Methods and apparatuses for fuel consumption prediction

1. A method of fuel consumption prediction, the method comprising the acts of:

extracting via processing circuitry of the vehicle, from historic travel data for at least one historic travel path of a vehicle, a respective historic fuel consumption for each historic travel path;

extracting via the processing circuitry of the vehicle, from the historic travel data for each historic travel path, at least one historic fuel-related quantity influencing the respective historic fuel consumption;

adjusting via the processing circuitry of the vehicle, model parameters of a machine learning model for predicting the fuel consumption based on the extracted respective historic fuel consumption and the at least one historic fuel-related quantity for each historic travel path;

predicting via the processing circuitry of the vehicle, the fuel consumption for a planned trip using the machine learning model with the adjusted model parameters; and

providing a recommendation via an output device of the vehicle based on the predicted fuel consumption for the planned trip, wherein

adjusting the model parameters of the machine learning model further comprises:

determining a first parameter corresponding to a base fuel consumption rate of the vehicle based on the historic travel data, and

determining a second parameter corresponding to a speed at which the vehicle achieves its maximum range based on the historic travel data, and

the historic travel data is acquired by at least one of a navigation system of the vehicle, a sensor of the vehicle, a mobile device of a driver of the vehicle, or an electronic control unit of the vehicle.

2. The method of claim 1 , wherein

predicting the fuel consumption for the planned trip comprises extracting, from travel data associated with the planned trip, at least one current fuel-related quantity and feeding the machine learning model with the current fuel-related quantity.

3. The method of claim 1 , wherein

the provided recommendation is related to refueling the vehicle based on the predicted the fuel consumption.

4. The method of claim 1 , wherein

the provided recommendation is related to selecting a route and/or a time for the planned trip based on the predicted the fuel consumption.

5. The method of claim 1 , wherein

extracting the at least one fuel-related quantity comprises extracting at least one of a vehicle type, vehicle settings, trip length, travel speed, a number of stops, traffic density, fuel type, driver behavior, road condition, road topology, and weather condition for each historic travel path or the planned trip.

6. The method of claim 1 , wherein

extracting the at least one fuel-related quantity comprises extracting an average speed as fuel-related quantity for each historic travel path and wherein adjusting the model parameters comprises adjusting parameters of a Gaussian function based on the extracted average speed.

7. The method of claim 6 , wherein adjusting parameters of the Gaussian function is performed according to:

arg

⁢

⁢

min

cons

,

a

,

b

,

c

⁢

∑

i

⁢

(

fcr

i

-

(

cons

+

a

×

e

-

(

frf

i

·

AvgSpeed

-

b

)

⁢

2

c

2

)

)

2

,

wherein cons, a, b, c respectively denote the first, the second, a third, and a fourth model parameter, fcr i denotes an extracted historic fuel consumption for historic travel path i and frf i ·AvgSpeed denotes the extracted average speed for historic travel path i.

8. A fuel consumption prediction apparatus, the apparatus comprising:

processing circuitry configured to:

extract, from historic travel data for at least one historic travel path of a vehicle, a respective fuel consumption for each historic travel path;

extract, from the historic travel data for each historic travel path, at least one historic fuel-related quantity influencing the respective historic fuel consumption;

adjust model parameters of a machine learning model for predicting the fuel consumption based on the extracted respective historic fuel consumption and the at least one historic fuel-related quantity for each historic travel path;

predict the fuel consumption for a planned trip using the machine learning model with the adjusted model parameters; and

provide a recommendation via an output device of the vehicle based on the predicted fuel consumption for the planned trip, wherein

the processing circuitry is configured to adjust the model parameters of the machine learning model by:

determining a first parameter corresponding to a base fuel consumption rate of the vehicle based on the historic travel data, and

determining a second parameter corresponding to a speed at which the vehicle achieves its maximum range based on the historic travel data, and

the historic travel data is acquired by at least one of a navigation system of the vehicle, a sensor of the vehicle, a mobile device of a driver of the vehicle, or an electronic control unit of the vehicle.

9. The method of claim 1 , wherein adjusting the model parameters further comprises:

determining at least one of a third parameter or a fourth parameter such that an error component of the predicted fuel consumption for the planned trip is minimized, wherein

the error quantity is a function of the first, second, third, and fourth parameters.

10. The apparatus of claim 8 , wherein the processing circuitry is further configured to adjust the model parameters by:

determining at least one of a third parameter or a fourth parameter such that an error component of the predicted fuel consumption for the planned trip is minimized, wherein

the error quantity is a function of the first, second, third, and fourth parameters.

11. The method of claim 1 , wherein

the first parameter summed with the second parameter is less than or equal to a maximum fuel consumption rate of the vehicle.

12. The apparatus of claim 8 , wherein

the first parameter summed with the second parameter is less than or equal to a maximum fuel consumption rate of the vehicle.

13. The apparatus of claim 8 , wherein the processing circuitry is further configured to predict the fuel consumption for the planned trip by:

extracting, from travel data associated with the planned trip, at least one current fuel-related quantity and feeding the machine learning model with the current fuel-related quantity.

14. The apparatus of claim 8 , wherein

the provided recommendation is related to refueling the vehicle based on the predicted the fuel consumption.

15. The apparatus of claim 8 , wherein

the provided recommendation is related to selecting a route and/or a time for the planned trip based on the predicted the fuel consumption.

16. The apparatus of claim 8 , wherein the processing circuitry is further configured to extract the at least one fuel-related quantity by:

extracting at least one of a vehicle type, vehicle settings, trip length, travel speed, a number of stops, traffic density, fuel type, driver behavior, road condition, road topology, and weather condition for each historic travel path or the planned trip.

17. The apparatus of claim 8 , wherein the processing circuitry is further configured to extract the at least one fuel-related quantity by:

extracting an average speed as fuel-related quantity for each historic travel path and wherein adjusting the model parameters comprises adjusting parameters of a Gaussian function based on the extracted average speed.

18. The apparatus of claim 17 , wherein the processing circuit is further configured to adjust parameters of the Gaussian function according to:

arg

⁢

⁢

min

cons

,

a

,

b

,

c

⁢

∑

i

⁢

(

fcr

i

-

(

cons

+

a

×

e

-

(

frf

i

·

AvgSpeed

-

b

)

⁢

2

c

2

)

)

2

,

wherein cons, a, b, c respectively denote the first, the second, a third, and a fourth model parameter, fcr i denotes an extracted historic fuel consumption for historic travel path i and frf i ·AvgSpeed denotes the extracted average speed for historic travel path i.