Abnormal combustion detection method for internal-combustion engines

• US 8,068,972 B2
• Filed: 09/28/2009
• Issued: 11/29/2011
• Est. Priority Date: 10/09/2008
• Status: Expired due to Fees

• Current Assignee: IFP Energies Nouvelles
• Original Assignee: IFP Energies Nouvelles
• Inventors: Pagot, Alexandre, Zaccardi, Jean-Marc, Auclair, Dominique, Thirouard, Benoist, Duval, Laurent
• Primary Examiner(s): HUYNH, HAI H
• Application Number:
  US12/567,934
• Publication Number:
  US 20100094528A1
• Time in Patent Office: 792 Days
• Field of Search: 701/111, 701/106, 701/107, 701/113, 123/435, 123/406.13, 123/406.27, 123/406.37, 123/479, 123/480, 73/353, 73/356, 73/357, 73/351.2, 731/140.2
• US Class Current: 701/107
• CPC Class Codes: F02D 19/12 peculiar to engines working with non-fuel substances or with anti-knock agents, e.g. with anti-knock fuel
• CPC Class Codes: F02D 2041/286 comprising means for signal processing
• CPC Class Codes: F02D 35/023 by determining the cylinder pressure
• CPC Class Codes: F02D 35/027 using knock sensors
• CPC Class Codes: F02D 35/028 by determining the combustion timing or phasing
• CPC Class Codes: F02D 41/0002 Controlling intake air
• CPC Class Codes: F02D 41/0025 Controlling engines characterised by use of non-liquid fuels, pluralities of fuels, or non-fuel substances added to the combustible mixtures
• CPC Class Codes: F02D 41/402 Multiple injections
• CPC Class Codes: F02M 26/20 Feeding recirculated exhaust gases directly into the combustion chambers or into the intake runners
• CPC Class Codes: G01M 15/08 by monitoring pressure in cylinders

Abstract

The invention is an abnormal combustion detection method for spark-ignition internal-combustion engines. For each engine cycle, a parameter characterizing a distribution of N combustion indicator values, CA10 for example, acquired over N cycles preceding the cycle in progress, is determined while ignoring extreme values. A threshold is defined from this parameter for the combustion indicator. The start of an abnormal combustion is then detected by comparing the combustion indicator with this threshold and the course of the abnormal combustion detected in the combustion chamber is controlled.

Claims

1. A method for controlling combustion of a spark-ignition internal-combustion engine, wherein an abnormal combustion is detected in a combustion chamber of at least one cylinder of the engine, a signal representative of a state of the combustion is recorded by a detector disposed in the engine and at least one indicator of the combustion at each engine cycle is determined from the signal for each engine cycle comprising:

determining at least one parameter characterizing a distribution of a number of combustion indicator values acquired over the number of engine cycles preceding an engine cycle in progress either with an estimator or a method of modeling the distribution, by ignoring, at least some of the indicator values which are obtained during pre-ignition;

defining at least one threshold for the at least one indicator by parameters characterizing the distribution;

detecting a start of an abnormal combustion by comparing the at least one indicator of the combustion with the threshold; and

controlling, in response to detecting the start of abnormal combustion, a course of the detected abnormal combustion in the at least one combustion chamber.

2. A method as claimed in claim 1, wherein the signal is a continuous pressure measurement within the cylinder.

3. A method as claimed in claim 2, wherein the at least one indicator is the crank angle at which part of the energy introduced has been released.

4. A method as claimed in claim 3, wherein the estimator is selected from a trimmed mean, a median and an estimator m resulting from a minimization of the following function f:

f(V,m)=(|V(1)−

m|)a+ . . . +(|V(N)−

m|)a,

wherein a is a real number ranging between 0 and 2, and V(1) . . . V(n) are the N combustion indicator values.

5. A method as claimed in claim 3, wherein the method of modelling the distribution comprises:

a) removing indicator values from the number of values of a distribution to be modeled;

b) modelling the distribution of remaining indicator values by selecting an a priori experimental distribution law;

c) calculating a modelling quality criterion; and

d) if the criterion indicates an insufficient modelling quality, removing other indicator values from remaining indicator values defined by a difference of the number of indicator values and a number of removed indicator values and repeating b).

6. A method as claimed in claim 5, wherein a criterion, used for modeling the distribution is a sum of rms deviations between the a priori experimental distribution and the modelled distribution, or to a sum of absolute deviations between the a priori experimental distribution and the modelled distribution.

7. A method as claimed in claim 3, wherein the at least one threshold is determined by constraints defined from the at least one threshold value obtained from testing and/or from threshold values obtained in previous engine cycles.

8. A method as claimed in claim 3, wherein combustion indicators are determined and the at least one threshold is also determined for each combustion indicator.

9. A method as claimed in claim 2, wherein the estimator is selected from a trimmed mean, a median and an estimator m resulting from a minimization of the following function f:

f(V,m)=(|V(1)−

m|)a+ . . . +(|V(N)−

m|)a,

wherein a is a real number ranging between 0 and 2, and V(1) . . . V(n) are the N combustion indicator values.

10. A method as claimed in claim 2, wherein the method of modelling the distribution comprises:

a) removing indicator values from the number of values of a distribution to be modeled;

b) modelling the distribution of remaining indicator values by selecting an a priori experimental distribution law;

c) calculating a modelling quality criterion; and

d) if the criterion indicates an insufficient modelling quality, removing other indicator values from remaining indicator values defined by a difference of the number of indicator values and a number of removed indicator values and repeating b).

11. A method as claimed in claim 10, wherein a criterion, used for modeling the distribution is a sum of rms deviations between the a priori experimental distribution and the modelled distribution, or to a sum of absolute deviations between the a priori experimental distribution and the modelled distribution.

12. A method as claimed in claim 2, wherein the at least one threshold is determined by constraints defined from the at least one threshold value obtained from testing and/or from threshold values obtained in previous engine cycles.

13. A method as claimed in claim 2, wherein combustion indicators are determined and the at least one threshold is also determined for each combustion indicator.

14. A method in accordance with claim 2 wherein the part of the energy is 10%.

15. A method as claimed in claim 1, wherein the at least one indicator is the crank angle at which part of the energy introduced has been released.

16. A method as claimed in claim 15, wherein the estimator is selected from a trimmed mean, a median and an estimator m resulting from a minimization of the following function f:

f(V,m)=(|V(1)−

m|)a+ . . . +(|V(N)−

m|)a,

wherein a is a real number ranging between 0 and 2, and V(1) . . . V(n) are the N combustion indicator values.

17. A method as claimed in claim 15, wherein the method of modelling the distribution comprises:

a) removing indicator values from the number of values of a distribution to be modeled;

b) modelling the distribution of remaining indicator values by selecting an a priori experimental distribution law;

c) calculating a modelling quality criterion; and

d) if the criterion indicates an insufficient modelling quality, removing other indicator values from remaining indicator values defined by a difference of the number of indicator values and a number of removed indicator values and repeating b).

18. A method as claimed in claim 17, wherein a criterion, used for modeling the distribution is a sum of rms deviations between the a priori experimental distribution and the modelled distribution, or to a sum of absolute deviations between the a priori experimental distribution and the modelled distribution.

19. A method in accordance with claim 15 wherein the part of the energy is 10%.

20. A method as claimed in claim 15, wherein the at least one threshold is determined by constraints defined from the at least one threshold value obtained from testing and/or from threshold values obtained in previous engine cycles.

21. A method as claimed in claim 15, wherein combustion indicators are determined and the at least one threshold is also determined for each combustion indicator.

22. A method in accordance with claim 15 wherein the part of the energy is 10%.

23. A method in accordance with claim 15 wherein the part of the energy is 10%.

24. A method as claimed in claim 1, wherein the estimator is selected from a trimmed mean, a median and an estimator m resulting from a minimization of the following function f:

f(V,m)=(|V(1)−

m|)a+ . . . +(|V(N)−

m|)a,

wherein a is a real number ranging between 0 and 2, and V(1) . . . V(n) are the N combustion indicator values.

25. A method as claimed in claim 1, wherein the method of modelling the distribution comprises:

a) removing indicator values from the number of values of a distribution to be modeled;

b) modelling the distribution of remaining indicator values by selecting an a priori experimental distribution law;

c) calculating a modelling quality criterion; and

d) if the criterion indicates an insufficient modelling quality, removing other indicator values from remaining indicator values defined by a difference of the number of indicator values and a number of removed indicator values and repeating b).

26. A method as claimed in claim 25, wherein a criterion, used for modeling the distribution is a sum of rms deviations between the a priori experimental distribution and the modelled distribution, or to a sum of absolute deviations between the a priori experimental distribution and the modelled distribution.

27. A method as claimed in claim 1, wherein the at least one threshold is determined by defining a tolerance for the parameter characterizing the distribution.

28. A method as claimed in claim 1, wherein the at least one threshold corresponds to a sum of a trimmed mean of the number of indicator values and of k times a standard deviation of the number of values, wherein the real number is a positive or negative real number.

29. A method as claimed in claim 1, wherein the at least one threshold is determined by constraints defined from the at least one threshold value obtained from testing and/or from threshold values obtained in previous engine cycles.

30. A method as claimed in claim 1, wherein combustion indicators are determined and the at least one threshold is also determined for each combustion indicator.