Method for determining fuel volatility and consequently performing cold starting of an internal combustion engine

• US 7,904,234 B2
• Filed: 12/21/2005
• Issued: 03/08/2011
• Est. Priority Date: 12/23/2004
• Status: Expired due to Fees

• Current Assignee: Magneti Marelli Powertrain S.p.A. (Magneti Marelli S.p.A.)
• Original Assignee: Magneti Marelli Powertrain S.p.A. (Magneti Marelli S.p.A.)
• Inventors: Ruggiano, Renzo, Farina, Raffaele, Zambelli, Anna, Mazet, Henri
• Primary Examiner(s): Cronin; Stephen K
• Assistant Examiner(s): Bacon; Anthony L
• Application Number:
  US11/794,066
• Publication Number:
  US 20090025689A1
• Time in Patent Office: 1,903 Days
• Field of Search: 123/491, 123/492, 123/478, 123/486, 123/179.15, 123/179.16, 123/179.17, 701/103, 701/104, 701/113, 701/115
• US Class Current: 701/113
• CPC Class Codes: F02D 19/0631 by estimation, i.e. without using direct measurements of a corresponding sensor
• CPC Class Codes: F02D 19/0636 by estimation, i.e. without using direct measurements of a corresponding sensor
• CPC Class Codes: F02D 19/0649 Liquid fuels having different boiling temperatures, volatilities, densities, viscosities, cetane or octane numbers
• CPC Class Codes: F02D 2200/0612 determined by estimation
• CPC Class Codes: F02D 2200/1012 Engine speed gradient
• 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 41/047 Taking into account fuel evaporation or wall wetting; special correction after fuel cut-off F02D41/126
• CPC Class Codes: F02D 41/062 for starting F02D41/061 takes precedence
• CPC Class Codes: F02D 41/1497 With detection of the mechanical response of the engine
• CPC Class Codes: F02D 41/22 Safety or indicating devices for abnormal conditions in air/fuel ratio feedback systems F02D41/1495, in electric control linkage F02D11/107, in purge control systems F02M25/0809
• CPC Class Codes: Y02T 10/30 Use of alternative fuels, e.g. biofuels

Abstract

A method for determining fuel volatility and consequently performing cold starting of an internal combustion engine; in the event of cold starting, the method provides for the determination of an enrichment percentage as a function of a stored value for fuel volatility; the determination of a predicted value for starting quality before performing starting; the starting of the engine using the previously determined enrichment percentage; the determination of a measured value for starting quality; the determination of a correction value of the stored value for fuel volatility as a function of the comparison between the measured value for starting quality and the predicted value for starting quality; and the updating of the stored value for fuel volatility by applying the correction value to said stored value, so consequently modifying the amount of enrichment in force.

Claims

1. A method for determining fuel volatility and consequently performing cold starting of an internal combustion engine, the method comprises the steps of:

determining, before starting the engine, an enrichment percentage (% Enrich) as a function of a stored value (Vmem) for fuel volatility and as a function of the temperature (TH2O) of a liquid coolant of the engine;

starting the engine using the previously determined enrichment percentage (% Enrich);

determining a predicted value (MarkPred) for starting quality before performing starting as a function of the stored value (Vmem) for fuel volatility and as a function of the temperature (TH2O) of the liquid coolant of the engine on starting;

determining a measured value (MarkMeas) for starting quality during an initial rise in the speed (RPM) of the engine and at the beginning of the third TDC counting from the one at which the rise in the speed (RPM) of the engine occurs;

determining a correction value (Vcorr) of the stored value (Vmem) for fuel volatility as a function of the comparison between the measured value (MarkMeas) for starting quality and the predicted value (MarkPred) for starting quality; and

updating the stored value (Vmem) for fuel volatility by applying the correction value (Vcorr) to said stored value (Vmem);

wherein, if cold starting of the engine is completed satisfactorily, the measured value (MarkMeas) for starting quality is determined as a function of a delay with which the rise in the speed (RPM) of the engine occurred and as a function of a rate of rise in the speed (RPM) of the engine during an initial rise in the speed (RPM) of the engine calculated at the beginning of the third TDC counting from the one at which the rise in the speed (RPM) of the engine occurs.

2. The method according to claim 1, wherein, once the stored value (Vmem) for fuel volatility has been updated by applying the correction value (Vcorr) to said stored value (Vmem), further comprising the step of updating the currently used enrichment percentage (% Enrich) as a function of the new stored value (Vmem) for fuel volatility.

3. The method according to claim 1, wherein the stored value (Vmem) for fuel volatility is increased if the predicted value (MarkPred) for starting quality is worse than the measured value (MarkMeas) for starting quality, and the stored value (Vmem) for fuel volatility is reduced if the predicted value (MarkPred) for starting quality is better than the measured value (MarkMeas) for starting quality.

4. The method according to claim 1, wherein the rise in engine speed is identified by detecting the presence of a first effective explosion and by detecting a difference between the value of the engine speed (rpm) evaluated after a given number of TDCs counting from the one at which the first effective explosion occurred and the value of the engine speed (rpm) preceding said first effective explosion.

5. The method according to claim 4, wherein the rise in engine speed is identified only if the difference between the value of the engine speed (rpm) evaluated after a given number of TDCs counting from the one at which the first effective explosion occurred and the value of the engine speed (rpm) preceding said first effective explosion is greater than a predetermined threshold value.

6. The method according to claim 1, wherein the rate of rise in the speed (RPM) of the engine is determined at the beginning of the third TDC counting from the one at which the rise in the speed (RPM) of the engine occurs as the difference between the current value of engine speed (rpm) and the value of the engine speed (rpm) preceding the TDC where said rise occurs.

7. The method according to claim 1, wherein the delay with which the rise in the speed (RPM) of the engine occurred is determined by measuring the interval that elapses between the beginning of starting and the moment at which a significant rate of rise in the speed (RPM) of the engine is detected.

8. The method according to claim 1, wherein the delay with which the rise in the speed (RPM) of the engine occurred is determined by measuring the interval that elapses between the beginning of starting and the moment at which the speed (RPM) of the engine exceeds a predetermined threshold value.

9. The method according to claim 1, wherein the measured value (MarkMeas) for starting quality is estimated to be equal to a constant, positive initial value from which is subtracted a first positive adjustment value that is a function of the delay with which the rise in the speed (RPM) of the engine occurred and to which is algebraically summed a second adjustment value, which is a function of the rate of rise in the speed (RPM) of the engine.

10. The method according to claim 9, wherein the first adjustment value is obtained by subtracting a constant penalty value for each TDC that elapses between the beginning of starting the engine and the moment at which a sufficiently significant rate of rise in the speed (RPM) of the engine is detected.

11. The method according to claim 9, wherein the first adjustment value is obtained by subtracting a constant penalty value for each TDC that elapses between the beginning of starting of the engine and the moment at which the speed (RPM) of the engine exceeds a predetermined threshold value.

12. The method according to claim 10, wherein the initial value is equal to 9 and the constant used for calculating the first adjustment value is equal to 0.38.

13. The method according to claim 9, wherein the second adjustment value is selected from a range by comparing the rate of rise in the speed (RPM) of the engine with a reference rate of rise.

14. The method according to claim 13, wherein the range used for calculating the second adjustment value is between +1 and −

2.75.

15. The method according to claim 1, wherein the measured value (MarkMeas) for starting quality is estimated to be equal to a constant, positive initial value from which a first positive adjustment value is subtracted if an appreciable delay in the rise in the speed (RPM) of the engine has been detected, or to which a second adjustment value is algebraically summed if no appreciable delay in the rise in the speed (RPM) of the engine has been detected.

16. The method according to claim 15, wherein the first adjustment value is applied if no significant rise in engine speed is detected after a given number of explosions from starting;

the first adjustment value is increased by a constant penalty value for each TDC that elapses between the moment at which starting of the engine begins and the moment at which the speed (RPM) of the engine exceeds a predetermined threshold value.

17. The method according to claim 16, wherein the initial value of the measured value (MarkMeas) for starting quality is equal to 9 and the constant penalty value of the first adjustment value is equal to 0.38 for each TDC.

18. The method according to claim 15, wherein the second adjustment value is a function of the rate of rise in the speed (RPM) of the engine and is selected from a range by comparing the rate of rise in the speed (RPM) of the engine with a reference rate of rise.

19. The method according to claim 18, wherein the rate of rise in the speed (RPM) of the engine is determined at the beginning of the third TDC, at the first TDC subsequent to the third explosion in absolute terms from starting and identified as the difference in revolutions between the value obtained at the third explosion and the most recent driven value preceding the first explosion of the engine in absolute terms, if greater than a predetermined threshold value.

20. The method according to claim 9, wherein the initial value is equal to 9 and the range used for calculating the second adjustment value is between +1 and −

2.75.

21. The method according to claim 9, wherein the measured value (MarkMeas) for starting quality is saturated in order always to be within a predefined range.

22. The method according to claim 21, wherein the predefined range of the measured value (MarkMeas) for starting quality is between 0 and 10.

23. The method according to claim 1, wherein, if cold starting of the engine is not completed satisfactorily, i.e. the engine stops while an evaluation is under way, the measured value (MarkMeas) for starting quality is set to the minimum full scale value.

24. The method according to claim 1, wherein the enrichment percentage (% Enrich) is variable over time between an initial maximum value in absolute terms and a final, zero value.

25. The method according to claim 24, wherein the enrichment percentage (% Enrich) is implemented by means of a first contribution, which acts on the objective mixture concentration and a second residual correction contribution, which acts directly on the objective amount of petrol.

26. The method according to claim 25, wherein each contribution falls in value over time in accordance with an exponential decay law.

27. The method according to claim 25, wherein a first decay time constant of the first concentration contribution is less than a second decay time constant of the second residual correction contribution.

28. The method according to claim 25, wherein the first time constant and the second time constant are calculated as a function of the stored value (Vmem) for fuel volatility and of the initial value of a temperature (TH2O) of a liquid coolant of the internal combustion engine.

29. The method according to claim 28, wherein the first decay time constant of the first concentration contribution is approximately one third of the second decay time constant of the second residual correction contribution.

30. The method according to claim 24, wherein, following updating of the stored value (Vmem) for fuel volatility, the enrichment under way is modified accordingly, either by updating the amount of the enrichment as a function of the new stored value (Vmem) for fuel volatility or by updating the dynamics of the extinction profile of said enrichment.

31. The method according to claim 30, wherein the amount of the enrichment is updated on the basis of the value of the ratio between the new enrichment percentage (% Enrich) determined on the basis of the new stored value (Vmem) for fuel volatility and the enrichment percentage (% Enrich) used on initial starting.

32. The method according to claim 30, wherein the dynamics of the extinction profile of said enrichment are updated by updating the values of the time constants as a function of the new stored value (Vmem) for fuel volatility.

33. The method according to claim 1, wherein the start is deemed significant for the purposes of updating the stored value (Vmem) for fuel volatility only if at least a predetermined time interval has elapsed since the internal combustion engine was last turned off.

34. The method according to claim 33, wherein the time interval is determined in such a manner as to ensure that all the components of the engine reach a temperature equivalent to the outside temperature.

35. The method according to claim 1, wherein the stored value (Vmem) for fuel volatility is updated, using the correction value Vcorr, only if the initial temperature (TH2O) of a liquid coolant for the engine before starting is within a given temperature range.

36. The method according to claim 35, wherein the temperature range has a lower limit, which ensures that each component of the starting system is perfectly functional and behaves repeatably (lower limit), and an upper limit, which ensures that fuel volatility has an influence on starting quality.

37. The method according to claim 36, wherein the stored value (Vmem) for fuel volatility is updated, using the correction value Vcorr, only if the initial temperature (TH2O) of a liquid coolant of the engine before starting is between −

25°

C. and 10°

C.

38. The method according to claim 1, wherein the stored value (Vmem) for fuel volatility is updated, using the correction value Vcorr, only if atmospheric pressure is greater than a predetermined threshold value.

39. The method according to claim 1, wherein the stored value (Vmem) for fuel volatility is not updated, using the correction value Vcorr, if there are engine malfunction signals present that may have an impact on starting quality.

40. The method according to claim 39, wherein, if engine malfunction signals identified by a standard diagnostic system are present that may have an impact on starting quality, then the enrichment percentage (% Enrich) is determined as a function of a stored reference value (Vref) for fuel volatility that is dependent on the values used in previous starts.

41. The method according to claim 1, wherein the stored value (Vmem) for fuel volatility is not updated if any malfunction condition not detectable by the standard diagnostic system and that may have an impact on starting quality is determined.

42. The method according to claim 41, wherein, if a malfunction condition that may have an impact on starting quality is determined, then the enrichment percentage (% Enrich) is determined as a function of a predetermined recovery value (V_recovery) for fuel volatility, which recovery value (V_recovery) makes it possible to start the car within an acceptable number of attempts, irrespective of the volatility value of the fuel actually present in the tank.

43. The method according to claim 1, wherein, if the mean of the most recent measured values (MarkMeas) for starting quality is below a given threshold value, an engine malfunction condition is determined.

44. The method according to claim 1, wherein, if the mean of the differences between the most recent predicted values (MarkPred) for starting quality and the corresponding measured values (MarkMeas) for starting quality is greater than a given threshold value, an engine malfunction condition is determined.

45. The method according to claim 1, wherein an error index is calculated, which is incremented by an amount depending on the number of TDCs of each failed start that resulted in the engine coming to a stop;

and if the value of the error index is greater than a predetermined threshold value an engine malfunction condition is determined.

46. The method according to claim 45, wherein the error index is set to zero after a predetermined number of consecutive and unproblematic cold starts.

47. The method according to claim 45, wherein the error index is set to zero when an unproblematic start/restart is detected that occurs within a given maximum number of attempts.

48. The method according to claim 1, wherein, if after a predetermined number of TDCs from the beginning of starting, the engine has not started, then the currently used enrichment percentage (% Enrich) is incremented.

49. The method according to claim 1, wherein a cold start of the engine is judged problematic if the difference between the predicted value (MarkPred) for starting quality and a current estimate of the measured value (MarkMeas) for starting quality is greater than a predetermined threshold value.

50. The method according to claim 1, wherein a cold start of the engine is judged problematic if the number of TDCs characterising the step that precedes the onset of established running of the engine exceeds a predetermined threshold value.

51. The method according to claim 49, wherein, once a problematic starting situation has been determined, then emergency action is implemented in which the currently used enrichment percentage (% Enrich) is updated as a function of an emergency value (Vemergency) for fuel volatility, which is equal to the fuel volatility value used at the beginning of starting decremented by a predetermined decrement value (DeltaVemergency).

52. The method according to claim 51, wherein the decrement value (DeltaVemergency) is such as to cause the emergency value (Vemergency) for fuel volatility to take a value close to the minimum possible value.

53. The method according to claim 1, wherein a problematic cold restart condition is determined when the engine is not yet warmed up, the temperature (TH2O) of a liquid coolant is below an upper limit that permits identification of fuel volatility, and the difference between the predicted value (MarkPred) for starting quality and a current estimate of the measured value (MarkMeas) for starting quality is greater than a predetermined threshold value;

in the case of a problematic cold restart condition, if the stored value (Vmem) for fuel volatility is high enough to be considered potentially problematic due to the consequent thinning because it is greater than the value that permits starting within an acceptable number of attempts irrespective of the volatility of fuel actually present in the tank, if the presence of system malfunctions not signalled by the standard diagnostics and responsible for the problematic starting can be excluded with sufficient certainty, and if the previous start was also problematic, then emergency action is taken in which the currently used enrichment percentage (% Enrich) is determined as a function of an emergency value (Vemergency) for fuel volatility, which is equal to the fuel volatility value used at the beginning of starting decremented by a predetermined decrement value (DeltaVemergency).

54. The method according to claim 53, wherein the decrement value (DeltaVemergency) is such as to cause the emergency value (Vemergency) for fuel volatility to take a value close to the minimum possible value.

55. The method according to claim 1, wherein a critical start/restart condition is determined when a start/restart occurs under any condition that is not perfectly covered by a case history of experimental data appropriate for defining the behaviour of the system.

56. The method according to claim 1, wherein, in the case of critical starting/restarting, if the stored value (Vmem) for fuel volatility is high enough to be considered potentially problematic due to the consequent thinning because it is greater than the value that permits starting within an acceptable number of attempts irrespective of the volatility of fuel actually present in the tank, and if the presence of system malfunctions not signalled by the standard diagnostics and responsible for the problematic starting can be excluded with some degree of certainty, then emergency action is taken in which the currently used enrichment percentage (% Enrich) is determined as a function of an emergency value (Vemergency) for fuel volatility, which is equal to the fuel volatility value used at the beginning of starting decremented by a predetermined decrement value (DeltaVemergency).

57. The method according to claim 56, wherein the decrement value (DeltaVemergency) is such as to cause the emergency value (Vemergency) for fuel volatility to take a value close to the minimum possible value.

58. The method according to claim 1, wherein the correction value (Vcorr) is obtained by multiplying the difference between the measured value (MarkMeas) for starting quality and the predicted value (MarkPred) for starting quality by a multiplicative constant.

59. The method according to claim 58, wherein the multiplicative constant takes two different values depending on whether the difference between the measured value (MarkMeas) for starting quality and the predicted value (MarkPred) for starting quality is positive or negative.

60. The method according to claim 58, wherein the correction value (Vcorr) is determined only if the number of successive TDCs that have occurred during the starting attempt is greater than a predetermined threshold value.

61. The method according to claim 60, wherein the correction value (Vcorr) is determined only if the number of successive TDCs that have occurred during the starting attempt is greater than 4.

62. The method according to claim 58, wherein, if the difference between the measured value (MarkMeas) for starting quality and the predicted value (MarkPred) for starting quality is in absolute terms below a given threshold value, then a value of zero is assigned to the correction value (Vcorr).

63. The method according to claim 1, wherein the stored value (Vmem) for fuel volatility is updated as a function of the correction value (Vcorr) by using a confidence automaton.

64. The method according to claim 63, wherein the stored value (Vmem) for fuel volatility is deemed to be confirmed if the difference between the measured value (MarkMeas) for starting quality and the predicted value (MarkPred) for starting quality is in absolute terms below a predetermined threshold value.

65. The method according to claim 64, wherein, when the stored value (Vmem) for fuel volatility is confirmed, then a reference value (Vref) for fuel volatility is assumed to be equal to the stored value (Vmem) for fuel volatility;

and the confidence automaton is based on the number of times for which the reference volatility value (Vref) has previously been confirmed consecutively.

66. The method according to claim 65 in which, for each confirmation, the confidence value is increased by a quantity that is dependent on the temperature (TH2O) of the liquid coolant on starting.

67. The method according to claim 65, wherein, when a certain stored value (Vmem) for fuel volatility is not confirmed, then the confidence value of the reference value (Vref) is reduced.

68. The method according to claim 64, wherein the new stored value (Vmem) for fuel volatility is calculated via a weighted mean between the reference value (Vref) for fuel volatility and the quantity obtained as a sum of the stored value (Vmem) for fuel volatility on initial starting with the correction value (Vcorr) that has just been calculated;

the weightings of the weighted mean are a function of the number of times for which the reference value (Vref) for fuel volatility has been confirmed in previous cold starts of the engine.

69. The method according to claim 64, wherein the number of times for which a certain stored value (Vmem) for fuel volatility has been confirmed is stored.

70. The method according to claim 69, wherein the number of times for which a certain stored value (Vmem) for fuel volatility has been confirmed is stored and the date of each confirmation is stored.

71. The method according to claim 70, wherein the oldest confirmations are eliminated.

72. The method according to claim 71, wherein the new stored value (Vmem) for fuel volatility is calculated via a weighted mean between the previously stored confirmed reference values for fuel volatility and the quantity obtained as a sum of the stored value (Vmem) for fuel volatility on initial starting with the correction value (Vcorr) that has just been calculated;

the weightings are a function of the number of times for which each reference value (Vref) for fuel volatility has been confirmed in previous cold starts of the engine and of the date on which said confirmations occurred, the more recent being preferred.

73. The method according to claim 63, wherein, in the case of emergency action, with the sole exclusion of critical starting/restarting for the purposes of repeatability of the evaluation thereof, the stored value (Vmem) for fuel volatility is calculated by means of a weighted mean between the reference value (Vref) and an emergency value (Vemergency) for fuel volatility, which is equal to the fuel volatility value used at the beginning of starting decremented by a predetermined decrement value (DeltaVemergency);

the weighting associated with the reference value (Vref) is equal to the confidence in said reference value (Vref).

74. The method according to claim 73, wherein the decrement value (DeltaVemergency) is such as to cause the emergency value (Vemergency) for fuel volatility to take a value close to the minimum possible value.

75. The method according to claim 63, wherein information regarding the level of fuel in a tank constitutes a further condition for enabling an emergency restart action and for reducing/zeroing the confidence in the reference volatility value (Vref).

76. The method according to claim 1, wherein a signal from a lambda probe (11) arranged in an exhaust system (10) of the engine is used to verify the correctness of the previously determined correction value (Vcorr).

77. The method according to claim 1, wherein the parameters used for compensation of the fluid film phenomenon are corrected as a function of the stored value (Vmem) for fuel volatility.