JPS62267574A - Knock controller for internal combustion engine - Google Patents

Abstract

PURPOSE:To enable correction of a knock decision level to the optimal level at all times by providing means for calculating a knock decision level which is compared with a knock strength level in a specific interval of a knock sensor signal and means for correcting said level. CONSTITUTION:A referential level VL satisfying a relation V<=VL with an approximately same probability with a knock decision probability Pk for a central value V50 in the distribution of knock strength levels V being outputted from a peak hold circuit 7 to an ignition timing control circuit 8 is obtained. A knock decision level Vref is determined such that the ratio between Vref and V50 is slightly larger than the ratio between V50 and VL. Consequently, the knock decision level variable according to manufacturing error, aging, etc. of an engine 1, a knock sensor 6, etc. can be corrected to the optimal level, while even under an operating condition where occurrence of knock is low, the knock decision level does not increase excessively and even upon transfer to an operating condition where occurrence of knock is high, occurrence of knock can be detected reliably.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention provides a knock control system that controls knock control factors such as ignition timing, boost pressure, air-fuel ratio, and EGR in accordance with the state of knock occurring in an internal combustion engine. This invention relates to a control device (hereinafter referred to as a knock control system).

[Conventional technology]

-A knock control system is an electrical signal from a knock sensor that detects engine vibration (hereinafter referred to as
When the knock sensor output signal (hereinafter referred to as the knock sensor output signal) exceeds a certain predetermined level (hereinafter referred to as the knock determination level), it is determined that knock has occurred, and the ignition timing is retarded. If no knock is detected, the ignition timing is advanced to keep the ignition timing close to the knock limit, thereby maximizing the fuel efficiency and output characteristics of the engine. (For example, JP-A-56-115861).

[Problems to be Solved by the Invention] In such a knock control system, the knock determination level has extremely important meaning.

If the knock determination level is too high, the ignition timing is advanced because knocking is not detected even though it is occurring, resulting in frequent knocking and even damage to the engine. On the other hand, if the knock detection level is too low,
Even though there is no knock, the ignition timing is retarded and the engine is unable to produce sufficient output.

Conventionally, in order to create an appropriate knock judgment level, for example, the output after passing the knock sensor signal through an integrating circuit,
Constant K carefully adapted to each engine speed
(hereinafter referred to as value) and further adds an offset voltage.

However, because there are manufacturing errors even in the same engine model, despite careful adaptation of the K value,
The knock determination level may be set to an inappropriate level, making accurate knock detection impossible.

As described above, the method of creating a knock determination level in a conventional knock control system requires careful value adaptation and has the problem that accurate knock detection cannot be achieved due to engine manufacturing errors.

In view of the above-mentioned problems, the present invention aims to provide a knock control device that detects knock with high precision and always realizes an optimal knock state, without being affected by variations in knock sensors or engines, changes over time, etc. .

In order to achieve such an object, the inventors of the present invention have developed a system in which the distribution shape of the logarithmic value (LOG') converted value of the knock intensity value has a predetermined shape, as described in Japanese Patent Application Laid-Open No. 60-243369. We have already invented a method for correcting the judgment level so that The present invention is a further development of the technology disclosed in Japanese Unexamined Patent Publication No. 60-243369.

[Means for solving problems]

Therefore, as shown in FIG. 1, the present invention includes a knock sensor A that detects knocking occurring in an internal combustion engine H, a knock determining means E that determines non-king based on the output signal of this knock sensor, and a result of this determination. a control value calculation means F for calculating a control value for controlling knock control factors such as ignition brightness or boost pressure, and a drive for an idler or the like that changes the value of the knock control factor according to this control value. a non-king control device for an internal combustion engine, comprising: a knock intensity value detection means B for detecting a knock intensity value V in a predetermined section of the knock sensor signal; and a knock determination level to be compared with the knock intensity value (■). The knock judgment level generating means C is configured to include a judgment level creating means C for calculating the knock judgment level, and a judgment level correcting means C for correcting the knock judgment level.

[Summary of the invention]

As described in Japanese Patent Application Laid-Open No. 60-243369, the maximum value ■ of a predetermined section of the knock sensor signal. , X is sampled multiple times and plotted on lognormal probability paper, we get
Figure 2 (al, fb) shows the state in which there is no knocking and the state in which knocking occurs at a certain frequency. Full throttle,
(This shows the characteristics of the fifth cylinder).

In addition, we believe that this property is not limited to the maximum value V, AX, but also the amount corresponding to the strength of the knock vibration component (hereinafter referred to as
It was confirmed that this holds true for the knock intensity values (referred to as knock intensity values V). Examples of knock strength values having such properties are listed below.

The first is a predetermined interval (for example, 10'' to 90' ATDC
) is the rectified and integrated value of the vibration output.

By changing the time constant of the integrator, this value represents the average vibration output of that cycle or the integrated value of the vibration output of that cycle, but in either case, it is a value that corresponds to the intensity of vibration of that cycle. becomes.

The second is the number of pulse trains generated when the vibration output in a predetermined section is compared with a certain threshold value (for example, a level slightly above the center of vibration), or the integrated value of the pulse trains. This value also corresponds to the vibration intensity of that cycle. This intensity value has a limited variation range compared to the range that the maximum wave height value VHAX can take, so it is slightly inferior in accuracy, but it has the advantage of being easy to handle.

The third example is a somewhat laboratory-level one (the implementation cost is a little high), but it is an effective value of the vibration output (R
MS). This RMS is the vibration level of the timepiece
(tl, the time length of the predetermined section is T〈TYi1 block〒)
This value also corresponds to the strength of the vibration of the cycle.

From the above three examples, the intensity value that generally corresponds to the strength of vibration■
It is natural and reasonable to think that the distribution in the non-knock state will be approximately logarithmic. for example,
A certain intensity value V is the maximum wave height value V that is a member of the same intensity value
Approximately V=a + (V)tall for WAX
) ' (a, n are arbitrary real numbers) If it can be approximated by the relationship, IogV=Ioga+n 1 o g
becomes VMAX, ' Og logV like VMAX
It is clear that also has a normal distribution (that is, ■ is a lognormal distribution). For reference, an example of RMS experimental data is shown in FIG.

The present invention utilizes this property to appropriately correct the determination level.

First, the basic concept of the present invention will be explained using FIG. 3. For the median value V50 of the distribution of knock intensity values ■, a reference level VL is determined that satisfies ■≦VL with approximately the same probability as the knock determination probability PK (details will be described later). And V5
7. So that the ratio of the judgment level V ref and V50 is slightly larger than the ratio A to 0. .

Create. For example, if D is a number greater than 1, then V
r * r = A X D X V S.

shall be.

If the current state is no knock (Fig. 3 (a)), the probability that ■≧V ref becomes ■≦■, that is, the knock judgment probability PK (the example in Fig. 3 So 5
%), the ignition timing advances (if
If V50,=AXV50 is created, the probability of V rot≧■ will be the same as PK, and it will be stable at the current ignition timing).

When the ignition timing advances from a no-knock state to a state where knock is occurring moderately (Fig. 3 (b)), Vrm
The probability that r≦V becomes approximately equal to PK, and the ignition timing is stabilized in this state. If the knock state becomes larger and the degree of bending in the distribution becomes larger than in Fig. 3(b), then V
The probability of rat≦V exceeds PK, and the ignition timing is retarded.

That is, even if the knocking state is too small or too large, the knocking state can be converged to an appropriate knocking state.

What is important here is to set A so that the probability of ■≦V50/A (=VL') is almost the same as PK.

A method of setting the knock determination probability PK and (2) will be explained. The knock determination probability referred to here is determined by the knock control method, and refers to how many knock determinations are performed in an average number of seconds or for each ignition. Specifically, the ignition timing is advanced or retarded to control knock, for example, by 1'C every second.
Assume that the retard amount is reduced by A and increased by 1° CA if there is knocking. In this way, the amount of retardation becomes stable in that knock determination is performed once per second on average, except when the most advanced angle is used and the latest angle is used.

This 1 depression/1 second is called the knock judgment probability.

■The setting method for , is the same as above, for example, decrease the value of by O,I every second, and set A to 0 for each time V≦VL.
．． If you increase it by 1, ■≦■.

This probability can be made the same as PK. In this way, if you increase or decrease A in the same way that you generally increase or decrease the amount of retardation, then
The probability of ≦■ can be made almost the same as the knock determination probability P3.

What has been described above is just an example, but the ratio of V tar and V50 should be slightly larger than the ratio of V50 and ■9. If you can set this, you can perform appropriate knock control. In the above example, only (D-1) is ■7°, and V
The ratio of V50 is larger than the ratio of V50 and ■.

〔Example〕

Hereinafter, the present invention will be explained with reference to embodiments shown in the drawings.

FIG. 5 is a block diagram showing one embodiment of the present invention.

In Fig. 5, 1 is a 4-cylinder 4-cycle engine, 2 is an air cleaner, 3 is an air flow meter that detects the intake air of the engine and outputs a signal in accordance with this, 4 is a throttle valve, and 5 is an engine 5 i1@ Crank angle position (
The distributor is equipped with a reference angle sensor 5A for detecting (for example, top dead center) and a crank angle sensor 5B that generates an output signal at every fixed crank angle of the engine. 6 is a knock sensor for detecting engine block vibration corresponding to the engine nosoter phenomenon using a piezoelectric element type (piezo element type), electromagnetic type (magnet, coil), etc.; 7 is a knock sensor that detects the peak output of the knock sensor for each cylinder. This is a peak hold circuit section that holds the peak. 9 is a water temperature sensor that generates a signal according to the engine cooling water temperature, 12 is a fully closed switch (idle switch) that outputs a signal when throttle valve 4 is fully closed, and 13 is a valve that is almost fully open. Fully open switch (power switch) for outputting a signal when the condition is 14, the air fuel ratio (A/F) of exhaust gas is rich compared to the stoichiometric air fuel ratio.
It is a 02 set/su which generates an output signal depending on whether it is lean or thin.

8 is an ignition timing control circuit for controlling the ignition timing and air-fuel ratio of the engine according to input/output signal states from each sensor and each switch; These are an igniter and an ignition coil that cut off electricity to the ignition coil. The high voltage generated by the ignition coil is applied to the spark plug of a predetermined cylinder at an appropriate time through the power distribution section of the distributor 5. Reference numeral 11 denotes an injector for injecting fuel into the intake manifold based on the fuel injection time (τ) determined by the control circuit 8.

Next, the detailed configuration of the peak hold circuit section 7 will be explained using FIG. 6. In FIG. 6, 701 is a filter such as a band pass or bypass for selecting and extracting only the knock frequency component from the output signal of the knock sensor 6, 702 is an amplifier, and 703
is a peak hold circuit that peak-holds the knock sensor signal output from the amplifier 702 based on the cylinder switching signal from the control circuit 8 using, for example, a capacitor, and outputs a knock intensity value.

Next, the detailed configuration and operation of the control circuit 8 will be explained with reference to FIG. In FIG. 7, 8000 is a central processing unit (CPU) for calculating ignition timing and fuel injection amount, which uses an 8-bit microprocessor. 80
01 is a read-only memory unit for storing control programs and limit constants necessary for calculations) (ROM)
, 8002 is a temporary storage unit for temporarily storing calculation data while the CPU 8000 is operating according to a program.
RAM). 8003 is a waveform shaping circuit for shaping the output signal of the reference angle sensor 5A, and 8004 is a waveform shaping circuit for shaping the output signal of the crank angle sensor 5B.

8005 is an interrupt control unit for causing the CPU to perform interrupt processing based on external or internal signals, and 8006 is a CPU
This is a 16-bit timer configured so that the count value increases by one at each clock cycle, which is the basic cycle of U operation. The engine rotation speed and crank angle position are detected by the timer 8006 and the interrupt control unit 8005 in the following manner. That is, the CPU reads the count value of the timer every time an interrupt occurs due to the output signal of the reference angle sensor 5A. The count value of the timer is the clock period (
For example, by calculating the difference between the count value at the current interrupt and the count value at the previous interrupt, the time interval of the reference angle sensor signal, that is, one revolution of the engine, is calculated. The time required can be measured. In this way, the engine speed is determined. In addition, the crank angle position is
Since the signal of the crank angle sensor 5B is output at every fixed crank angle (for example, 30'CA), the reference angle sensor 5A
You can know the crank angle at that time by setting the top dead center signal at Mtp to 30 degrees.

This crank angle signal every 30°A is used as a reference point for generating an ignition timing control signal and as a cylinder switching signal for a peak hold circuit.

8007 is a multiplexer that switches multiple analog signals at appropriate times and guides them to an analog-to-digital converter (A/D converter) 8008.
It is controlled by a control signal output from 10. In this embodiment, the output signal of the knock sensor signal from the peak hold circuit section 7, the intake air amount signal from the air flow meter 3, and the water temperature signal from the water temperature sensor 9 are manually input as analog signals. 8008 is an A/D converter for converting an analog signal into a digital signal. 800
9 is an input port for digital signals, and in this embodiment, the idle signal from the idle switch 12, the power signal from the power switch 13, and 0□
Rich and lean signals from the sensor 14 are manually input. 8
010 is an output port for outputting a digital signal.

This output boat outputs an ignition timing control signal for the igniter 10, a fuel injection signal for the injector 11, a cylinder switching signal for the peak hold circuit 7, and a control signal for the multiplexer 8007. 8011
is a CPU bus, and the CPU 8000 carries control number 13 and data signals on this bus signal line to control peripheral circuits and send and receive data.

The configuration of the apparatus for implementing the present invention has been described above, and the calculation of ignition timing, knock determination, and correction of the knock determination level will now be described using a flowchart as an example.

First, the overall flow of the present invention will be explained using the flowchart of FIG. 8 as an example, and then the steps that characterize the present invention will be explained in detail.

When the knock control routine starts at step 100, at step 101, the knock intensity value V in a predetermined section of the knock sensor signal is fetched. In step 102, it is determined whether the conditions for knock control are met based on intake air amount, engine speed, etc., and if YES, step 103
If the answer is NO, the process advances to step 107. Step 10
In 3, knock judgment level V50 and reference level ■,
Create. Based on this knock determination level V rat, a knock determination and a retard amount are calculated in step 104. In step 105, it is determined whether the engine is in a steady state based on changes in intake air amount, engine speed, etc. If YES, the process proceeds to step 106, where the knock determination level [E is performed; if NO, step Proceed to 107. In step 107, the percentage points of the distribution of ■ are updated, and in step 108, the knock control routine ends.

Next, steps that are a feature of the present invention will be explained in detail using FIG. 9.

First, in step 103-1, the knock determination level Vref is determined using the median value VSG of the distribution of the knock intensity ■, a predetermined variable A, and a predetermined number larger than 1: V rer =A X D X V S.

and create. Next, the process proceeds to step 103-2, where, using the median value V50 of the knock intensity distribution and the predetermined variable A,
A reference level ■ is created as v,=v50/A.

Next, the process proceeds to step 104, and in step 104-1 it is determined whether or not there is a knock by comparing the magnitudes of V and V,50. Proceed to. In step 104-2,
Increase the retard IR by a predetermined amount ΔR. In step 104-3, a timer or counter determines whether or not time has elapsed over a predetermined period T. If YES, the process proceeds to step 104-4. (At this time 2. The timer or counter is reset.) In step 104-4,
The retard angle i1R is reduced by a predetermined amount ΔR. Next step 1
Proceed to 04-5, and the maximum retardation amount R is determined by engine conditions such as engine speed and intake air amount, 1iRMAX.
Determine whether the limit has been exceeded, and if YES, proceed to step 10.
Proceed to 4-6 and substitute RMAX for R. Further, if it is determined NO in step 104-5, step 10
Proceeding to 4-7, it is determined whether R is smaller than the minimum retard IRMIN determined by the engine conditions. If this judgment is YES, the process advances to step 104-8, and RMI is applied to R.
Substitute N.

Next, the process proceeds to step 105, where it is determined whether the engine is in a steady state, and if YES, the process proceeds to step 106-1.
If the answer is No, the process advances to step 107-1. In step 106-1, ■≦■.

If YES, go to step 106-2.
If No, the process advances to step 106-3.

In step 106-2, A is increased by a predetermined amount ΔA, and the process proceeds to step 106-3. In step 106-3, it is determined whether the time has passed the predetermined period T, and if YES, the process advances to step 106-4 (N).

In this case, the process advances to step 107-1. Step 106-
4, A is reduced by a predetermined amount ΔA.

Next, the process proceeds to step 107-1, where it is determined whether V>V50. If YES, the process proceeds to step 107-2; if No, the process proceeds to step 107-3. In step 107-2, V50 is increased by a predetermined amount Δ■.

Further, in step 107-3, V<VS. If YES, the process proceeds to step 107-4; if NO, the process proceeds to step 108.

In step 107-4, V50 is reduced by a predetermined amount Δ■. The process then proceeds to step 108, and the knock control routine ends.

[Other Examples]

■Knock judgment level in step 103-1 ■□.

is as shown in Figure 10. ■2゜t　　(A+B)XV5゜ However, B is a positive predetermined number You can also create it as

In this case, the other steps may be the same as those in FIG. 9.

(2) Calculation of the percentage point of the distribution of V may be limited to the 50% point (median value V50). For example, as shown in FIG. 11, the 66% point may be used. That is, if in step 103 the knock judgment level ■, , is created as ■, , f=AXV6&, and in step 107 the ratio of increase/decrease of VS6 is set to 2, the probability that V>V, iV>V6° will be satisfied. The ratio is 1/2, that is, the value of the cumulative 66% point. To explain the reason why it may be done in this way, since V,, > V50, AXV&&/V50>AXVSO/V66■1゜, /V
50>V50/V L , and the ratio of V rat and V50 is ■5° and VL
This is because it satisfies the condition that the ratio is slightly larger than the ratio of V A &/A (in this case, V A &/A ).

■The increase/decrease in the amount of retardation is as shown in the example in Figure 9.
The method is not limited to a method in which the amount of retardation is reduced every predetermined period regardless of whether or not there is knocking. For example, a method may be used in which the amount of retardation is reduced when no knocking continues for a predetermined period. In that case, step 1
04 and 106 will change, so we will explain it again using FIG. 12.

First, in step 103, knock determination levels (2), . . . , f are created, and the process proceeds to step 104-1. Step 104
-1, it is determined whether there is knocking or not, and if there is knocking, proceed to step 104-2; if there is no knocking, proceed to step 104-2.
-Proceed to 3. In step 104-2, the retard angle IR is increased by ΔR. Further, in step 104-3, it is determined whether or not knocking has continued for a predetermined period. If YES, the process proceeds to step 104-4; if NO, the process proceeds to step 104-5.
Proceed to. In step 104-4, the retard IR is decreased by ΔR. Steps 104-5 to 104-8 are for guarding the maximum and minimum amount of retardation, and since they have already been explained with reference to FIG. 7, their explanation will be omitted.

Thereafter, if it is determined in step 105 that the state is steady, the process proceeds to step 106-1, where it is determined that ■≦■. Therefore, in the case of YES, the process proceeds to step 106-2, and A is increased by ΔA. Moreover, in step 106-3, ■≦■
, for a predetermined period of time, and if YES, the process proceeds to step 106-4, where 8 is decreased by ΔA. By doing so, the probability of ■≦■ can be made approximately equal to Pg. In this way, if the amount of retardation is increased or decreased in the same way, the present invention can be implemented regardless of the method of increasing or decreasing the amount of retardation.

■The value of the constant B is shown in Figure 13 (A), for example.

As shown in (B), it may be changed depending on the engine conditions (intake air amount, engine speed).

(2) The constants R and B may be changed for each cylinder as shown in Tables 1a and lb.

Table 1a (change the value of D for each cylinder) Table Lb (change the value of B for each cylinder) It may be changed depending on both.

Table 2 Change CD according to the cylinder] Table 3 [Change V3 according to the cylinder] ■Also, create a reference level VLZ (-Vso/A) where ■≦VL2 with a probability smaller than PK, and set Vraf =
The present invention can also be substantially implemented as an AXVso. That is, for the level V where ■≦VL with the same probability as PK, since ■Lz < V L, the relationship (2), llf/V50>V50/Vt substantially holds true.

In this embodiment, step 1.03-1 in FIG.

106-3 respectively in steps 103-1 and 103-1 in FIG.
It may be replaced with 106-3. That is, step 103
-1, set the judgment level V rat to V rQ1=Δ×V5
0, and the period for reducing A in step 106-3 is set to be larger than the period for eliminating the retard amount, for example, 2.
Just double it.

In the embodiment of 0 or more, only the retard amount was increased or decreased based on the knock determination, but there is no harm in controlling the A/F, boost pressure, etc. at the same time.

In the examples of 0 or more, the maximum value is used as the knock intensity value, but as mentioned above, it is of course possible to use the rectified/integrated value, the number of knock pulses, the integrated value of the knock pulse train, the effective value, etc. .

〔Effect of the invention〕

As described above, the present invention substantially operates at knock determination level V0. The ratio of the median value V50 of the distribution of the knock intensity value ■ is
V≦■, with almost the same probability as v5゜ and knock judgment probability PK.
By setting the ratio to be slightly larger than the level VL, the knock judgment level, which changes due to manufacturing errors in the engine and knock sensor, changes over time, etc., can be constantly corrected to the optimum value, resulting in optimal knock control. is possible.

Furthermore, JP-A No. 60-243369, which is the basis of the present invention,
In the publication, the knock detection level is corrected according to the state of knock occurrence, so the knock detection level is corrected if the driving condition where knock occurrence is low continues or when high octane fuel is used where knock occurrence is low. If the level gradually rises and you enter a driving state where knocking occurs frequently in this state, or if you switch to a low octane fuel, it will be impossible to detect knocking right after that. In the invention, the knock detection level is created based on the reference level at which the knock intensity value V or higher occurs with almost the same probability as the knock detection probability PK, so even in driving conditions where knock occurrence is low, the knock detection level is The level does not rise more than necessary, and there is an excellent effect that the occurrence of knocking can be reliably detected even when the operating state changes from an operating state where knocking occurs less to an operating state where knocking occurs more frequently.

[Brief explanation of drawings]

FIG. 1 is a diagram corresponding to claims of the present invention, and FIG. 2(a). (bl and Figure 3 (al), (b) are cumulative % characteristic diagrams showing an example of experimental data of LOG (V), Figure 4 is a graph of LOG (V)
Fig. 5 is a partial cross-sectional configuration diagram showing an example of the device of the present invention, and Fig. 6 is a more detailed diagram of the peak hold circuit section of the device shown in Fig. 5. 7 is a more detailed block diagram of the ignition timing control circuit in the device shown in FIG. 5, and FIGS. 8 and 9 are flowcharts 1 and 10 for explaining the operation of the circuit shown in FIG. - Figures 12 and 16 are flowcharts of essential parts in other embodiments of the apparatus of the present invention, respectively, and Figures 13 (A), (B), 14 and 15 are flow charts of other embodiments of the apparatus of the present invention. D value corresponding to engine speed-intake air amount in the example. It is a characteristic diagram of each D, value, and B1 value. DESCRIPTION OF SYMBOLS 1... Engine, 6... Non-ignition sensor, 7... Peak hold circuit part, 8... Ignition timing control circuit, 10
...The igniter and ignition coil that constitute the driving means. Representative Patent Attorney Takashi Okabe ■ Figure 2 Figure 3 Figure 8 Figure 10 Figure 11 Figure 12 (A) (B) Figure 14

Claims (9)

[Claims] (1) A knock sensor for detecting knock of an internal combustion engine, a knock intensity value detection means for detecting a knock intensity value V in a predetermined section from the signal of this knock sensor, and a determination level creation means for creating a knock determination level V_r_e_f and,
The knock intensity value V and the knock determination level V_r_e_f
In an internal combustion engine knock control device, the internal combustion engine knock control device includes a knock determination means for determining the presence or absence of knock by comparison with the knock intensity value, and a drive means for controlling knock control factors such as ignition timing or air-fuel ratio according to the determination result. The knock determination probability P is determined according to the knock intensity value V detected by the detection means.
A reference level creation means that creates a reference level V_L that is equal to or higher than the knock intensity value V with almost the same probability as __K, and a ratio between the knock determination level V_r_e_f and the median value V_5_0 of the distribution of the knock intensity values V is set at this center. Value V_5_0
1. A knock control device for an internal combustion engine, comprising: a determination level correcting means for correcting a knock determination level so that the knock determination level is slightly larger than the ratio between the knock determination level and the reference level V_L. (2) The reference level creation means includes the knock intensity value V
% point value calculation means for calculating at least one % point value of the distribution;
A level calculation means that creates V_L=V_P/A using _P and a certain variable A, and the knock judgment probability P_K.
and a variable correction means for correcting the variable A so that the knock intensity value V becomes equal to or higher than the reference level V_L with almost the same probability as the knock control for an internal combustion engine according to claim 1. Device. (3) The determination level correction means includes a median calculation means for calculating the median value V_5_0 of the distribution of the knock intensity values V, and the knock determination level V_r_e_f and the median value V_ of the distribution.
The knock determination level V_r is set so that the ratio with V_5_0 is slightly larger than the ratio between the median value V_5_0 and the reference level.
A knock control device for an internal combustion engine according to claim 1 or 2, characterized in that the knock control device for an internal combustion engine corrects _e_f. (4) The determination level correction means adjusts the knock determination level V_r_e_f by the variable A, the median value V_5_0 of the distribution, and a predetermined number D larger than 1.
The knock control device for an internal combustion engine according to claim 3, characterized in that it is created as f=A×D×V_5_0. (5) The knock control device for an internal combustion engine according to claim 4, wherein the predetermined value D changes depending on at least one of operating conditions of the internal combustion engine and cylinders. (6) The determination level correction means adjusts the knock determination level V_r_e_f using the variable A, the median value V_5_0 of the distribution, and a predetermined positive number B, V_r_e_f=(A+
The knock control device for an internal combustion engine according to claim 3, wherein the knock control device for an internal combustion engine is created as B)×V_5_0. (7) The knock control device for an internal combustion engine according to claim 6, wherein the positive number B changes depending on at least one of operating conditions of the internal combustion engine and cylinders. (8) The determination level correction means includes a large value calculation means for calculating a value V_M_N larger than the median value V_5_0 of the distribution of knock intensity values V, and the knock determination level V_r_
The knock determination level V_r_ is set so that the ratio between e_f and a value V_M_N larger than the median value is approximately equal to the ratio between the value V_M_N larger than the median value and the reference level V_L.
Claim 1 characterized in that e_f is corrected.
The knock control device for an internal combustion engine according to item 1 or 2. (9) A knock sensor for detecting knock of an internal combustion engine, a knock intensity value detection means for detecting a knock intensity value V in a predetermined section from the signal of this knock sensor, and a determination level creation means for creating a knock determination level V_r_e_f. and,
The knock intensity value V and the knock determination level V_r_e_f
A knock control device for an internal combustion engine comprising a knock determination means for determining the presence or absence of knock by comparison with the knock intensity, and a drive means for controlling knock control factors such as ignition timing or air-fuel ratio according to the determination result. median calculation means for calculating the median value V_5_0 of the distribution of values V; reference level creation means for creating a reference level V_L_2 that is equal to or higher than the knock intensity value V with a probability smaller than the knock determination probability P_K; A knock control device for an internal combustion engine, comprising a determination level correction means for correcting a knock determination level V_r_e_f so that the ratio with a median value V_5_0 is approximately equal to the ratio between the median value V_5_0 and a reference level V_L_2. .