Correlation based fuel tank leak detection

• US 9,857,266 B2
• Filed: 02/04/2014
• Issued: 01/02/2018
• Est. Priority Date: 02/04/2014
• Status: Active Grant

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Original Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Inventors: Tseng, Fling, Makki, Imad Hassan, Dudar, Aed M., Jentz, Robert Roy, Filev, Dimitar Petrov
• Primary Examiner(s): Rivera, Carlos A
• Assistant Examiner(s): Picon-Feliciano, Ruben
• Application Number:
  US14/172,108
• Publication Number:
  US 20150219522A1
• Time in Patent Office: 1,428 Days
• Field of Search: 701102-104, 701114, 123516, 123518-520, 123198 D, 123 19 D, 7311439, 7311443, 137587-589
• CPC Class Codes: F02D 2041/225 Leakage detection
• CPC Class Codes: F02D 2200/0414 Air temperature
• CPC Class Codes: F02D 2200/0602 Fuel pressure
• CPC Class Codes: F02D 2200/0606 Fuel temperature
• CPC Class Codes: F02D 2200/703 Atmospheric pressure
• 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: F02D 41/266 the computer being backed-up or assisted by another circuit, e.g. analogue
• CPC Class Codes: F02M 25/0809 Judging failure of purge control system
• CPC Class Codes: G01M 3/025 Details with respect to the testing of engines or engine parts
• CPC Class Codes: G01M 3/3272 for verifying the internal pressure of closed containers
• CPC Class Codes: Y02T 10/40 Engine management systems

Abstract

A method for an engine, comprising: sealing a fuel tank; and indicating fuel system degradation based on a comparison of a first correlation coefficient, determined based on a change in fuel tank pressure and a change in fuel tank temperature, to a second correlation coefficient determined based on a change in ambient pressure and a change in ambient temperature. In this way, leak tests, such as engine-off natural vacuum tests, may be performed without knowledge of the properties of the fuel stored in the fuel tank.

Claims

1. A method for an engine, comprising:

sealing a fuel tank by closing a fuel tank isolation valve; and

while the fuel tank is sealed and the engine is off,

indicating fuel system degradation based on a comparison of a first correlation coefficient, determined based on a change in fuel tank pressure and a change in fuel tank temperature, to a second correlation coefficient determined based on a change in ambient pressure and a change in ambient temperature;

the fuel tank pressure and the fuel tank temperature measured from a fuel tank pressure sensor and a fuel tank temperature sensor, respectively; and

the ambient pressure and ambient temperature measured from an ambient pressure sensor and an ambient temperature sensor, respectively,

wherein determining the first correlation coefficient includes determining if a fuel tank temperature entropy is greater than an entropy threshold;

wherein determining the first correlation coefficient further includes determining a first covariance between fuel tank temperature and fuel tank pressure based on recursive estimates of mean fuel tank temperature and mean fuel tank pressure over a duration, and determining recursive estimates of standard deviation of fuel tank temperature and standard deviation of fuel tank pressure over the duration; and

wherein determining the second correlation coefficient includes determining a second covariance between ambient temperature and ambient pressure based on recursive estimates of mean ambient temperature and mean ambient pressure over the duration, and determining recursive estimates of standard deviation of ambient temperature and standard deviation of ambient pressure over the duration.

2. The method of claim 1, further comprising:

indicating a fuel system leak if the first correlation coefficient is less than the second correlation coefficient; and

, after indication of the fuel system leak, uploading data to a cloud computing system coupled to a plurality of vehicles, the data including test data.

3. The method of claim 1, further comprising:

indicating an intact fuel system if the first correlation coefficient is greater than the second correlation coefficient; and

, after indication of the intact fuel system, uploading data to a cloud computing system, the data including test data;

wherein the ambient pressure and temperature sensors are coupled to a vehicle including the engine, and the fuel tank pressure and temperature sensors are coupled to the fuel tank.

4. The method of claim 1, further comprising:

aborting a fuel tank leak test if the first correlation is not greater than zero;

otherwise, determining a third correlation coefficient based on the change in fuel tank pressure and the change in ambient pressure; and

indicating a fuel system leak if the third correlation coefficient is greater than a correlation threshold;

wherein determining the third correlation coefficient includes determining a covariance between fuel tank pressure and ambient pressure based on recursive estimates of mean fuel tank pressure and mean ambient pressure over the duration, and determining recursive estimates of standard deviation of fuel tank pressure and standard deviation of ambient pressure over the duration.

5. The method of claim 1, where determining the first correlation coefficient includes retrieving a first correlation model based on one or more previous fuel system degradation tests.

6. The method of claim 5, where determining the second correlation coefficient includes retrieving a second correlation model based on the one or more previous fuel system degradation tests.

7. The method of claim 6, where the first and second correlation models are retrieved from a cloud computing system coupled to a plurality of vehicles.

20. The method of claim 1, wherein indicating fuel system degradation includes generating test results based on the comparison, the method further comprising uploading test result data to a cloud database for aggregation from multiple vehicles, and downloading updated leak test parameters based on the aggregated results.

8. A method for testing a vehicle fuel tank for leaks, comprising:

sealing the vehicle fuel tank by closing a canister vent valve and a canister purge valve;

while the fuel tank is sealed and an engine coupled to the fuel tank is off,

determining whether a fuel tank fill level is between a first fill level limit and a second fill level limit;

monitoring fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure for a duration based on indications from a fuel tank temperature sensor, a fuel tank pressure sensor, an ambient temperature sensor, and an ambient pressure sensor, respectively;

determining entropy levels for fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure over the duration;

determining recursive means and standard deviations for fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure over the duration;

determining a first correlation coefficient based on the recursive means and standard deviations for fuel tank pressure and fuel tank temperature over the duration;

determining a second correlation coefficient based on the recursive means and standard deviations for ambient pressure and ambient temperature over the duration; and

indicating a fuel tank leak if the first correlation coefficient is less than the second correlation coefficient.

9. The method of claim 8, further comprising:

after testing the vehicle fuel tank for leaks, uploading data to a cloud computing system coupled to a plurality of vehicles, the data including test data; and

further comprising;

if the first correlation coefficient is not greater than zero, aborting testing the fuel tank for leaks and setting a flag to test the fuel tank for leaks at a later time point;

if the first correlation coefficient is greater than zero, determining a third correlation coefficient based on the recursive means and standard deviations for fuel tank pressure and ambient pressure; and

indicating a fuel tank leak if the third correlation coefficient is greater than a correlation threshold.

10. The method of claim 8, where determining the first correlation coefficient includes determining if the entropy level for fuel tank temperature is greater than an entropy threshold.

11. The method of claim 10, where determining the first correlation coefficient includes retrieving a first correlation model based on one or more previous fuel system degradation tests.

12. The method of claim 11, where determining the second correlation coefficient includes retrieving a second correlation model based on the one or more previous fuel system degradation tests.

13. The method of claim 12, where the first and second correlation models are retrieved from a cloud computing system coupled to a plurality of vehicles.

14. The method of claim 8, further comprising:

determining whether a fuel sloshing amount is below a sloshing threshold prior to monitoring fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure for the duration.

15. A system for a vehicle, comprising:

a fuel tank temperature sensor coupled to a fuel tank;

a fuel tank pressure sensor coupled to the fuel tank;

a fuel level sensor coupled within the fuel tank;

an ambient temperature sensor;

an ambient pressure sensor;

one or more valves configured to seal the fuel tank when closed; and

a controller configured with instructions stored in non-transitory memory, that when executed cause the controller to;

seal the fuel tank;

while the fuel tank is sealed and the engine is off,

determine whether a fuel tank fill level is between a first fill level limit and a second fill level limit based on indications from the fuel level sensor;

monitor fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure for a duration based on indications from the fuel tank temperature sensor, the fuel tank pressure sensor, the ambient temperature sensor, and the ambient pressure sensor, respectively;

determine entropy levels for fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure over the duration;

determine recursive means and standard deviations for fuel tank temperature, fuel tank pressure, ambient temperature, and ambient pressure over the duration;

determine a first correlation coefficient based on the recursive means and standard deviations for fuel tank pressure and fuel tank temperature over the duration;

determine a second correlation coefficient based on the recursive means and standard deviations for ambient pressure and ambient temperature over the duration; and

indicate a fuel tank leak if the first correlation coefficient is less than the second correlation coefficient.

16. The system of claim 15, wherein the duration includes when the fuel tank temperature entropy level is greater than a threshold;

and where the controller is further configured with instructions stored in non-transitory memory, that when executed cause the controller to;

determine a third correlation coefficient based on the recursive means and standard deviations for fuel tank pressure and ambient pressure over the duration; and

indicate a fuel tank leak if the third correlation coefficient is greater than a correlation threshold;

record fuel tank leak testing data to the non-transitory memory of the controller; and

upload the fuel tank leak testing data to a cloud computing system coupled to a plurality of vehicles.

17. The system of claim 15, further comprising:

a communications module configured to communicate with a cloud computing system; and

where the controller is further configured with instructions stored in non-transitory memory, that when executed cause the controller to;

retrieve one or more correlation models from the cloud computing system via the communications module;

determine the first correlation coefficient based on the one or more correlation models;

determine the second correlation coefficient based on the one or more correlation models; and

abort testing the fuel tank for leaks and set a flag to test the fuel tank for leaks at a later time point if the first correlation coefficient is not greater than zero.

18. The system of claim 17, where the cloud computing system is coupled to two or more vehicles.

19. The system of claim 18, where the one or more correlation models are based on fuel tank leak test results from the two or more vehicles.