METHOD FOR AUTO BRAKING OF CAR USING LEARNING MODEL AND APPARATUS USING THE SAME

• Main Author: CHOI BUM SUK
• Format: Patent
• Language: English; Korean
• Published: 16.05.2019
• Subjects: ARRANGEMENT OF BRAKING ELEMENTS ON VEHICLES IN GENERAL; BRAKE CONTROL SYSTEMS OR PARTS THEREOF, IN GENERAL; CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE ORDIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; PERFORMING OPERATIONS; PORTABLE DEVICES FOR PREVENTING UNWANTED MOVEMENT OFVEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TOTHE CONTROL OF A PARTICULAR SUB-UNIT; TRANSPORTING; VEHICLE BRAKE CONTROL SYSTEMS OR PARTS THEREOF; VEHICLE MODIFICATIONS TO FACILITATE COOLING OF BRAKES; VEHICLES IN GENERAL

Disclosed are an automatic braking method of a car using a machine learning model and an apparatus therefor. An embodiment of the present invention provides the automatic braking method of the car, in which various kinds of sensing data with respect to the car are collected based on various kinds of sensors mounted in the car. The various kinds of sensing data are temporally synchronized based on timestamps to generate temporally synchronized data, a learning target frame related to braking of the car is extracted from the temporally synchronized data based on brake sensing data collected by a brake sensor among the various kinds of sensors. The learning target frame is used to educate a collision preventing model of the car, and the collision preventing model is used to analyze the various kinds of sensing data to automatically brake the car. The present invention can provide the collision preventing model customized for each driver through collision preventing learning according to a driving habit. 기계 학습 모델을 이용한 차량의 자동 제동 방법 및 이를 위한 장치가 개시된다. 본 발명의 일실시예에 따른 차량의 자동 제동 방법은 차량에 설치된 다종의 센서들을 기반으로 차량에 대한 다종의 센싱 데이터를 수집하고, 타임 스탬프를 기반으로 다종의 센싱 데이터를 시간을 기준으로 동기화하여 타임 동기화 데이터를 생성하고, 다종의 센서들 중 브레이크 센서에서 수집된 브레이크 센싱 데이터를 기반으로 타임 동기화 데이터 중 차량의 제동과 관련된 학습 대상 프레임을 추출하고, 학습 대상 프레임을 이용하여 차량의 충돌 방지 모델을 학습시키고, 충돌 방지 모델을 통해 다종의 센싱 데이터를 실시간으로 분석하여 차량을 자동으로 제동한다.