APPARATUS AND METHOD FOR CALCULATING PREDICTION SPEED BASED ON DEEP LEARNING IN A CAR ACCIDENT

• Main Authors: YOO HYUN SANG, PARK DONG SEOK
• Format: Patent
• Language: English; Korean
• Published: 06.07.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS; SIGNALLING; TRAFFIC CONTROL SYSTEMS

The present invention relates to an apparatus for calculating prediction speed based on deep learning according to the occurrence of an accident. The apparatus includes: an accident information collection part that collects accident information when an accident occurs; a control part that calculates a prediction speed according to a traffic condition by setting a temporal and spatial congestion influence area based on a convolutional neural network (CNN) based on the accident information; and a communication part that transmits the accident information or the prediction speed information to a navigation device. 본 발명은 사고발생에 따른 딥러닝 기반의 예측속도 산출 장치에 관한 것으로, 사고 발생 시, 사고 정보를 수집하는 사고 정보 수집부; 상기 사고 정보에 기초하여 CNN(Convolutional Neural Network) 기반으로 시공간적 정체 영향권 범위를 설정하여 교통상황에 따른 예측속도를 산출하는 제어부; 및 상기 사고 정보나 상기 예측속도 정보를 내비게이션 장치에 전달하는 통신부;를 포함한다.