misfire cycle is directly guided to the second photoelectric conversion element 48 and electrically converted. The electric converted signal is processed by the electric signal processing units 49 and 50, and the controller 27
Sent to. Here, in the case of this system, it is necessary to transmit light in the infrared region of 5.3 μm, so that the optical conductor such as the above-mentioned plastic fiber has a large transmission loss and cannot transmit sufficient light to the optical signal processing circuit 25. there is a possibility. Therefore, in that case, a fluorine-based optical fiber having excellent light transmittance in the infrared region may be used. When it is not possible to melt a plurality of fibers like a plastic fiber to make them into one, they are integrated by an optical element in the optical signal processing circuit section or each photoelectric conversion element corresponding to a cylinder (combustion chamber). It is only necessary to take either measure.

The misfire signal and the NOx signal obtained in this way are guided to the controller 27 as shown in FIG. 8 and, together with a plurality of other engine parameter signals 51, are subjected to arithmetic processing, the air-fuel ratio 52, the ignition timing 53. Is controlled.

FIG. 10 shows the relationship between the NOx concentration and the combustion light intensity signal in the 5.3 μm band when the method shown in FIGS. 8 and 9 is used, and since they are in a substantially proportional relationship, the signal output terminal 49a of FIG. The electric signal output from the device has a value almost corresponding to the NOx concentration.

FIG. 11 shows an example of the detection waveform of the misfire state when the method shown in FIGS. 8 and 9 is used. The electric signal of the combustion light intensity and the combustion pressure signal output from the signal output end 50a of FIG. 9 were simultaneously detected, and the misfire was discriminated. As is clear from the figure, when a misfire occurs, the intensity of the burning light becomes zero and no peak-shaped waveform occurs. On the other hand, the combustion pressure is also only the compression pressure, and the pressure rise due to combustion is not seen. Therefore,
A slice level is set as shown by S in FIG. 11, and if a signal below this value is considered to be a misfire, by forming a misfire signal pulse for this misfire cycle and transmitting it to the controller 27, misfire information is sent to the successive controller. .

FIG. 12 is a diagram in which the average value V _P of several cycles of the peak value V _P for each cycle of the combustion light intensity signal shown in FIG. 11 is represented by contour lines with respect to the air-fuel ratio and the ignition timing. _P1 shown
The smaller the number of subscripts, the larger the peen value, and the smallest _Pn . As is clear from Fig. 12,
The contour line of _P has a shape that corresponds well to the curve of the misfire limit (broken line portion), and the misfire determination may be performed using the average value _P for several cycles.

FIG. 13 is a flowchart for determining the NOx allowable limit. The signal (step 51) photoelectrically converted through the photoelectric conversion element 47 and the electric signal processing circuit 49 of FIG. 9 is guided to the controller 27, and is input as a NOx signal at step 52. In step 52, NOx is considered in consideration of the characteristics shown in FIG.
Calculate the concentration. In this case, it is better to obtain the average value of several tens of combustion cycles. In step 54, the NOx allowable value in that operating condition (value satisfying the area on the right side of the NOx limit line in Fig. 4) is compared with the NOx value calculated in step 53, and the NOx value calculated within the allowable value is entered. It is judged whether it is connected. If it is within the allowable value, go to step 55 in normal operation mode,
If the value is out of the allowable value, the correction operation mode 56 is entered and various controls are executed.

FIG. 14 is a flow chart for judging the state of misfire.
The signal (step 57) photoelectrically converted through the photoelectric conversion element 48 and the electric signal processing circuit 50 of FIG. 9 is an electric signal processing circuit.
At 50, the step 50 misfire pulse formation is performed. That is, in step 58, the combustion light intensity signal waveform is input as shown in FIG. 11, and the combustion light intensity signal waveform generated in each combustion cycle is applied to the predetermined slice level S as shown in FIG. Determine whether the peak value is large or small. If it is smaller than S, it is determined that the cycle is a misfire,
A misfire pulse is generated and output for a predetermined period within the cycle.

The misfire pulse thus obtained is input to the controller 27 at step 59. The misfire pulse generated from step 59 is counted in step 60 during the predetermined number of combustion cycles, and the count value is sent to step 61. In step 61, it is judged whether the count value is a predetermined value or more, and if it is the predetermined value or more, the correction operation mode 62
If the value is equal to or less than the predetermined value, the normal operation mode 63 is entered.

The part A of FIG. 13 and the part A ′ of FIG. 14 are executed by the controller 27. FIG. 15 is a flow chart in which the control of the both is integrated and the details of the correction operation mode are described.

The controller 27 receives signals of engine speed N _E and air flow rate Q _A (load information such as intake pipe pressure, throttle valve opening, etc.) at step 64, and at step 65 the air-fuel ratio.
Signals of A / F (this value is X) and ignition timing Adv (this value is Y) are input.

Next, at step 66, the misfire pulse (M) formed by the electric signal processing circuit 50 of FIG. 9 and the NOx (N) signal formed by the electric signal processing circuit 49 of FIG. 9 are input. The value of N is a waveform that is peak-held for each combustion cycle.

At step 67, the concentration of NOx is determined by the NOx input at step 66.
The signal (N) is converted using the characteristics shown in FIG. Furthermore, in this step, for each converted combustion cycle,
The NOx concentration value is averaged over several tens of times and the average value is calculated. In Block 68, this is N _E , Q _A
NO determined in advance corresponding to the operating state of the engine determined by
It is determined whether the x density is larger than the allowable value, and if it is larger, the process proceeds to step 69 and step 70. If it is smaller, the process proceeds to steps 71 and 72.

On the other hand, the misfire pulse M is counted in step 73 how many times there are during a predetermined number of predetermined combustion cycles, and the count number M _X is sent to step 74 to determine whether this number is a predetermined number or less. , M _X is larger Steps 72 and 71
If M _X is smaller, the process proceeds to steps 69 and 71. The steps 69, 70, 70 and 72 are adapted to operate only when signals are transmitted from both of the two input signal lines.

Step 69 operates under conditions where NOx is out of tolerance and misfire is within tolerance. This is the case where points such as P ₃ , P ₅ , and P ₉ shown in FIG. 4 are applicable. In this case, X = X + 1 is set, the air-fuel ratio is increased by a predetermined value to be diluted, and the control signal is set to A /
Send to F correction control step 75. In step 75, the air amount or the fuel amount is controlled by this control signal, and the air-fuel ratio of the air-fuel mixture supplied to the engine is sent as leaner than the last time. Here, in the case of P ₃ in FIG. _4, by repeating this mode, the control target region satisfying both conditions is entered as shown by P ₄ , and the normal operation mode of step 71 is entered. In the case of P ₅ , P ₉ , repeating the mode through block 69 results in P ₆ , P
_Although NOx is within the allowable value as in ₈ , the misfire enters the area where the value exceeds the allowable value. That is, the condition of step 72 is satisfied. In this case, at step 72, the preset ignition timing corresponding to the engine operating condition determined from N _E and Q _{A is} set.
The deviation ΔY between Y ₀ and the ignition timing Y at this time is taken in the form of ΔY = Y−Y ₀ , and it is determined at block 76 whether ΔY is positive, that is, whether the ignition timing is ahead of the set ignition timing. When ΔY is positive, that is, in the state of P _{6 in} FIG. 4, first, the air-fuel ratio is set to the Rich side in the form of X = X-0.5 in step 77, and this signal is sent to step 75 and A / Executes F correction control. Further, in parallel with this, the process proceeds to step 78, the ignition timing is set to side delays in the form of Y = Y-1, comprising the signal running Adv correction control sends to block 79 in such a state of P ₇ To control.

On the other hand, if ΔY is negative, that is, in the state of P ₈ in FIG. 4, the process proceeds to step 80 and the air-fuel ratio is set to X = X-0.5.
Set to the latch side in the form of, and send this signal to step 75 to execute the A / F correction control, and at step 81 set the ignition timing to the advance side in the form of Y = Y + 1 and set this signal to the step. Send to 79 and execute Adv correction control, and control so as to be in a state like P ₁₀ .

Next, when both NOx and misfire are above the allowable values, the routine proceeds to step 70, where the deviation of the ignition timing is obtained in the form of ΔY = Y−Y ₀ , and at step 82 it is judged whether ΔY is positive. If positive, i.e., if the conditions such as P ₁₁ of FIG. 4, slightly set Ritsuchi side air-fuel ratio in the form of X = X-0.1 at step 83, sends the signal to step 75 A / F In addition to the correction control, the ignition timing is set slightly retarded in the form of Y = Y-1.5 in step 84, and in that case sent to step 79.
Adv correction control is executed and control is performed to enter the control target range. On the other hand, when ΔY is negative, that is, in the case of the state like P ₁₂ in FIG. 4, the routine proceeds to step 85, where the air-fuel ratio is X = X
Set slightly to the lean side in the +0.2 form, send the signal to step 75 to execute A / F correction control, and proceed to step 86 to make a large advance in the ignition timing Y = Y + 1.5 form. It is set to the No side, the signal is sent to step 79, Adv correction control is executed, and control is performed so as to enter the control target range.

In this way, if it is possible to enter the control target area, the map table created from N _E , Q _A will contain the X, Y at that time.
Is stored, and the same control is executed next time, and a new X,
When Y is obtained, it is effective to add a function to update and learn it. Further, as shown in FIG. 4, it is also effective to find and store the misfire limit line and the NOx limit line on the X and Y map table by repeating the above control.

Next, the method (2) described above, that is, the method of detecting and controlling misfire and NOx by detecting the combustion temperature will be described.

FIG. 16 is an example of details of the quartz glass fiber tip portion 36a of the combustion detecting end 21 for detecting the combustion temperature. 16 (a) is an axial cross section thereof, and FIG. 16 (b) is a cut cross section of XX. The tip of the quartz glass fiber 36 is arbitrarily selected so as to have a large heat radiation surface. The surface is covered with a black body film 87, and the outer periphery thereof is covered with a member having excellent heat resistance, for example, quartz glass 88, ceramics 88, etc., and fusion bonded to the quartz glass fiber 36. The black body film is formed on the tip of the quartz glass fiber 36 by a vapor deposition method or a sputtering method, and the reinforcing body 88 is also formed on the surface of the black body film 87 and the quartz glass fiber 36 by a vapor deposition method, a sputtering method or the like. Just go. Note that a sapphire rod may be used instead of the quartz glass fiber 36. Black body membrane 87 and reinforcement membrane 88 are 2
A thin film of about 5 μm is desirable from the viewpoint of heat capacity. However, if sufficient bondability cannot be obtained, a film thicker than the above is used.

Black bodies are known to generate radiant energy with respect to temperature, and Fig. 17 shows the relationship between spectral radiant energy of black bodies, wavelength, and temperature. The temperature can be estimated by detecting the radiant energy for a certain wavelength. For example, Fig. 17P
1000-50 by detecting the radiant energy of the wavelength of the point
A temperature of 00 ° C can be determined.

When the temperature of the black body film 87 shown in FIG. 16 rises due to the combustion gas, that is, when the combustion temperature changes, the radiant energy (radiant light) corresponding to it is emitted from the black body film 87, so that This is an optical filter (point P in Fig. 17)
An electric signal corresponding to the combustion temperature can be obtained by receiving the light with a photoelectric conversion element through the one having a transmittance of 500 nm).

FIG. 18 shows the characteristics of the photocurrent with respect to the combustion temperature of the photoelectric conversion element. The photocurrent changes with the log scale with temperature change. Therefore, when detecting a wide temperature range, methods such as logarithmic compression and linearization of this photocurrent via a log diode are added.

FIG. 19 is a relationship diagram between combustion temperature and NOx. The higher the temperature and the larger the excess air ratio λ, that is, the larger the air-fuel ratio A / F, and the leaner the air-fuel ratio, the larger the NOx concentration. However, the pressure in the combustion chamber is 1 atm and 50 atm
It shows that the effect of pressure is negligible under normal combustion conditions. Therefore, if the air-fuel ratio at that time can be detected, NOx can be easily calculated from the combustion temperature at that time. As shown in FIG. 5, the air-fuel ratio can be detected as an average value by the air-fuel ratio sensor 28, and the NOx concentration can be calculated if the combustion temperature can be detected.

FIG. 20 shows the configuration of the signal processing circuit 25 in the case of the combustion temperature detection method. The thermal radiation light from the black body film 87 of the tip end portion 36a guided by the single-core integrated fiber 43 has a specific wavelength (for example, 7
(00 nm) and is guided to the photoelectric conversion element 90 via the optical filter 89 that allows only the light to pass therethrough. Since this radiant light intensity is a function of temperature, it means that the information on the combustion temperature is obtained by this photoelectric conversion. In the case of the misfire cycle, the combustion temperature naturally does not rise. Therefore, in the case of a signal below a predetermined voltage level, it is possible to output a misfire signal by passing through the comparator and the waveform shaper as a misfire cycle. Further, since the absolute value of the combustion temperature is almost proportional to the NOx concentration, it is possible to output a signal corresponding to the NOx concentration from the peak value or integrated value of the photoelectrically converted electric signal.
These are performed by the electric signal processing circuit 91.

That is, in the electric signal processing circuit 91, a combustion temperature signal waveform corresponding to each combustion cycle is formed by utilizing the characteristics shown in FIGS. 17 and 18, and is almost equivalent to the combustion light intensity signal waveform shown in FIG. Create a signal waveform. By utilizing the fact that the combustion temperature does not rise at the time of misfire, an arbitrary slice level can be set, and if the combustion temperature signal for each combustion cycle does not exceed this level, a misfire pulse will be sent once within that combustion cycle. It is generated and the signal is output from the output terminal 92. The output terminal 93 is configured to output a signal in which the peak value of the combustion temperature signal is peak-held for each combustion cycle.

FIG. 21 is a flow chart for determining the NOx allowable limit. A signal (step 94) which is photoelectrically converted by the photoelectric element 90 in FIG. 20 and peak-held by the electric signal processing circuit 91 for each combustion cycle is transmitted to the controller 27 and is input as a combustion temperature signal by the block 95. Again step
At 96, the air-fuel ratio signal from the air-fuel ratio sensor 28 shown in FIG. 5 is input. In step 99, the NOx concentration is calculated using the characteristic curve of NOx with respect to the air-fuel ratio (excess air ratio) at the combustion temperature as shown in FIG. Furthermore, the average value for 10 combustion cycles is calculated, and the average NOx value is calculated as step 100.
Send this to. In Step 100, N in the operating condition
The Ox allowable value is compared with this average NOx value to determine whether or not there is an allowable value. If it is within the allowable value, normal operation mode 101
If not, shift to the correction operation mode 102,
The same control as in FIG. 15 is executed.

It is a flow chart for judging the misfire state of FIG.
Photoelectric conversion is performed by the photoelectric conversion element 90 shown in FIG. 20 (step 94).
The misfire pulse (step) formed by the electric signal processing circuit 91.
103) is transmitted to the controller 27 and input as a misfire pulse at step 104. In step 105, the number of misfire pulses is counted for a predetermined number of combustion cycles, and it is determined whether this numerical value is a preset numerical value or more. If it is less than the predetermined value, the operation proceeds to step 101 because it is not misfiring, and if it is more than the predetermined value, it goes to step 102 because it is misfiring.
The same control as in the figure is executed.

The controller 27 executes B in FIG. 21 and B ′ in FIG.

The details of the control flow of the B and B'sections are the same as those in FIG. 15, so description thereof will be omitted here.

〔The invention's effect〕

According to the present invention described above, it becomes possible to easily control to an extremely narrow control target range determined from the NOx limit and the misfire limit, which were the problems of the conventional lean burn control device, and the misfire limit with the aging of the engine, NOx. Even if there is a limit change, a NOx limit change due to a change in vehicle weight, or a misfire limit change due to a change in fuel properties, optimal control can always be performed.

This not only improves fuel economy and exhaust (NOx) performance, but also improves drivability. Further, even if the above-mentioned various conditions, state changes, atmospheric condition changes, etc. occur, a suitable engine operating state can always be obtained.

[Brief description of drawings]

FIG. 1 is a flow chart showing an embodiment of the present invention, FIG. 2 is a configuration diagram corresponding to the flow chart of FIG. 1, FIG. 3 is a characteristic diagram of air-fuel ratio, torque, NOx, and FIG. Is a characteristic diagram of air-fuel ratio and ignition timing, FIG. 5 is a system diagram, FIG. 6,
FIG. 7 is a sectional view of the combustion light sensor, FIG. 8 is a detailed configuration diagram of the present invention, FIG. 9 is a partial configuration diagram of FIG. 8, FIG. 10, and FIG.
Fig. 12 is a characteristic diagram, Fig. 13, Fig. 14 and Fig. 15 are detailed flow charts of the present invention, Fig. 16 is a sectional view of a black body sensor, Fig. 17, Fig. 18, Fig. 19 The figure is a characteristic diagram, FIG. 20 is a processing circuit diagram, and FIGS. 21 and 22 are detailed flow charts. 6 ... Engine, 7 ... Misfire detector, 8 ... NOx detector, 9 ... Digital arithmetic unit.

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Toshiharu Nogi 4026 Kuji Town, Hitachi City, Hitachi, Ibaraki Hitachi Research Laboratory, Hitachi, Ltd. (56) References JP-A-60-17239 (JP, A) JP-A-61 -229950 (JP, A) JP 60-135634 (JP, A) JP 62-30932 (JP, A) JP 61-87945 (JP, A)

Claims (2)

[Claims] 1. (a). Misfire state determination means for determining the misfire state of the engine; (b). NOx concentration determination means for determining the concentration of NOx generated in the engine; (c). (1) A function of determining whether or not the current combustion state is in an allowable combustion region defined by the misfire limit and the NOx concentration limit, based on both information from the misfire state determination means and the NOx concentration determination means; Function to calculate the fuel control signal and the ignition control signal for maintaining the current combustion state in the allowable combustion range; (3) Correct the fuel control signal so as to increase or decrease the current fuel if not in the allowable combustion range Function of calculating; (4) Having a function of correcting and calculating the ignition control signal so that the ignition timing is advanced or retarded even if the fuel control signal is corrected and calculated but is not within the allowable combustion range. Digital arithmetic means that is repeatedly executed to control the fuel and ignition timing so as to enter the allowable combustion region; (d). A lean burn engine control device comprising: an output means for supplying a fuel control signal and an ignition control signal calculated by the digital calculation means to a fuel injection valve and an ignition coil. 2. (a). A step of obtaining a misfire state of the engine; (b). Determining the concentration of NOx generated by the engine; (c). Determining whether or not the current combustion state is in an allowable combustion region defined by the misfire limit and the NOx concentration limit, based on both the misfire state and NOx concentration information; (d). In the allowable combustion region, a step of calculating a fuel control signal and an ignition control signal for maintaining the current combustion state; (e). Correcting the fuel control signal so as to increase or decrease the current fuel if it is not within the allowable combustion range; (f). Even if the fuel control signal is corrected, if it is not within the allowable combustion range, the step of correcting the ignition control signal to advance or retard the ignition timing is repeatedly executed to maintain the combustion state of the engine within the allowable combustion range. A lean burn engine control method characterized by controlling fuel and ignition timing so as to be able to perform.