System and method of calculating distance to empty of vehicle
View Patent ↗In a method and system for determining a DTE of a vehicle, real-time transportation situation information from a current position of the vehicle to a destination is transmitted to a vehicle. The transmitted transportation situation information and obtaining vehicle forward driving situation information to the destination are received. Predicted battery consumption energy consumed for traveling from the current position to the destination using a battery usage prediction model to which the obtained vehicle forward driving situation information and current vehicle state information obtained from the vehicle are input are predicted and determined. A DTE at the destination as a predicted value at the current position is determined using the determined predicted battery consumption energy to the destination.
1. A method of determining a distance to empty (DTE) of a vehicle, the method comprising:
in a state in which a travel route to a destination is set, transmitting real-time transportation situation information from a current position of the vehicle to the destination to the vehicle by a vehicle external system;
receiving the transmitted transportation situation information and obtaining vehicle forward driving situation information to the destination, on which a real-time transportation situation is reflected, by a navigation terminal or a controller provided in the vehicle;
predicting and determining, by the controller, predicted battery consumption energy consumed for traveling from the current position to the destination using a battery usage prediction model to which the obtained vehicle forward driving situation information and current vehicle state information obtained from the vehicle are input; and
determining, by the controller, a DTE at the destination as a predicted value at the current position using the determined predicted battery consumption energy to the destination,
wherein the current vehicle state information includes a residual battery capacity.
2. The method of claim 1 , wherein the current vehicle state information further includes an electric load except for energy consumed for traveling of the vehicle, and an ambient temperature detected by a sensor.
3. The method of claim 1 ,
wherein the real-time transportation situation information to the destination includes road link-specific distances and average driving times of the vehicle on a travel route from the current position to the destination of the set travel route, and
wherein the vehicle forward driving situation information to the destination includes a speed and an acceleration of the vehicle on the travel route to the destination, determined from the road link-specific distances and the average driving times.
4. The method of claim 1 ,
wherein the vehicle forward driving situation information to the destination includes a speed and an acceleration of the vehicle on a travel route from the current position to the destination, and
wherein the navigation terminal or the controller of the vehicle is configured to obtain the speed and the acceleration of the vehicle on the travel route from the current position to the destination as time-series information and to provide the speed and the acceleration of the vehicle as inputs to the battery usage prediction model.
5. The method of claim 4 , wherein the speed and the acceleration are combined as values of a same time window, and the current vehicle state information obtained at the same time as the speed and the acceleration is processed by being allocated to the same time window as the speed and the acceleration.
6. The method of claim 5 , wherein the current vehicle state information further includes an electric load except for energy consumed for traveling of the vehicle, and an ambient temperature detected by a sensor.
7. The method of claim 5 , wherein the battery usage prediction model is a machine learning-based prediction model.
8. The method of claim 7 , wherein the battery usage prediction model is configured to:
extract feature information of the speed and the acceleration by processing time-series data of the speed and the acceleration using a first neural network; and
output the predicted battery consumption energy by processing the extracted feature information and the current vehicle state information as inputs using a second neural network.
9. The method of claim 7 , wherein the battery usage prediction model is configured to:
extract feature information of the speed and the acceleration while moving and updating time-series data of the speed and the acceleration at a predetermined time shift using a long short-term memory of a first neural network; and
output the predicted battery consumption energy by processing the extracted feature information and the current vehicle state information as inputs using a deep learning neural network as a second neural network.
10. A method of determining a distance to empty (DTE) of a vehicle, the method comprising:
in a state in which a travel route to a destination is set, transmitting real-time transportation situation information from a current position of the vehicle to the destination to the vehicle by a vehicle external system;
receiving the transmitted transportation situation information and obtaining vehicle forward driving situation information to the destination, on which a real-time transportation situation is reflected, by a navigation terminal or a controller provided in the vehicle;
predicting and determining, by the controller, predicted battery consumption energy consumed for traveling from the current position to the destination using a battery usage prediction model to which the obtained vehicle forward driving situation information and current vehicle state information obtained from the vehicle are input; and
determining, by the controller, a DTE at the destination as a predicted value at the current position using the determined predicted battery consumption energy to the destination,
wherein the determination of the DTE at the destination includes:
determining a SOC-based maximum DTE from an initial state of charge (SOC) value of a battery of the vehicle after charging of the battery and a maximum DTE at full charging of the battery;
determining a DTE reduced up to the present from SOC-based battery consumption energy used up to the present after the charging of the battery and energy efficiency information up to present;
determining a DTE to be reduced from the current position to the destination using the predicted battery consumption energy to the destination; and
determining the DTE at the destination to be a value obtained by subtracting the DTE reduced up to the present and the DET to be reduced from the current position to the destination from the determined initial SOC-based maximum DTE.
11. The method of claim 10 , wherein, in the determination of the DTE to be reduced from the current position to the destination,
the DTE to be reduced from the current position to the destination is determined to be a value obtained by multiplying SOC-based predicted consumption energy to be used from the current position to the destination and a predicted energy efficiency value to the destination,
the SOC-based predicted consumption energy to be used from the current position to the destination is determined to be a value obtained by multiplying a battery SOC value predicted to be reduced from the current position to the destination and maximum available battery energy at the full charging of the battery, and
the battery SOC value predicted to be reduced from the current position to the destination is obtained using the predicted battery consumption energy determined by the battery usage prediction model.
12. The method of claim 11 ,
wherein the battery SOC value predicted to be reduced from the current position to the destination is a value obtained by subtracting a finally-predicted battery SOC at the destination from a current battery SOC, and
wherein the finally-predicted battery SOC at the destination is determined to be a value obtained by dividing the predicted battery consumption energy determined by the battery usage prediction model with the maximum available battery energy at the full charging of the battery.
13. The method of claim 11 , wherein the predicted energy efficiency value to the destination is determined to be a value obtained by dividing a remaining travel distance from the current position to the destination with the predicted battery consumption energy determined by the battery usage prediction model.
14. The method of claim 10 , wherein, in the determination of the DTE reduced up to the present,
the DTE reduced up to the present is determined to be a value obtained by multiplying the SOC-based battery consumption energy used up to the present after the charging of the battery and an energy efficiency up to the present, and
the SOC-based battery consumption energy used up to the present is determined to be a value obtained by multiplying a battery SOC reduced up to the present after the charging of the battery and the maximum available battery energy at the full charging of the battery.
15. A system for determining a distance to empty (DTE) of a vehicle, the system comprising:
a vehicle external system providing real-time transportation situation information from a current position of the vehicle to a destination on a travel route to the destination;
a navigation terminal provided in the vehicle to transmit destination information received from a driver to the vehicle external system and receive information regarding the travel route to the destination and a map and the real-time transportation situation information from the vehicle external system; and
a controller configured for predicting and determining predicted battery consumption energy consumed for traveling from the current position to the destination using a battery usage prediction model to which vehicle forward driving situation information on which a real-time transportation situation is reflected and current vehicle state information obtained from the vehicle are input, and determining a DTE at the destination as a predicted value at the current position using the determined predicted battery consumption energy to the destination,
wherein the vehicle forward driving situation information to the destination, on which the real-time transportation situation is reflected, is obtained from the information regarding the travel route to the destination and the map and the received transportation situation information by the navigation terminal provided in the vehicle or the controller,
wherein the current vehicle state information includes a residual battery capacity.
16. The system of claim 15 , wherein the current vehicle state information further includes an electric load except for energy consumed for traveling of the vehicle, and an ambient temperature detected by a sensor.
17. The system of claim 15 ,
wherein the real-time transportation situation information to the destination includes road link-specific distances and average driving times of the vehicle on a travel route from the current position to the destination of a set travel route, and
wherein the vehicle forward driving situation information to the destination includes a speed and an acceleration of the vehicle on the travel route to the destination, determined from the road link-specific distances and the average driving times.
18. The system of claim 17 , wherein the speed and the acceleration are combined as values of a same time window, and the current vehicle state information obtained at the same time as the speed and the acceleration is processed by being allocated to the same time window as the speed and the acceleration.
19. The system of claim 18 , wherein the battery usage prediction model is a machine learning-based prediction model.
20. The system of claim 19 , wherein the battery usage prediction model is configured to:
extract feature information of the speed and the acceleration by processing time-series data of the speed and the acceleration using a first neural network; and
output the predicted battery consumption energy by processing the extracted feature information and the current vehicle state information as inputs using a second neural network.