초록 ▼

A knocking detecting apparatus for an internal combustion engine, which can more accurately perform the knocking determination based on frequency components of the output signal of the knock sensor, is provided. A frequency spectrum analysis of the knock sensor output signal is performed at predeter

A knocking detecting apparatus for an internal combustion engine, which can more accurately perform the knocking determination based on frequency components of the output signal of the knock sensor, is provided. A frequency spectrum analysis of the knock sensor output signal is performed at predetermined angular intervals of the crankshaft rotation angle. Intensities of a plurality of frequency components obtained by the frequency spectrum analysis are stored as time series data. The time series data of the frequency component intensity are binarized, and it is determined whether or not a knocking has occurred based on the binarized time series data.

대표청구항 ▼

1. A knocking detecting apparatus for an internal combustion engine having knocking determining means for performing a knocking determination based on an output signal of a knock sensor mounted on said engine, said knocking detecting apparatus comprising: frequency spectrum analyzing means for perfo

1. A knocking detecting apparatus for an internal combustion engine having knocking determining means for performing a knocking determination based on an output signal of a knock sensor mounted on said engine, said knocking detecting apparatus comprising: frequency spectrum analyzing means for performing a frequency spectrum analysis of the knock sensor output signal at predetermined angular intervals of the crankshaft rotation angle of said engine;data storing means for storing intensities of a plurality of frequency components obtained by the frequency spectrum analysis as time series data; andbinarizing means for binarizing the time series data of the frequency component intensity;wherein said knocking determining means determines whether or not a knocking has occurred based on the binarized time series data. 2. A knocking detecting apparatus according to claim 1, wherein said knocking determining means performs the knocking determination by comparing the binarized time series data with a master pattern data corresponding to a state where a knocking has occurred. 3. A knocking detecting apparatus according to claim 2, wherein said knocking determining means includes: noise component calculating means for calculating time series data of noise components based on the binarized time series data; andnoise correcting means for correcting the binarized time series data with the time series data of noise components,wherein said knocking determining means performs the knocking determination based on the corrected time series data. 4. A knocking detecting apparatus according to claim 2, further comprising: noise component calculating means for calculating time series data of noise components based on the time series data of the frequency component intensity; andnoise correcting means for correcting the time series data of the frequency component intensity with the time series data of noise components,wherein said binarizing means binarizes the corrected time series data of the frequency component intensity. 5. A knocking detecting apparatus according to claim 2, wherein said knocking determining means includes weighting means for performing a weighting of the binarized time series data and the master pattern data according to a frequency of the frequency component, and said knocking determining means performs the knocking determination using the weighted time series data and the weighted master pattern data. 6. A knocking detecting apparatus according to claim 2, wherein said knocking determining means calculates a fitting ratio of the binarized time series data with respect to the master pattern data, and determines that a knocking has occurred when the fitting ratio exceeds a determination threshold value. 7. A knocking detecting apparatus according to claim 2, wherein the master pattern data are set according to an operating condition of said engine. 8. A knocking detecting apparatus according to claim 1, further comprising frequency group time series data generating means for generating frequency group time series data by selecting a maximum strength at one sampling timing with respect to each of frequency groups consisting of a plurality of frequencies, wherein said binarizing means binarizes the frequency group time series data. 9. A knocking detecting apparatus according to claim 8, further comprising: group noise component calculating means for calculating time series data of group noise components corresponding to each of the frequency groups based on the frequency group time series data; andgroup noise correcting means for correcting the frequency group time series data with the time series data of group noise components,wherein said binarizing means binarizes the corrected frequency group time series data. 10. A knocking detecting method for an internal combustion engine, which performs a knocking determination based on an output signal of a knock sensor mounted on said engine, said knocking detecting method comprising the steps of: a) performing a frequency spectrum analysis of the knock sensor output signal at predetermined angular intervals of the crankshaft rotation angle of said engine;b) storing intensities of a plurality of frequency components obtained by the frequency spectrum analysis as time series data;c) binarizing the time series data of the frequency component intensity; andd) determining whether or not a knocking has occurred based on the binarized time series data. 11. A knocking detecting method according to claim 10, wherein the knocking determination is performed in said step d) by comparing the binarized time series data with a master pattern data corresponding to a state where a knocking has occurred. 12. A knocking detecting method according to claim 11, wherein said step d) includes the steps of: e) calculating time series data of noise components based on the binarized time series data;f) correcting the binarized time series data with the time series data of noise components; andg) performing the knocking determination based on the corrected time series data. 13. A knocking detecting method according to claim 11, further comprising the steps of: e′) calculating time series data of noise components based on the time series data of the frequency component intensity; andf′) correcting the time series data of the frequency component intensity with the time series data of noise components,wherein the corrected time series data of the frequency component intensity are binarized in said step c). 14. A knocking detecting method according to claim 11, wherein said step d) includes the step of h) performing a weighting of the binarized time series data and the master pattern data according to a frequency of the frequency component, and the knocking determination is performed using the weighted time series data and the weighted master pattern data. 15. A knocking detecting method according to claim 11, wherein said step d) includes the step of i) calculating a fitting ratio of the binarized time series data with respect to the master pattern data, and it is determined that a knocking has occurred when the fitting ratio exceeds a determination threshold value. 16. A knocking detecting method according to claim 11, wherein the master pattern data are set according to an operating condition of said engine. 17. A knocking detecting method according to claim 10, further comprising the step of j) generating frequency group time series data by selecting a maximum strength at one sampling timing with respect to each of frequency groups consisting of a plurality of frequencies, wherein the frequency group time series data are binarized in said step c). 18. A knocking detecting method according to claim 17, further comprising the steps of: k) calculating time series data of group noise components corresponding to each of the frequency groups based on the frequency group time series data; andl) correcting the frequency group time series data with the time series data of group noise components,wherein the corrected frequency group time series data is binarized in said step c). 19. A computer program embodied on a computer-readable medium for causing a computer to implement a knocking detecting method for an internal combustion engine, which performs a knocking determination based on an output signal of a knock sensor mounted on said engine, said knocking detecting method comprising the steps of: a) performing a frequency spectrum analysis of the knock sensor output signal at predetermined angular intervals of the crankshaft rotation angle of said engine;b) storing intensities of a plurality of frequency components obtained by the frequency spectrum analysis as time series data;c) binarizing the time series data of the frequency component intensity; andd) determining whether or not a knocking has occurred based on the binarized time series data. 20. A computer program according to claim 19, wherein the knocking determination is performed in said step d) by comparing the binarized time series data with a master pattern data corresponding to a state where a knocking has occurred. 21. A computer program according to claim 20, wherein said step d) includes the steps of: e) calculating time series data of noise components based on the binarized time series data;f) correcting the binarized time series data with the time series data of noise components; andg) performing the knocking determination based on the corrected time series data. 22. A computer program according to claim 20, wherein said knocking detecting method further comprises the steps of: e′) calculating time series data of noise components based on the time series data of the frequency component intensity; andf′) correcting the time series data of the frequency component intensity with the time series data of noise components,wherein the corrected time series data of the frequency component intensity are binarized in said step c). 23. A computer program according to claim 20, wherein said step d) includes the step of h) performing a weighting of the binarized time series data and the master pattern data according to a frequency of the frequency component, and the knocking determination is performed using the weighted time series data and the weighted master pattern data. 24. A computer program according to claim 20, wherein said step d) includes the step of i) calculating a fitting ratio of the binarized time series data with respect to the master pattern data, and it is determined that a knocking has occurred when the fitting ratio exceeds a determination threshold value. 25. A computer program according to claim 20, wherein the master pattern data are set according to an operating condition of said engine. 26. A computer program according to claim 19, wherein said knocking detecting method further comprises the step of j) generating frequency group time series data by selecting a maximum strength at one sampling timing with respect to each of frequency groups consisting of a plurality of frequencies, wherein the frequency group time series data are binarized in said step c). 27. A computer program according to claim 26, wherein said knocking detecting method further comprises the steps of: k) calculating time series data of group noise components corresponding to each of the frequency groups based on the frequency group time series data; andl) correcting the frequency group time series data with the time series data of group noise components,wherein the corrected frequency group time series data is binarized in said step c).