KR20210073839A - Method For Engine Combustion Diagnosis Based On Mechanical Abnormality Using Engine Vibration Signal and Combustion Diagnosis System Thereof - Google Patents

Abstract

According to the present invention, a mechanical combustion disorder diagnosis/determination method using an engine vibration signal of a combustion disorder diagnosis system (1) for an engine (10) includes engine vibration analysis control operations (S10-S60) enabling a diagnosis controller (30) to perform modulation frequency transform with respect to a vibration signal measured including a vibration component by the rotation oscillation of the engine (10); confirming a frequency multiple peak as an odd-number difference multiple of issue occurrence condition vibration signal analysis in contrast to normal condition vibration signal analysis of an integer difference multiple; and determining a combustion disorder issue occurrence cylinder based on an area peak value which is no less than a set threshold value. Therefore, the present invention can accurately determine a cylinder having a combustion disorder signal at a measurement moment based on an odd-number multiple difference component in contrast to an integer multiple difference extracted from the vibration signal.

Description

Method For Engine Combustion Diagnosis Based On Mechanical Abnormality Using Engine Vibration Signal and Combustion Diagnosis System Thereof

The present invention relates to diagnosing engine combustion abnormalities, and more particularly, to a combustion abnormality diagnosis system that ensures accuracy in determining which cylinders generate major signals of failures due to combustion abnormalities from engine vibration signal analysis control using engine vibration signals. will be.

In general, engine combustion control at vehicle age satisfies stable combustion and combustion noise control aspects by controlling combustion robustness in disturbance conditions such as environment, fuel difference, engine aging, and high compression ratio engines such as diesel engines and passenger diesel engines. It is an example of engine control technology of great importance.

Therefore, in the engine combustion control, it is very important to determine whether or not a current failure occurs with the combustion signal for each cylinder (ie, cylinder) of the engine, and to determine in which cylinder the main signal of the failure occurs.

As an example, the engine combustion abnormality determination method is a method in which an abnormality is determined based on the size of a problem frequency domain by FFT (Fast Fourier Transform) transformation using a signal of a knock sensor mounted on an engine block. That is, the magnitude of the amplitude of the response value in a specific frequency domain through FFT transformation is summed in a predetermined band pass filter, and when the value exceeds a specified threshold, it is judged as an abnormal state. way to do it

Therefore, in the engine combustion abnormality determination method, it is determined whether a problem occurs as to whether the magnitude of the vibration amplitude of the frequency band related to the problem occurrence exceeds a threshold through the measured signal through a band pass filter.

US Patent Publication US 2004-0050363 (2004,03,18)

However, the engine combustion abnormality determination method is a method that may be affected by disturbances and the environment because it determines whether there is a problem with the cylinder abnormality based on a reference value.

Therefore, even if the amplitude of vibration in the frequency band related to the problem exceeds the threshold, even if there is a deviation in the amplitude of vibration in the problem frequency region, it is difficult to accurately determine which cylinder is the problematic cylinder of the engine. have no choice but to give

Therefore, the present invention in consideration of the above points can accurately determine the combustion abnormality signal generation cylinder at the time of measurement from the odd multiple difference component of the irregular vibration compared to the integer multiple difference extracted from the vibration signal that is advantageous in terms of delay compared to the noise signal while being able to analyze based on the crank angle. In particular, by applying the TDC (Top Dead Center) standard classification for each cylinder in the large problem vibration area according to the crank angle standard to calculate the sum of odd-numbered difference components, accurate judgment is made as to the cylinder that generates the combustion abnormal signal. An object of the present invention is to provide a method for diagnosing mechanical combustion abnormality using an engine vibration signal and a combustion abnormality diagnosis system.

In the present invention for achieving the above object, in an internal combustion engine equipped with a knocking sensor, a crank position sensor, and a cam position sensor, the region in which vibration occurs based on the crank angle of the vibration signal is determined by TDC (TOP DEAD CENTER) for each cylinder. ), characterized in that the mechanical combustion abnormality diagnosis determination method, characterized in that it is determined whether the internal combustion engine is abnormal by checking the size.

In a preferred embodiment, the vibration signal is measured by a vibration sensor mounted on an engine block of the engine, the crank angle is detected at the crankshaft of the engine, and the camshaft angle is detected at the camshaft of the engine.

As a preferred embodiment, the determination of whether the internal combustion engine is abnormal is performed by converting the modulation frequency of the vibration signal measured together with the signal component by the rotational excitation of the engine by the diagnostic controller, selecting the frequency multiple peak of the odd multiple compared to the integer multiple, It consists of engine vibration signal analysis control in which the region peak value for the vibration signal, which reflects the crank angle and the offset angle, is applied to the determination of the cylinder with the combustion abnormality problem.

As a preferred embodiment, the engine vibration signal analysis control selects a correlation proportional analysis frequency band in which the correlation value increases under the problem occurrence condition through the noise-vibration correlation analysis of the problem occurrence condition and the normal condition in the vibration signal to convert the modulation frequency This is performed modulation frequency analysis control, vibration signal-based analysis control in which the frequency multiple peak is confirmed as the odd multiple of the problem occurrence condition vibration signal analysis compared to the normal condition vibration signal analysis of integer multiples using the correlation proportional analysis frequency band, A crank angle-based analysis control is performed in which a cylinder having a combustion abnormality problem is sold with the region peak value greater than or equal to a set value using the odd multiple frequency.

As a preferred embodiment, the correlation proportional analysis frequency band is set to 500 ~ 3000 Hz through a BPF (Band Pass Filter). The odd-order multiple frequency is applied to the 0.5th order of the integer multiple.

In a preferred embodiment, the modulation frequency analysis control includes the steps of performing downsampling a plurality of times in the correlation proportional analysis frequency band, performing Hilbert transform followed by Envelope transform between downsampling multiple times, and between downsampling multiple times. A step of performing low-pass filter processing in , and a step of performing FFT (Fast Fourier Transform) transformation after down-sampling the plurality of times.

As a preferred embodiment, the plurality of downsampling is performed on the data set in the correlation proportional analysis frequency band, the first downsampling, the second downsampling performed on the envelope-converted data, and the low-pass filter-processed data. It is divided into 3rd order downsampling performed on

As a preferred embodiment, the low-pass filter is set to 300 Hz. The FFT transform applies Hanning Window Overlap 66%.

As a preferred embodiment, the vibration signal-based analysis control includes the steps of detecting a peak average value obtained by using a plurality of odd multiple frequencies as the frequency multiple peak, and setting a peak index using the frequency multiple peak detection count as a peak index A step of comparing the value (Threshold), the peak index is out of the set value of the peak index, and a step of confirming the occurrence of a problem due to abnormal combustion is performed.

In a preferred embodiment, the peak index setting value is 10.

As a preferred embodiment, the crank angle-based analysis control includes the steps of performing modulation frequency analysis on the odd multiple frequency, calculating the region peak value by dividing the crank angle reference for each cylinder region of the engine to determine the occurrence of a problem (Threshold) ), and the region peak value deviates from the problem occurrence determination set value and is determined as the problem occurrence cylinder.

As a preferred embodiment, the modulation frequency analysis of the crank angle-based analysis control includes the steps of performing first down sampling on the odd multiple frequency, performing the Hilbert transform followed by the envelope transform, and the envelope transformed It consists of a step of performing secondary downsampling on data, a step of performing low-pass filter processing, and a step of performing third-order downsampling performed on the low-pass-filtered data.

As a preferred embodiment, the calculation of the region peak value is carried out through envelope analysis by combining a time-based vibration signal with a crank/cam angle signal to obtain a crank angle-based vibration analysis result.

In a preferred embodiment, the problem occurrence determination setting value is 50.

And, in the combustion abnormality diagnosis system of the present invention for achieving the above object, modulation frequency conversion is performed on the vibration signal measured by including a signal component due to rotational excitation of the engine, and the frequency multiple peak is the normal value of the integer multiple. It is confirmed as an odd multiple of the conditional vibration signal analysis compared to the conditional vibration signal analysis, and when the area peak value for the vibration signal of the cylinder reflecting the crank angle and the opening angle is greater than the set value (Threshold), it determines the cylinder with the combustion problem. diagnostic controller; It is characterized in that it includes a vibration sensor that is mounted on the engine block of the engine to measure the vibration generated in the engine when the engine is operating as a frequency signal.

In a preferred embodiment, the diagnostic controller detects the crank angle as a crank angle signal from a crankshaft of the engine, and detects the offset angle as an offset angle signal from a camshaft of the engine.

In a preferred embodiment, in the diagnostic controller, modulation frequency analysis and time-based frequency analysis are performed using the vibration frequency of the vibration sensor, and the crank angle-based analysis is performed by combining the crank angle and the cam angle with the vibration frequency. When the area peak value is higher than the set value (Threshold), it is determined that the cylinder has a combustion error problem.

The method for diagnosing mechanical combustion abnormality using the engine vibration signal applied to the combustion abnormality diagnosis system of the present invention implements the following actions and effects.

First, the crank angle-based analysis can be made with the characteristics of the vibration signal advantageous in terms of delay compared to the noise signal while taking advantage of the use of vibration signals that enable continuous monitoring in real time for each cylinder of the engine. Second, by analyzing the vibration signal based on the crank angle, it is possible to check the location of the problem vibration based on the crank angle, so that the problem cylinder can be classified by comparing the magnitude of the vibration signal generated at that time. Third, it is possible to perform abnormal diagnosis by performing crank angle-based vibration analysis and time-based vibration analysis in parallel, and to more accurately determine the cause of the problem through estimating the problem cylinder. Fourth, by judging with the size at which vibrations due to components other than the periodic excitation force generated by the engine are generated due to a failure, a more reliable cylinder classification is possible by comparing the size of a specific frequency band, unlike the existing logic that is vulnerable to disturbances. Fifth, the performance of the combustion abnormality diagnosis system can be improved without changing the engine system by using a knock sensor mounted on the engine block or a separate vibration sensor additionally mounted on the engine block to determine whether a mechanical abnormality occurs.

1 and 2 are flowcharts of a method for diagnosing mechanical combustion abnormality using an engine vibration signal according to the present invention, and FIG. 3 is a combustion abnormality diagnosis system in which a mechanical combustion abnormality diagnosis determination control using an engine vibration signal according to the present invention is performed. 4 is an example of vibration data 1 and 2 diagrams in which a correlation value increases under a problem occurrence condition from a measurement signal of a vibration sensor according to the present invention, and FIG. 5 is a modulation frequency analysis control according to the present invention. It is an example of calculating the magnitude by the modulation frequency through the modulation frequency, FIG. 6 is an example of determining a condition for generating a problem compared to a normal condition based on a vibration signal according to the present invention, and FIG. 7 is an example of determining a condition for generating a problem according to the present invention. It is an example of the comparison of the judgment performance of the rust sensor and the vibration sensor for the determination of the occurrence condition, FIG. 8 is an example of detecting a cylinder with a problem calculated from the crank angle-based analysis control according to the present invention, and FIG. 9 is the crank angle according to the present invention It is the result state of the problematic cylinder determined by the basic analysis control.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying illustrative drawings, and since such an embodiment may be implemented in various different forms by those of ordinary skill in the art to which the present invention pertains as an example, it will be described herein It is not limited to the embodiment.

1 and 2 , in the method for diagnosing mechanical combustion abnormality, the abnormality diagnosis is made by the engine vibration signal analysis control (S10 to S60), so that the combustion abnormal cylinder is accurately determined from the occurrence of the combustion abnormality problem in the engine cylinder.

In particular, the engine vibration signal analysis control (S10 to S60) is subjected to down sampling for the vibration frequency selection control (S10 to S20) of the vibration signal measured during engine operation, and is converted to FFT (Fast Fourier Transform) under normal conditions. Frequency analysis control (S30) to modulate the vibration level of the odd multiples of the 0.5th order (0.5, 1.5, 2.5,3.5, 4.5, 5.5, ...) compared to the rotation integer multiples (1,2,3,4,5, …) of ), and after accurately determining the main signal generating cylinder of the current failure with the vibration signal-based analysis control (S40) from the odd multiple, the crank angle-based analysis control (S50) ensures the accuracy of determining the cylinder for combustion abnormality.

As a result, the mechanical combustion abnormality diagnosis determination method is characterized as a mechanical combustion abnormality diagnosis determination method using an engine vibration signal, and the mechanical combustion abnormality diagnosis determination method using the engine vibration signal can implement the following advantages.

First, based on the signal analyzed based on the crank angle, it is possible to determine whether a current failure has occurred and to determine which cylinder the main signal of the failure occurs in. Second, by judging based on the crank angle, it is possible to classify the area where the problem vibration occurs greatly based on the TDC for each cylinder. Third, when a problem vibration occurs, it is possible to accurately determine whether there is a problem by the size of the sum of components having an odd multiple of 0.5th order by paying attention to the fact that the 0.5th order component is large because the vibration occurs irregularly, not at the multiplier of the natural number of engine rotation. have.

Referring to FIG. 3 , the combustion abnormality diagnosis system 1 includes an engine 10 , vibration signal measuring devices 21 , 23 , 25 , and a diagnosis controller 30 .

Specifically, the engine 10 is a conventional gasoline or diesel engine, and the vibration signal measuring devices 21 , 23 , and 25 are mounted on the engine block and detect vibrations generated in the engine 10 as a frequency signal during engine operation. a vibration sensor 21 to perform a motion, a crank angle sensor 23 for detecting a rotation angle of the crankshaft rotated by the piston stroke of the engine 10, and a cam sensor for detecting a rotation angle of a camshaft operating the valve system in the cylinder head It consists of (25).

In addition, the vibration sensor 21, the crank angle sensor 23, and the cam sensor 25 use the diagnostic controller 30 and a controller area network (CAN) as a communication network, and the CAN is the vibration sensor ( 21), the crank angle signal of the crank angle sensor 23, and the cam angle signal of the cam sensor 25 are provided to the diagnostic controller 30 as input data.

Specifically, the diagnostic controller 30 utilizes the vibration signal, crank angle, and cam angle as input data to determine a vibration signal-based mechanical abnormality diagnosis, determine a problem cylinder (ie, cylinder) using a crank angle-based analysis of the vibration signal, Logic processing is performed to determine whether there is a problem through time-based modulation frequency analysis.

In particular, the diagnostic controller 30 is equipped with a program or algorithm for modulation frequency analysis control (S30), vibration signal-based analysis control (S40), and crank angle-based analysis control (S50) in a memory, and logic processing of the program or algorithm It operates as a central processing unit that implements

To this end, the program or algorithm performs the modulation frequency analysis and time-based frequency analysis of the vibration sensor 21 based on the vibration frequency, and the vibration frequency of the vibration sensor 21 and the crank angle and the cam of the crank angle sensor 23 . The cam angle analysis of the sensor 25 is divided into a part that is performed based on the crank angle. To this end, the diagnosis controller 30 may include a vibration analysis module for performing vibration frequency-based performance and a crank analysis module for performing crank angle-based performance.

Hereinafter, a method for diagnosing mechanical combustion abnormality using the engine vibration signal of FIGS. 1 and 2 will be described in detail with reference to FIGS. 3 to 9 . In this case, the control subject is the diagnostic controller 30 , and the control object is the engine 10 and the injector fuel injection pattern.

First, the diagnostic controller 30 performs vibration frequency selection control (S10 to S20) of the engine 10 as a vibration signal measurement step according to engine operation in S10 and an analysis frequency region selection step in S20.

Referring to FIG. 3 , the diagnostic controller 30 includes the vibration sensor 21 of the engine block together with the crank angle detected by the crank angle sensor 23 and the cam angle detected by the cam sensor 25 in the engine 10 in operation. ) by checking the detected vibration signal as input data through CAN to measure the vibration signal according to engine operation (S10). Therefore, the diagnostic controller 30 reads the crank angle, the cam angle, and the vibration signal together with the vibration signal to measure the vibration signal according to the engine operation ( S10 ).

In addition, the diagnostic controller 30 analyzes the noise data of the vibration signal continuously measured by the vibration analysis module 31 through BPF (Band Pass Filter) processing and selects the frequency region (S20) to determine the problem occurrence condition and the normal condition. Frequency domain definition is made through noise/vibration correlation analysis. In this case, the BPF process applies a low pass filter of 500 Hz and a high pass filter of 3000 Hz. In particular, the BPF (Band Pass Filter) may be replaced with a Butterworth Filter.

Referring to FIG. 4 , the diagnostic controller 30 may acquire vibration data 1 and 2 having a large correlation value through BPF processing, and from the vibration data 1 and 2, the correlation value increases in the 500 to 3000 Hz band under the problem occurrence condition. it can be seen that there is

From this, the analysis frequency domain selection (S20) is a frequency domain definition through the noise/vibration correlation analysis of the problem occurrence condition and the normal condition for the vibration signal according to the operation of the engine 10, and the correlation value in the problem occurrence condition This increasing frequency region of the 500 ~ 3000 Hz band is selected as the correlation proportional analysis frequency region.

Next, the diagnostic controller 30 repeats the modulation frequency analysis control (S30) of the down sampling step of S31, the Hilbert conversion step of S32, the envelope conversion step of S33, and the down sampling of S34. It is performed as a performing step, a low pass filter processing step of S35, a re-performing step of down sampling in S36, and a Fast Fourier Transform (FFT) transformation step of S37. In this case, the down sampling refers to a signal processing process of lowering the original sampling rate of a signal in order to reduce the size of data that cannot be contained by implying everything.

Referring to FIG. 5 , the diagnostic controller 30 uses time data 2 sec before the measurement time reference as a correlation proportional analysis frequency domain for real-time calculation with a vibration analysis program or logic in a 500 to 3000 Hz band by a modulation frequency. Perform size calculations.

To this end, an amplitude modulation signal is calculated as a modulation frequency magnitude using a modulation frequency and a modulation index.

Signal: x(t) = A _c (t)x(1 + m cos(ω _m t) x cos(ω _m t)

Modulation frequency formula: f _m (HZ) = ω _m /2π

Modulation index formula: m = A _m /A _c

where “t” is the time in seconds, “A _m “ is the amplitude of the high-frequency excitation, “A _c “ is the amplitude of the low-frequency excitation, and “m” is the modulation representing the ratio of the low-frequency excitation amplitude to the high-frequency excitation amplitude. index, “ω _m ” is the angular frequency of the modulated wave, and “f _m (HZ)” is the modulation frequency representing the low-frequency component that transmits the high-frequency component.

As an example, the down-sampling (S31) is a primary down-sampling process for lowering the sampling rate, which is the amount of data collected for 1 second for a signal (that is, the correlation proportional analysis frequency domain of the 500 to 3000 Hz band), the Repeat downsampling (S34) is a secondary downsampling process that lowers the sampling rate, which is the amount of data collected for 1 second for the Hilbert/Envelope converted signal, especially considering the low pass filter frequency. The down-sampling re-performing ( S36 ) is a tertiary down-sampling process for lowering the sampling rate, which is the data collection amount for 1 second for the low-pass filter-processed signal.

For example, the Hilbert transformation (S32) converts cos(ω _m t) into sin(ω _m t) with a 90 degree phase change, draws the converted amplitude, and extracts an envelope formation. The envelope transformation (S33) extracts a line drawn so as to surround the waveform by connecting the ends of the waveform to each other.

For example, in the low pass filter processing ( S35 ), 300 Hz is set as a Low Pass Frequency by passing a low frequency component while blocking a high frequency component of the signal. The FFT transform (S37) is used for high-speed calculations such as frequency domain analysis of discrete time signals, convolution calculations in frequency domain, correlation function calculation, numerical analysis, and the like. In this case, in the FFT transformation, Hanning Window Overlap 66% is applied in Hanning Window, which converts an aperiodic signal into a periodic signal with a window, so that data overlap is applied to prevent omission of the window effect.

From this, the modulation frequency analysis control (S30) is performed from the correlation proportional analysis frequency region (eg, 500 to 3000 Hz) obtained in the vibration frequency selection control (S10 to S20) in which the measured vibration signal is filtered with a band pass filter (BPF). By classifying the signal component by the excitation force generated in the engine, it contributes to the confirmation of the magnitude of the vibration caused by the excitation when a problem occurs due to the occurrence of mechanical abnormalities.

Thereafter, the diagnostic controller 30 performs the vibration signal-based analysis control (S40) in the frequency multiple peak detection step of S41, the peak index application step of S42, the peak index determination step of S43, and the problem occurrence confirmation step of S44 due to abnormal combustion. do.

Referring to FIG. 6 , the diagnostic controller 30 exemplifies the vibration signal analysis result under the problem occurrence condition compared to the vibration signal analysis result under the normal condition.

As shown, in the normal condition, it is exemplified that the vibration level due to the integer multiple of rotation (1,2,3,4,5, ...) is greatly generated, but in the above problematic condition, 0.5 in rotation compared to the normal condition It is exemplified that the vibration level caused by the odd multiples (0.5, 1.5, 2.5, 3.5, 4.5, 5.5, ...) of the difference is relatively large. The reason for this is that the level of vibration caused by excitation in a specific cylinder (eg, a cylinder with problem parts) is large.

For example, in the frequency multiple peak detection (S41), the average value of the peak obtained from the summation of the 0.5 order odd multiple difference (0.5, 1.5, 2.5, 3.5, 4.5, 5.5, ...) in rotation compared to the normal condition is used as the frequency multiple peak. set

For example, the peak index application ( S42 ) applies a peak index accumulation formula, and the peak index determination ( S43 ) applies a peak index discriminant.

Peak index cumulative formula: I

Peak index discriminant: Index > K

Here, "N" applies about 10 as the number of HOMI (Half Order Modulation Index), and "K" applies about 10 (Dimensionless) as the peak index set value (Threshold).

As an example, the check for occurrence of a problem due to the abnormal combustion ( S44 ) is performed when the accumulated value of the peak index is greater than 10.

Referring to FIG. 7 , as a result of HOMI6 analysis using a knock sensor signal, the 1300RPM condition exemplifies the best abnormal diagnosis judgment performance, but it has been experimentally proven that a vibration sensor with high sensitivity exhibits better judgment performance.

HOMI _N : Half Order Modulation Index (N)

From this, the vibration signal-based analysis control (S40) uses a vibration signal that is advantageous in terms of delay compared to noise signal while being able to analyze based on crank angle, so that the occurrence of a combustion abnormality problem is confirmed based on the signal analyzed based on the crank angle, and such combustion It improves the accuracy of determining the cylinder that generates the main signal of the current failure through the occurrence of an abnormal problem.

Finally, the diagnostic controller 30 confirms the problem cylinder analysis result in the problem cylinder confirmation step of S60 from the result of performing the crank angle based analysis control ( S50 ).

As a result, the diagnostic controller 30 returns to the vibration signal measurement step of S10 and repeats the combustion abnormal diagnosis procedure when the occurrence of a combustion abnormality problem is not confirmed in the problem occurrence cylinder confirmation (S60), while the combustion abnormality problem occurrence is confirmed. In this case, after terminating the diagnostic logic, a buzzer or text notification is performed for the cylinder with the problem with combustion, and then the cylinder with the problem with combustion is stored in the memory.

On the other hand, referring to FIG. 2 , the diagnostic controller 30 modulates the crank angle-based analysis control ( S50 ) and modulates the frequency analysis control ( S30 ) to determine the cylinder where the combustion abnormality occurs for the problem occurrence situation, and then the cylinder area of S51 The step of discriminating based on the crank angle for each star, the step of determining the peak value of S52, and the step of determining the cylinder with the problem of S53 are performed.

Referring to FIG. 8 , the modulation frequency analysis control ( S30 ) repeats the down sampling step of S31 , the Hilbert transform step of S32 , the envelope conversion step of S33 , and down sampling of S34 repeatedly. Step, a low pass filter processing step of S35, and a re-performing step of down sampling of S36 are performed. Therefore, the diagnostic controller 30 uses the result value of down sampling re-performation (S36) in the crank angle reference classification step for each cylinder area of S51, so that the FFT conversion step of S37 is performed in the crank angle based analysis control (S50). There are differences that are not performed.

In addition, the diagnostic controller 30 uses the time data 2 sec prior to the measurement time as a time-based vibration signal for real-time calculation of the cylinder of the problem occurrence with the crank analysis program or logic, and uses this as a time-based vibration signal to calculate the crank angle and the cam angle. The crank angle-based vibration analysis result obtained by combining with the crank angle signal for the crank angle is obtained and the crank angle-based envelope analysis result is obtained by performing the crank angle reference classification (S51) for each cylinder area by the envelope analysis. In this case, the crank angle-based envelope analysis result is the crank angle of cylinder #1 (Cyl1), cylinder #3 (Cyl3), cylinder #4 (Cyl4), and cylinder #2 (Cyl2) for the four cylinders of the engine 10 . Divided by angle.

As a result, the crank angle reference classification (S51) for each cylinder area is the area peak value (Cylinder) of the envelope analysis value having a set value (Threshold) or more that can specify a specific cylinder as a problem cylinder among cylinders #1 to #4. Peak Value) is calculated. In this case, it is confirmed that the Cylinder Peak Value above the threshold occurred in cylinder #3 and cylinder #4.

For example, the peak value determination (S520) applies the regional peak value discriminant, and the problem occurrence cylinder determination (S53) is determined to be a cylinder with a region peak value greater than or equal to a set value (Threshold).

Area peak value discriminant: Peak value within the area > A

Here, “A” is the problem occurrence determination threshold, and about 50 (Dimensionless) is applied.

Referring to FIG. 9 , the experimental results for the problematic cylinder obtained by the crank angle based analysis control ( S50 ) are exemplified.

As shown, after passing the BPF (Band Pass Filter) of the measured vibration signal, the envelope obtained through Down Sampling and Hilbert Transform is analyzed based on the crank angle and classified according to the TDC standard of each cylinder. is exemplified

Therefore, from the filtered vibration signal (crank-based analysis) and the envelope analysis of the filtered vibration signal (crank-based analysis), cylinder #3(Cyl3)/cylinder #4(Cyl4) compared to the normal engine whose Envelope peak value is less than Threshold(100) Cylinder #3 (Cyl3) and cylinder #4 (Cyl4) or cylinder #3 (Cyl3) or cylinder #4 (Cyl4) in the area of It can be seen that this result is exemplified by the analysis.

From this, the crank angle-based analysis control (S50) is performed by using the sum size of the odd multiple difference component of the irregular vibration as a TDC (Top Dead Center) standard for each cylinder in the large problem vibration area according to the crank angle standard. It greatly improves the accuracy of identifying the problem cylinder that causes combustion abnormality among the 4 areas.

As described above, the method for diagnosing mechanical combustion abnormality using the engine vibration signal applied to the combustion abnormality diagnosis system 1 of the engine 10 according to the present embodiment is the rotational excitation of the engine 10 by the diagnostic controller 30 . Modulation frequency transformation is performed on the measured vibration signal including the signal component by And by including the engine vibration signal analysis control (S10~S60) in which the cylinder having a combustion abnormality problem is judged by the peak value of the region above the set value (Threshold), the combustion abnormality at the time of measurement from the integer multiple difference compared to the integer multiple difference extracted from the vibration signal Signal generation cylinder determination is made accurately.

1: Combustion abnormal diagnosis system 10: Engine
21: vibration sensor 23: crank angle sensor
25: cam sensor 30: diagnostic controller

Claims (20)

In an internal combustion engine equipped with a knocking sensor, a crank position sensor, and a cam position sensor,
A method for diagnosing mechanical combustion abnormality, characterized in that it is determined whether the internal combustion engine is abnormal by checking the size of the area where vibration occurs based on the crank angle based on the TDC (TOP DEAD CENTER) for each cylinder.
The method according to claim 1, wherein the vibration signal is measured by a vibration sensor mounted on the engine block of the engine, the crank angle is detected in the crankshaft of the engine, and the camshaft angle is detected in the camshaft of the engine. A method for diagnosing mechanical combustion abnormalities using
The method according to claim 1, wherein the determination of whether the internal combustion engine is abnormal is a modulation frequency transformation of a vibration signal measured together with a signal component due to rotational excitation of the engine by a diagnostic controller, Frequency multiple peak selection, engine vibration signal analysis control applied to the area peak value for vibration signal reflecting crank angle and opening angle to determine combustion abnormality problem cylinder
Method for diagnosing mechanical combustion abnormality, characterized in that consisting of.
The method according to claim 3, wherein the engine vibration signal analysis control selects a correlation proportional analysis frequency band in which the correlation value increases under the problem occurrence condition through noise-vibration correlation analysis between the problem occurrence condition and the normal condition in the vibration signal to convert the modulation frequency This is performed modulation frequency analysis control, a vibration signal-based analysis control in which a frequency multiple peak is identified as an odd multiple of a problem occurrence condition vibration signal analysis compared to a normal condition vibration signal analysis of an integer multiple using the correlation proportional analysis frequency band; Crank angle-based analysis control in which a cylinder having a combustion abnormality problem is determined by the region peak value above a set value using the odd multiple frequency
Method for diagnosing mechanical combustion abnormality, characterized in that it is performed as
The method according to claim 4, wherein the correlation proportional analysis frequency band is set to 500 ~ 3000 Hz through a BPF (Band Pass Filter).
The method according to claim 4, wherein the odd-order multiple frequency is applied as a 0.5-order multiple of the integer multiple.
The method according to claim 4, wherein the modulation frequency analysis control comprises the steps of performing down sampling a plurality of times in the correlation proportional analysis frequency band, performing Hilbert transform followed by Envelope transform between the multiple times of downsampling, and the plurality of A step of performing low-pass filter processing between downsampling times, and a step of performing FFT (Fast Fourier Transform) transformation after downsampling a plurality of times.
A method for diagnosing mechanical combustion abnormality, characterized in that it is performed as
The method according to claim 7, wherein the plurality of down-sampling is a first-order down-sampling performed on the data set to the correlation proportional analysis frequency band, a second-order down-sampling performed on the envelope-converted data, and the low-pass filter-processed data. Method for diagnosing mechanical combustion abnormality, characterized in that it is divided into tertiary downsampling performed for
The method of claim 7, wherein the low-pass filter is set to 300 Hz.
The method for diagnosing mechanical combustion abnormality according to claim 7, wherein the FFT conversion is performed by applying a Hanning Window Overlap of 66%.
The method according to claim 4, wherein the vibration signal-based analysis control comprises the steps of: detecting a peak average value obtained by using a plurality of odd-order multiple frequencies as the frequency multiple peak; setting a peak index using the frequency multiple peak detection count as a peak index Comparing with a value (Threshold), the peak index is out of the set value of the peak index, checking the occurrence of a problem due to abnormal combustion
A method for diagnosing mechanical combustion abnormality, characterized in that it is performed as
The method of claim 11, wherein 10 is applied to the peak index setting value.
The method according to claim 4, wherein the crank angle-based analysis control comprises the steps of performing modulation frequency analysis on the odd multiple frequency, calculating a region peak value by dividing the crank angle reference for each cylinder region of the engine to determine the occurrence of a problem (Threshold) ) and the step of determining that the region peak value is out of the problem occurrence determination set value as the problem occurrence cylinder
A method for diagnosing mechanical combustion abnormality, characterized in that it is performed as
The method according to claim 13, wherein the modulation frequency analysis comprises performing a first down sampling step on the odd multiple frequency, performing an envelope transform following Hilbert transform, and performing a second downsampling on the envelope transformed data. Sampling is performed, low-pass filter processing is performed, and third-order down-sampling performed on the low-pass filter-processed data is performed
Method for diagnosing mechanical combustion abnormality, characterized in that consisting of.
The method as set forth in claim 13 , wherein the calculation of the region peak value is carried out through envelope analysis by combining a time-based vibration signal with a crank/cam angle signal to obtain a crank angle-based vibration analysis result.
The method according to claim 13, wherein 50 is applied to the set value for determining the occurrence of the problem.
Modulation frequency transformation is performed on the measured vibration signal by including the signal component due to the rotational excitation of the engine, and the frequency multiple peak is an integer multiple of the normal condition vibration signal analysis compared to the normal condition vibration signal analysis. A diagnostic controller that determines a cylinder having a combustion problem when the peak value of the vibration signal of the cylinder, which is checked as an odd multiple, and the crank angle and the opening angle is greater than or equal to the set value (Threshold);
A vibration sensor that is mounted on the engine block of the engine and measures the vibration generated by the engine as a frequency signal during engine operation
Combustion abnormal diagnosis system, characterized in that it is included.
The combustion abnormality diagnosis system according to claim 17, wherein the diagnostic controller detects the crank angle as a crank angle signal from a crankshaft of the engine, and detects the offset angle as an offset signal from a camshaft of the engine. .
The combustion abnormality diagnosis system according to claim 17, wherein the diagnosis controller performs modulation frequency analysis and time-based frequency analysis using the vibration frequency of the vibration sensor.
The method according to claim 17, wherein the diagnostic controller determines the cylinder in which the combustion abnormality occurs when the peak value of the region is greater than or equal to a set value (Threshold) through a crank angle-based analysis that combines the crank angle and the cam angle with the vibration frequency. combustion abnormality diagnosis system.

Images