JP4782035B2 - Combustion control device for internal combustion engine - Google Patents

Description

  The present invention relates to a combustion control apparatus that enables optimal combustion control by extracting only a knock signal by eliminating spike noise in an internal combustion engine such as an automobile engine.

It is generally known that ions are generated when an air-fuel mixture is burned in a combustion chamber of an internal combustion engine. Recently, research on combustion control focusing on the ion current corresponding to this ion has been actively conducted, and an attempt has been made to extract a knock signal superimposed on the ion current and perform combustion control such as retardation control (Patent Literature). 1).
JP 9-338295 A

  This Patent Document 1 discloses a knock intensity detecting means for detecting the strength of knocking generated in an internal combustion engine, an ignition timing retarding means for retarding the engine ignition timing according to the knock intensity during operation, and a response according to the knock intensity during operation. When the engine knock is detected by the valve timing retarding means for retarding the opening / closing timing of the engine intake valve and the knock intensity detecting means, the engine has priority over the ignition timing retarding means when the knock intensity is smaller than a predetermined judgment value. And a knock control means for activating the ignition timing retarding means in preference to the valve timing retarding means when the valve timing retarding means is greater than a determination value. Are listed.

  By adopting such a configuration, it is possible to prevent engine damage by suppressing engine knocking, and to minimize the occurrence of engine output decline, fuel consumption deterioration, exhaust temperature rise, etc. It is said to be.

  Combustion control as described above is based on the premise that only the knock signal can be accurately extracted. However, with conventional methods, when spike noise such as corona noise is superimposed on the ion current, the control performance as intended is achieved. There was a problem that could not be realized. For example, the corona noise is generated by discharging the electric charge charged in the insulator portion of the spark plug, but has a characteristic that it is generated irregularly according to the charged state of the insulator portion. Moreover, since the frequency of the corona noise often coincides with the frequency band of the knock signal, the noise component cannot be reliably removed by filter processing or the like.

  The present invention has been made in view of these problems, and an object of the present invention is to provide a combustion control device that can extract only a knock signal regardless of the frequency band of spike noise.

  In order to achieve the above object, a combustion control apparatus for an internal combustion engine according to claim 1 includes a signal detection unit that detects an ion current corresponding to ions generated during combustion of an air-fuel mixture in a combustion chamber, and the signal detection unit includes: A detection signal to be output is acquired for each sampling period, and a data acquisition unit that stores signal input data, a differential calculation of the signal input data, and a determination as to whether the differential calculation result exceeds a reference value TH A position detection unit that identifies a sudden change start position at which the current suddenly changes, and a noise identification unit that identifies a sudden change start position due to noise by comparing the occurrence timing of the sudden change start position with a knock signal cycle specified in advance. A data correction unit that corrects the signal input data and substantially eliminates noise components before and after the sudden change start position due to the specified noise. There.

The combustion control apparatus for an internal combustion engine according to the present invention includes a signal detection unit that detects an ion current corresponding to ions generated during combustion of an air-fuel mixture in a combustion chamber, and a detection signal output by the signal detection unit. A sudden change start position where the ion current changes abruptly by a data acquisition unit that stores the signal input data acquired every time, and differentially calculates the signal input data and determines whether the differential calculation result exceeds the reference value TH A noise detection unit that identifies a sudden change start position due to noise by comparing a differential calculation result of the sudden change start position with a differential calculation result at a location separated by a predetermined knock signal period. And a data correction unit that corrects the signal input data and substantially eliminates noise components before and after the sudden change start position due to the specified noise. Composed of Te.

  In each of the above inventions, the sampling period τ is preferably set to τ <τn / 2 based on a knock signal period τn specified in advance. Further, it is preferable that the reference value TH is determined based on a statistical value derived from a calculation result of the differential calculation. As the statistical value, it is preferable to employ an average value or standard deviation value of the differential operation result. More preferably, the average value and the standard deviation value are calculated only from the positive value of the differential result.

  The data acquisition unit operates to acquire ion currents in the time range of the cut window, while the position and range on the time axis of the cut window correspond to dynamically changing operating conditions of the internal combustion engine. Are preferably determined.

  According to the present invention described above, it is possible to realize a combustion control apparatus that can accurately extract a knock signal regardless of the frequency band of spike noise.

  Hereinafter, the present invention will be described in detail based on examples.

  FIG. 1A is a circuit diagram illustrating a combustion control device EQU according to an embodiment. This combustion control unit EQU is connected to an ignition coil 1 in which a primary coil 1P and a secondary coil 1S are electromagnetically coupled, a switching transistor 2 that intermittently drives the ignition coil 1, and a secondary coil 1S of the ignition coil. And the one-chip microcomputer 4 that controls the ON / OFF of the switching transistor 2 and receives the analog detection signal SG from the ion current detection circuit 3. A spark plug 5 is connected between the secondary coil 1S of the ignition coil and the ground line.

  As shown, the base terminal of the switching transistor 2 is connected to the one-chip microcomputer 4, the collector terminal is connected to the primary coil 1P of the ignition coil, and the emitter terminal is connected to the ground line.

  The ion current detection circuit 3 includes a bias capacitor C that is charged by the discharge current of the spark plug 5, a Zener diode ZD that is connected in parallel to the capacitor C and regulates the charging voltage of the capacitor C, and is connected in series to the Zener diode ZD. And amplifying unit AMP connected to both ends of diode D1.

  The anode terminals of the Zener diode ZD and the diode D1 are directly connected to each other, and the cathode terminal of the diode D1 is connected to the ground line. The cathode terminal of the Zener diode ZD is connected to the secondary coil 1S.

  The amplification unit AMP of the ion current detection circuit 3 includes an amplification element Q1 having an inverting terminal, a non-inverting terminal, and an output terminal, an input resistor R1 connected to the inverting terminal of the amplification element Q1, and an inverting terminal of the amplification element Q1. The feedback resistor R2 is connected between the output terminals. A diode D2 for protecting the amplifying element Q1 may be connected between the inverting terminal of the amplifying element Q1 and the ground line.

  In this embodiment, an OP amplifier is used as the amplifying element Q1. The input impedance of the OP amplifier is almost infinite (≈∞), and the inverting terminal and the non-inverting terminal are virtually short-circuited (imaginary short). Therefore, the current I shown in FIG. 1B flows in common to the input resistor R1 and the feedback resistor R2, and the output voltage Vout of the amplifier AMP is the product of the current I and the feedback resistor R2 (Vout = I * R2). That is, in this amplification unit AMP, the feedback resistor R2 functions as a detection resistor for the input current I.

  In the circuit configuration of FIG. 1, when a negative high voltage is generated in the secondary coil 1S, the spark plug 5 is ignited and the ignition current charges the capacitor C as shown in FIG. At this time, since the Zener diode ZD is connected in parallel to the capacitor C, the voltage across the capacitor C matches the breakdown voltage Vz of the Zener diode ZD. During this discharge, the diode D1 is short-circuited (ON), so that the current flowing through the input resistor R1 and other circuit elements can be ignored.

  After that, when the high voltage of the secondary coil 1S disappears (see FIG. 1D), the bias voltage charged in the capacitor C is discharged through the path shown in FIG. This discharge current is nothing but the ionic current I (see FIG. 1 (e)). The ionic current I is the output terminal of the amplifying element Q1, the feedback resistance R2, the input resistance R1, the capacitor C, the secondary coil 1S, and the ignition. It flows through the path of the plug 5.

  As described above, since the relationship of the output voltage Vout = R2 × I is established, a voltage proportional to the ion current I is obtained from the amplifying unit AMP.

  The one-chip microcomputer 4 includes a CPU core 4a, an A / D converter 4b, an output port 4c, and a memory unit 4d. The A / D converter 4b directly receives the analog detection signal SG from the ion current detection circuit 3 and converts it into digital data. Further, an ignition pulse is output from the output port 4 c toward the base terminal of the switching transistor 2. In the illustrated combustion control device EQU, the ion current detection circuit 3 and the one-chip microcomputer 4 are directly connected, but a sample hold circuit or the like may be provided in the middle.

  Next, the operation content of the one-chip microcomputer 4 will be described based on the flowchart of FIG. As shown in FIGS. 1 and 2A, first, the one-chip microcomputer 4 receives the analog detection signal SG obtained from the ion current detection circuit 3 directly by the A / D converter 4b, and stores the digitally converted data in the memory. Store in the unit 4d (ST1). In this data acquisition process, since it is necessary to accurately acquire a knock signal superimposed on the ion current, the sampling frequency is set to be twice or more the frequency of the knock signal.

  Here, although the frequency of the knock signal can be experimentally specified based on the structure of the internal combustion engine, in the following description, it is assumed that the frequency of the knock signal is, for example, 8 KHz. Therefore, in this embodiment, the sampling frequency F is about 30 KHz, in other words, the sampling period τ (= 1 / F) is 33 μS as a value that satisfies the condition of the sampling frequency F> 2 × 8 KHz.

  Further, this data acquisition process is executed within the range of the extraction window Win that is set so that the ion current can be reliably captured. FIG. 2B illustrates the relationship between the analog signal waveform obtained from the ion current detection circuit 3 and the extraction window Win with a time width T (= 33 μS × (N−1)). As shown in the figure, a knock signal and spike noise are superimposed after the peak position of the ion current.

  The ion current is generated after the discharge of the spark plug 5 is completed (see FIG. 1 (e)). The discharge timing of the spark plug 5 and the disappearance timing of the generated ion current are determined by the operation such as the engine speed and the engine load. It changes dynamically according to conditions. Therefore, in this embodiment, the relationship between the operating condition of the internal combustion engine and the ion current generated under the operating condition is experimentally obtained in advance, and a window calculation table TBL indicating the relationship is provided in the memory unit 4d. Yes. During actual driving, the window calculation table TBL1 is searched based on data from various sensors indicating driving conditions, and the start position and end position of the cut window Win are determined in real time based on the search result. ing.

  Now, when the data acquisition in the necessary range is completed by the processing of step ST1 in FIG. 2A, next, the N signal input data SG (i) stored in the memory unit 4d is subjected to differential processing by differential calculation. Apply (ST2). As shown in the flowchart of FIG. 3 (a), the differentiation process (ST2) in FIG. 2 (a) performs D (i) ← SG (i) for N data SG (i) of i = 1. ) -SG (i-1) difference calculation (ST21 to ST23).

  FIG. 3B shows a differential waveform obtained by differential operation.

  When the differentiation process in step ST2 in FIG. 2A is completed, the differential value D (i) is compared with the differentiation threshold TH to calculate the steep slope flag Tn (i) (ST3). FIG. 4A is a flowchart showing a calculation process of the steep slope flag Tn (i).

  As shown in the figure, the differential value D (i) and the threshold value TH are compared in ascending order for i = 1 to n (ST32). If D (i) ≧ TH, the previous differential value D (i− On the condition that 1) is smaller than the threshold value TH (ST33), the steep slope flag Tn (i) is set to 1 (ST34). On the other hand, when the differential value D (i) is smaller than the threshold value TH, or even if D (i) ≧ TH, including the previous differential value D (i−1), D (i−1) ≧ If it is TH, the steep slope flag Tn (i) is set to 0 (ST35).

  In short, the above processing means that determination is made in ascending order with respect to i = 1 to n, and when D (i) ≧ TH for the first time, steep inclination flag Tn (i) = 1 is set (see FIG. 4 (c)). Therefore, even if the values of the differential values D (i + 1), D (i + 2)... Are continuously greater than or equal to the threshold value TH, the steep slope flags Tn (i + 1), Tn (i + 2). .. are all set to 0. By such processing, the position where the ion current waveform starts to rise rapidly can be detected. Note that the ion current waveform starts to rise rapidly because a knock signal is superimposed on the ion current or spike noise is superimposed.

  The threshold value TH used in the differentiation process of FIG. 4A is appropriately set based on the acquired ion current waveform and the statistical value of the differentiated waveform. For example, only the differential value (D (i)> 0) indicating a positive value is subjected to a product-sum operation to obtain an average value AV (AV = ΣD (j) / M), and an appropriate constant is added to the calculated average value AV. The threshold TH may be obtained by integrating γ (TH = γ × AV). Alternatively, the average value AV and the standard deviation σ are calculated for the differential value D (i) indicating a positive value, and the threshold value TH is determined based on these values.

  In any case, once the steep slope flag Tn (i) is calculated by the process of FIG. 4A, next, for N steep slope flags Tn (i), a time interval whose value is 1 is determined. . That is, there should be a certain time period α at the steep start position by the knock signal, and the time period α is specified in advance based on the structure of the internal combustion engine. On the other hand, since corona noise occurs randomly, a steep slope start position is searched, and if it occurs at a predetermined cycle τn = α (in this example 1/8 KHz), it is a knock signal. Otherwise, it can be determined that the noise is spike noise.

  The determination process of whether the steep start position is due to a knock signal or spike noise is as shown in FIG. First, it is determined whether or not the steep slope flag Tn (i) is 0 or 1 (ST41). If the steep slope flag Tn (i) = 1, then, before the predetermined period α (Tn (i−α)). Alternatively, it is determined whether or not a steep slope start position exists after a predetermined period α (Tn (i + α)) (ST42). In this embodiment, since the knock signal period is 125 μS (= 1/8 KHz) and the sampling period τ is 33 μS, α≈3.8 (= 125/33), and the integer value is Since it is necessary to adopt, all of Tn (i−3), Tn (i−4), Tn (i + 3), and Tn (i + 4) are to be determined.

  Then, if any one of the steep slope flags Tn (i-3), Tn (i-4), Tn (i + 3), and Tn (i + 4) is present, the steep slope flag Tn (i) = 1 is determined by the knock signal. It is determined that it means the steep start position, and the noise flag NS (i) is set to NS (i) = 0 (ST44). On the other hand, if all of the steep slope flags Tn (i-3), Tn (i-4), Tn (i + 3), and Tn (i + 4) are 0, the steep slope flag Tn (i) = 1 is caused by spike noise. It is determined that it means the steep start position, and the noise flag NS (i) is set to NS (i) = 1 (ST43). In FIG. 4C, spike noise is detected at the position i = X, and NS (X) = 1.

  When spike noise is specified as described above, spike noise is deleted by correcting data that is considered to be spike noise from the signal input data SG (i) (ST5 in FIG. 2). FIG. 5 is a flowchart for specifically explaining the noise removal processing.

  First, it is determined whether or not the noise flag NS (i) is 1 in ascending order for i = 1 to N (ST52). For example, if the noise flag NS (X) is NS (X) = 1, the start reference position BGN and the end reference position END are set in front of the time axis and rearward of the time axis with respect to the position of i = X. Determine (ST53).

  The specific method for determining the start reference position BGN and the end reference position END is appropriate. For example, the start reference position is set to BGN = X-2 on the condition that the steep slope flag Tn (BGN) = 0. It is determined. However, if the steep slope flag Tn (BGN) = Tn (X−2) = 1, BGN = X−1. In this embodiment, since the processing of steps ST32 to ST33 in FIG. 4A is provided, Tn (X) is a steep start position, and Tn (X-1) = It will never be 1.

  On the other hand, the end reference position END is set to END = X + α on condition that Tn (X + 1) to Tn (X + α−1) are all 0. Since α = 3.78 (= 125/33) in this embodiment, α = 4 is used in the determination of Tn (X + 1) to Tn (X + α−1), and the end reference position is END = X + 4. However, if there is a non-zero value in the range of Tn (X + 1) to Tn (X + α-1), for example, if Tn (X + β) is first 1 on the time axis, the end reference position END is set to one. It is assumed that (X + β-1) immediately before.

  If the start reference position BGN and the end reference position END are determined as described above (ST53), SG (BGN) and SG (END) are virtually connected with a straight line for the signal input data SG (i). Then, the signal input data of SG (BGN + 1) to SG (END-1) is linearly interpolated (ST54). FIG. 5C illustrates the interpolation process, and spike noises SG (BGN + 1) to SG (END-1) are substantially removed based on the values of SG (BGN) and SG (END). Has been.

  As described above, since only the noise component can be removed while leaving the knock signal, the BPF processing corresponding to the knock frequency specified in advance (8 KHz in this example) is executed, and the signal input data SG (i ) To extract a knock signal (ST6 in FIG. 2).

  As described above, in the present embodiment, spike noise is eliminated by paying attention to the periodicity of the knock signal. Therefore, spike noise can be eliminated even if the frequency bands of the spike noise and the knock signal match.

  The specific contents described above do not particularly limit the present invention, and various modifications can be made without departing from the spirit of the present invention. In particular, the method of noise detection and the method of removing noise components are not limited to those illustrated.

  For example, in the algorithm of step ST4 of FIG. 4, the knock signal and the noise are distinguished based on only the knock signal generation period α (= τn) specified in advance. It is preferable to extract noise using the fact that noise has a higher frequency than a signal.

  FIG. 6 is a flowchart for explaining another method for noise detection. As shown in the figure, first, after the counter variable CNT is cleared to zero (ST60), the value of the steep inclination flag Tn (i) is set for i = 1 to n as in step ST3 of FIG. However, in this case, the threshold value TH 'is set larger than in the case of step ST3. This threshold value TH ′ is a threshold value for specifically extracting only a sharp rise of noise, and corresponds to operating conditions that change in real time based on the fact that the noise rises sharper than the rise characteristic of the knock signal. The threshold value TH ′ is experimentally determined in advance. However, the threshold value TH ′ may also be determined based on a statistical value derived from the calculation result of the differential calculation. Examples of the statistical value include an average value and a standard deviation value of the differential calculation result, and it is preferable to calculate the average value and the standard deviation value only from a positive value of the differential result.

  In any case, in the algorithm of FIG. 6, the number of the steep slope flags Tn (i) = 1 is counted with the counter variable CNT (ST66). When the processes of steps ST51 to ST67 are completed, it is determined whether or not the value of the counter variable CNT is zero (ST68). If CNT = 0, the process is terminated.

  On the other hand, if the counter variable CNT ≠ 0 in the determination in step ST68, it is determined whether or not the counter variable CNT is 1 (ST69). When the counter variable CYT = 1, it means that a signal (high frequency noise) indicating a particularly steep change has been detected only once, and the process proceeds to the data interpolation process of step ST72. The data interpolation process is the same as in FIG. 5, and it is determined as noise based on the signal input data SG (BGN) to SG (END) before and after the steep slope flag Tn (i) = 1. The signal input data SG (BGN + 1) to SG (END-1) are corrected.

  If it is determined in step ST69 that the counter variable CNT> 1, the generation time interval T of a plurality of detected steep slope flags Tn (i) is calculated (ST70). If the threshold value TH ′ in the determinations in steps ST62 and ST63 is set appropriately, the steep slope flag Tn (i) = 1 is likely to mean noise detection. However, if there are a plurality of locations where the steep slope flag Tn (i) = 1, these may be knock signals. Therefore, the time interval T calculated in the process of step ST70 is compared with the knock period τn specified in advance, and when the time interval T is significantly lower than the knock period τn, the steep slope flag Tn (i ) = 1 is determined to be noise. That is, when the time interval T is T <τn−α that is significantly lower than the knock period τn, the process proceeds to step ST72. On the other hand, if T ≧ τn−α, it is determined that a knock signal has been detected, and the process ends without doing anything.

  The process of FIG. 6 is completed as described above. However, regardless of which route is passed, the process proceeds to the BPF process (ST6) shown in FIG. Thus, if the algorithm of FIG. 6 is used, noise can be detected more precisely, and only the knock signal can be specifically extracted.

  In the above algorithm, since the signal input data SG (i) is differentiated, the spike noise differential value D (i) rebounds positively and negatively, and in some cases cannot be distinguished from the frequency of the knock signal. There is. FIG. 7 illustrates this exceptional state. Since the period of the steep slope flag Tn exceeding the threshold TH coincides with the period of the knock signal, spike noise cannot be removed by the above algorithm. Shows the case.

  On the other hand, although a normal knock signal has a differential value D (i) continuously exceeding a threshold value, a knock signal of a type called a hard knock has a differential value D (i) similar to spike noise. The threshold may not be exceeded continuously. Therefore, if the algorithm described above is used as it is, the knock signal may be mistaken for spike noise and erased.

  Therefore, an algorithm that can cope with any of the above problems is required, and an algorithm that can reliably eliminate spike noise while not erasing hard knocks is desired. Based on the above, further examination has revealed that even when the differential value D (i) of the signal input data after the differential processing exceeds the threshold value TH, by determining the differential value level without erasing the hard knock, It became possible to eliminate only spike noise.

  FIG. 8 (a) illustrates this algorithm. The processing of steps SP1 to SP3 and SP6 to SP7 is the same as the processing of steps ST1 to ST3 and ST5 to ST6 in FIG. 2A, but is characterized by the processing of steps SP4 to SP5.

  The processing content of step SP4 is as shown in FIG. 8B, and among the steep inclination flags Tn (i) of i = 1 to N, Tn (i) = 1, that is, the differential value D ( A starting position i at which i) starts increasing due to a steep slope is extracted (SP40 to SP42).

  Then, the differential values D (i) to D (i + β) are evaluated by an appropriate number β from the start position i of the steep slope toward the rear of the time axis, and the maximum value MAX is extracted (SP43 to SP46). β is set based on the frequency of the noise signal and the knock signal, and is set to about ½ of the signal cycle of the signal having the lower frequency.

  As described above, even when the differential values D (i) to D (i + β) continuously exceed the threshold value TH, the steep slope flag Tn (i) is set to 1 only for the increase start position i, and thereafter The steep inclination flags Tn (i + 1) to Tn (i + β) are all zero. Therefore, there is a possibility that the maximum peak of the differential value is generated behind the steep start position i on the time axis.

  Therefore, for j = i to i + β, the value of the variable MAX initially set to zero is compared with each differential value D (j) (SP44), and the value of the variable MAX is obtained only when D (j)> MAX. Is rewritten (SP45). Then, after this process is completed for i to i + β, the maximum value MAX of the sections i to i + β is stored as the extracted value PK (i) of the steep slope point i. On the other hand, when Tn (i) = 0 among the steep inclination flags Tn (i) of i = 1 to N, the differential value D (i) at that position i is stored as the extracted value PK (i). (SP49).

  FIG. 9B shows the extracted values PK (1) to PK (N) determined in this way. However, actually, it is not always necessary to provide N arrays PK (1) to PK (N), and the differential value D (i) is extracted from the extracted value PK (only for the location i where Tn (i) = 1. It is enough because it was rewritten in i). That is, the arrays PK (1) to PK (N) substitute for the arrays D (1) to D (N).

  In any case, in FIG. 9B, only the steep slope flags Tn (10) and Tn (14) are 1, and only the extracted values PK (10) and PK (14) are the differential values of the peripheral positions. The maximum value MAX is rewritten with D (i) to D (i + β) taken into account.

  Next, the noise determination algorithm will be described based on FIG. As shown in the figure, for i = 1 to N, a place where the steep slope flag Tn (i) = 1 is searched, and the extracted value PK (i) at the place i where Tn (i) = 1, the i-α position, and The extracted values PK (i−α) and PK (i + α) at the i + α position are compared (SP53). As described above, α is the time period α of the knock signal specified in advance based on the structure of the internal combustion engine. In this embodiment, since α≈3.8 (= 125/33), all of PK (i−3), PK (i−4), PK (i + 3), and PK (i + 4) are to be determined. It becomes.

  Then, the ratio RTO with the extracted value PK (i) is calculated, and PK (i) / PK (i-3), PK (i) / PK (i-4), PK (i) / PK (i + 3), If all the ratios RTO calculated by the calculation of PK (i) / PK (i + 4) are values within the range of the reference value 1 / BS to BS, the noise flag NS (i) is set to NS (i) = 0. (SP54). That is, when all the calculated ratios RTO are 1 / BS <RTO <BS in relation to the two reference values 1 / BS and BS, NS (i) = 0.

  On the other hand, any ratio of PK (i) / PK (i-3), PK (i) / PK (i-4), PK (i) / PK (i + 3), PK (i) / PK (i + 4) If RTO becomes a value outside the range of the reference value 1 / BS to BS, the noise flag NS (i) is set to NS (i) = 1 (SP55). That is, when any of the calculated ratios RTO is RTO ≦ 1 / BS in relation to the two reference values 1 / BS and BS, or when RTO ≧ BS, NS (i) = 1. To do.

  In the embodiment shown in FIG. 9, Tn (10) and Tn (14) are 1 at the interval of the time period α of the knock signal. Therefore, in the algorithm of FIG. 2, the noise flags SN (10) and SN (14) are both zero, and cannot be removed although they are spike noise.

  However, according to the algorithm of FIG. 8, it can be correctly recognized that the noise is spike noise. That is, the extracted values PK (10) and PK (14) are at the levels shown in FIG. 9C, and PK (10) / PK (14)> BS, so the noise flag SN (i) is Correctly SN (10) = 1. Since PK (14) / PK (10) <1 / BS, the noise flag SN (i) is correctly set to SN (14) = 1.

  Since the noise flag NS (i) is set correctly in this way, spike noise can be removed by the processing of the subsequent step SP6, and only the knock signal can be accurately extracted by the BPF processing (SP7). . That is, according to the algorithm shown in FIG. 8, even when the differential value of spike noise exceeds the threshold value TH in the time period α of the knock signal, spike noise can be reliably removed.

  Further, according to the algorithm shown in FIG. 8, even when the differential value of the signal input data SG (i) exceeds the threshold value only once (in the case of hard knock), this is not mistaken for noise. That is, in the case of hard knock, PK (i) / PK (i-3), PK (i) / PK (i-4), PK (i) / PK (i + 3), PK (i) / PK ( Since any ratio RTO of i + 4) is a value within the range of the reference value 1 / BS to BS, the noise flag SN (i) = 0 is set (SP54), and SN (i) = 1 is not set. (See FIG. 9 (d)).

It is a circuit diagram which shows the circuit structure of the combustion control apparatus which concerns on an Example. It is the flowchart which shows the outline | summary of control operation of a one-chip microcomputer, and the detection signal waveform of an ion current detection circuit. It is the flowchart which shows a differentiation process, and the wave form diagram of a differentiation waveform. It is a flowchart illustrating a setting process of a steep slope flag and a noise detection flag, and a drawing for explaining an algorithm. It is drawing which shows the flowchart which shows a data interpolation process, and a count algorithm. It is a flowchart explaining another algorithm for noise detection. The noise waveform is difficult to determine. It is a flowchart which shows the outline | summary of another control operation | movement of a one-chip microcomputer. 9 is a flowchart for explaining a part of FIG. 8 in detail, and a drawing for explaining the operation thereof.

Explanation of symbols

EQ Combustion control device 3 Signal detection unit (ion current detection circuit)
SG (i) Signal input data ST1 Data acquisition unit ST2 to ST3 Position detection unit ST4 Noise identification unit ST5 Data correction unit

Claims (5)

A signal detector for detecting an ion current corresponding to ions generated during combustion of the air-fuel mixture in the combustion chamber;
A data acquisition unit that acquires a detection signal output by the signal detection unit for each sampling period and stores signal input data;
A position detection unit that differentiates the signal input data and identifies a sudden change start position at which the ionic current suddenly changes by determining whether or not the differential calculation result exceeds a reference value TH;
A noise specifying unit that specifies a sudden change start position due to noise by comparing the generation timing of the sudden change start position and a knock signal period specified in advance,
A combustion control apparatus for an internal combustion engine, comprising: a data correction unit that corrects the signal input data and substantially eliminates noise components before and after the sudden change start position due to the specified noise.   The combustion control apparatus according to claim 1, wherein the sampling period τ is set to τ <τn / 2 based on a knock signal period τn specified in advance. The combustion control device according to claim 1 or 2 , wherein the reference value TH is determined based on a statistical value derived from a calculation result of the differential operation. While the data acquisition unit operates to acquire the ion current in the time range of the cut window,
The combustion control device according to any one of claims 1 to 3 , wherein a position and a range on the time axis of the cutout window are determined in accordance with an operating condition of the internal combustion engine that dynamically changes. A signal detector for detecting an ion current corresponding to ions generated during combustion of the air-fuel mixture in the combustion chamber;
A data acquisition unit that acquires a detection signal output by the signal detection unit for each sampling period and stores signal input data;
A position detection unit that differentiates the signal input data and identifies a sudden change start position at which the ionic current suddenly changes by determining whether or not the differential calculation result exceeds a reference value TH;
A noise identification unit that identifies the sudden change start position due to noise by comparing the differential calculation result of the sudden change start position with the differential calculation result at a location separated by a predetermined knock signal period;
A combustion control apparatus for an internal combustion engine, comprising: a data correction unit that corrects the signal input data and substantially eliminates noise components before and after the sudden change start position due to the specified noise.

Images