Method for predicting a passage time of an intersection in consideration of irregular rotation and signal waiting, and device using the same
Disclosed are an apparatus and a method for predicting a passage time and a driving time of vehicles in a waiting line section, which pass through the waiting line section disposed behind an intersection. The apparatus inputs information on speeds of the vehicles to a trained prediction model and more accurately predicts the time for users to arrive at his or her destinations.
1. A method of predicting a passage time comprising:
collecting traffic information comprising speeds of vehicles that pass through a waiting line section disposed behind an intersection in each of a plurality of directions of progress of the vehicles in the waiting line section;
calculating passage times of vehicles that pass through the waiting line section in each direction of the progress based on the traffic information and a length of the waiting line section;
calculating an average passage time for each direction of the progress based on the passage times;
inputting the average passage time to a prediction model trained to predict a future passage time;
deriving a predicted passage time for each direction of the progress from the trained prediction model; and
displaying a route over specific geographic areas of a map based on the predicted passage time for each direction of the progress.
2. The method of claim 1 , wherein:
the average passage time is derived by calculating passage times corresponding to 10 minutes before and after derivation timing of the vehicles that have passed through the waiting line section in each direction of the progress by using an arithmetic mean method, and
the derivation timing indicates timing of a 5-minute unit at which the average passage time is derived.
3. The method of claim 1 , wherein:
the prediction model is trained by receiving average passage times that have been accumulated from the past according to time series,
the average passage times that have been accumulated from the past are derived by calculating passage times corresponding to 10 minutes before and after derivation timings of the vehicles that have passed through the waiting line section in each direction of the progress by using an arithmetic mean method,
the derivation timing indicates timing of a 5-minute unit at which the average passage time is derived, and
the predicted passage time is derived by being divided in a unit of 5 minutes up to an hour in a future from an actual time for each direction of the progress.
4. The method of claim 1 , wherein the waiting line section is differently set for each direction of the progress of the vehicles.
5. The method of claim 1 , wherein the trained prediction model is a model trained by deep learning based on an artificial neural network.
6. The method of claim 1 , further comprises generating waiting line section information comprising a length, a road grade, a road type, and a number of lanes of the waiting line section, a signal type of the intersection in each direction of the progress and environment information comprising information on a day of a week, a time, and an administrative district,
wherein the average passage time, the waiting line section information, and the environment information are input into the trained prediction model.
7. An apparatus for predicting a passage time, the apparatus comprising:
a collection circuit configured to collect traffic information, comprising speeds of vehicles that pass through a waiting line section disposed behind an intersection in each of a plurality of directions of progress of the vehicles in the waiting line section;
an arithmetic circuit configured to calculate passage times of vehicles that pass through the waiting line section in each direction of the progress based on the traffic information and a length of the waiting line section and configured to calculate an average passage time for each direction of the progress based on the passage times;
a prediction circuit comprising a prediction model trained to predict a future passage time and configured to input the average passage time to the trained prediction model and configured to derive a predicted passage time in each direction of the progress; and
a display screen which displays a route over specific geographic areas of a map based on the predicted passage time in each direction of the progress.
8. The apparatus of claim 7 , wherein:
the arithmetic circuit is configured to derive the average passage time by calculating passage times corresponding to 10 minutes before and after derivation timing of the vehicles that have passed through the waiting line section in each direction of the progress by using an arithmetic mean method, and
the derivation timing indicates timing of a 5-minute unit at which the average passage time is derived.
9. The apparatus of claim 7 , further comprising: a management circuit configured to generate waiting line section information comprising a length, a road grade, a road type, and a number of lanes of the waiting line section, a signal type of the intersection for each direction of the progress and environment information comprising information on a day of a week, a time, and an administrative district.
10. The apparatus of claim 7 , further comprising: a transmission circuit configured to transmit the predicted passage time to a vehicle navigation management server.
11. The apparatus of claim 7 , further comprising: a learning circuit configured to train the prediction model by inputting, to the prediction model, average passage times that have been accumulated from the past according to time series,
wherein the average passage times that have been accumulated from the past are derived by calculating passage times corresponding to 10 minutes before and after derivation timings of the vehicles that have passed through the waiting line section in each direction of the progress by using an arithmetic mean method, and
the derivation timing indicates timing of a 5-minute unit at which the average passage time is derived.
12. A method of predicting a driving time, comprising:
collecting traffic information comprising speeds of vehicles that pass through waiting line sections disposed behind intersections in each of a plurality of directions of progress of the vehicles in the waiting line sections;
calculating passage times of vehicles that pass through the waiting line sections for each waiting line section and in each direction of the progress based on the traffic information and lengths of the waiting line sections;
calculating an average passage time in each direction of the progress based on the passage times;
inputting the average passage time to a prediction model trained to predict a future passage time;
deriving predicted passage times for each direction of the progress from the trained prediction model;
deriving a route having a smallest sum of predicted passage times, among the waiting line sections through which the vehicles pass through in order to arrive at their destinations; and
displaying the route over specific geographic areas of a map based on the route having the smallest sum of predicted passage times.
13. The method of claim 12 , wherein:
the average passage time is derived by calculating passage times corresponding to 10 minutes before and after derivation timing of the vehicles that have passed through the waiting line sections in each direction of the progress by using an arithmetic mean method, and
the derivation timing indicates timing of a 5-minute unit at which the average passage time is derived.
14. The method of claim 12 , further comprising generating waiting line section information comprising a length, a road grade, a road type, and a number of lanes of the waiting line section, a signal type of the intersection for each direction of the progress and environment information comprising information on a day of a week, a time, and an administrative district,
wherein the trained prediction model is a model trained by deep learning based on an artificial neural network, and
the average passage time, the waiting line section information, and the environment information are input into the trained prediction model.
15. The method of claim 12 , wherein:
the prediction model is trained by receiving average passage times that have been accumulated from the past according to time series,
the average passage times that have been accumulated from the past are derived by calculating passage times corresponding to 10 minutes before and after derivation timings of the vehicles that have passed through the waiting line section in each direction of the progress by using an arithmetic mean method,
the derivation timing indicates timing of a 5-minute unit at which the average passage time is derived, and
the predicted passage time is derived by being divided in a unit of 5 minutes up to an hour in a future from an actual time for each direction of the progress.