• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2006년도 하계종합학술대회
• /
• pp.909-910
• /
• 2006

In this paper, the transfer function to the vehicle is derived from using system identification algorithm in connection with the driving vehicle. We design the adaptive cruise controller using the derived transfer function, and make it possible to monitoring and control the vehicle in real time using embedded system and technology of Internet.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.1123-1124
• /
• 2023

본 논문은 실시간 차선 인식을 기반으로 한 자율주행 알고리즘을 제안한다. 자율주행 알고리즘은 크게 차선 인식과 의사결정으로 구분된다. 차선 인식 부분에서는 직관적인 판단을 위해 버드 아이 뷰로 영상데이터를 변환하여 안정적 차선 인식을 위하여 차선 영역을 추출하고 노이즈를 제거하는 전처리과정을 거친다. 이렇게 처리된 영상에서 Hough 변환을 통하여 차선을 검출한다. 의사결정 부분에서는 검출된 차선과 현재 위치를 기반으로 진행할 경로를 결정한다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.1175-1176
• /
• 2023

전기차의 자율주행을 위해선 차선 인식과 모터 제어가 필요하다. 카메라로 입력된 영상에 허프 변환을 적용하고, 변환된 이진 이미지에 Enet 및 DeepLabv3+ 구조를 활용한 LaneNet 모델을 적용하여 차선을 학습시키고, Fuzzy 제어 기법을 활용하여 모터의 조향이 원활이 되도록 하였다. 기존의 Rule base 기법에 비하여 차선 인식 정확도가 월등히 향상되었으며, 주행 결과 Real-Time 주행환경 판단에 대한 여지를 남겼다.

• 한국항만학회지
• /
• 제14권3호
• /
• pp.291-301
• /
• 2000

T his paper presents the lateral and longitudinal control algorithm for the driving of a 4WS AGV(Automated Guided Vehicle). The control law to the lateral and longitudinal control of the AGV includes adaptive agin tuning ability, that is the controller gain of the gravity compensated PD controller can be changed on a real-time. The gain tuning law is derived from the Lyapunov direct method using the output error of the reference model and the actual model, And to show the performance of the presented lateral and longitudinal control algorithm, we simulate toe nonlinear AGV equations of the motion by deriving the Newton-Euler Method, The read path is from quay yard area to docking position in loading yard area. The quay yard area is where the quay crane loads the container to the AGV and the docking position is where the container is transferred to the gantry crane. The road types are constructed in a straight line and J-turn. When driving the straight line, the driving velocity is 6㎧ and the J-turn is 3㎧.

• 한국자동차공학회논문집
• /
• 제13권3호
• /
• pp.35-40
• /
• 2005

In the present study, the influence of operating conditions on fuel economy for hybrid electric vehicle was analyzed. In order to accomplish this, vehicle speed, engine speed, battery current and voltage, SOC (state of charge),motor speed and torque, generator speed and torque, engine coolant temperature etc. were measured in real time. The tests were carried out under different driving cycles which are urban and highway cycles, KOREA CITY cycle and on-road driving, and also under various operating conditions such as different initial SOC, with or without regenerative braking etc.. Generally, conventional gasoline engines show a poor fuel economy at stop and go driving, because braking energy is wasted and the engine is operated in low thermal efficiency regions. However, in case of hybrid vehicles, higher fuel economy can be obtained because of utilizing the maximum thermal efficiency regions of engine, idling stop of engine, and regenerative braking etc..

• 전자공학회논문지SC
• /
• 제40권3호
• /
• pp.127-134
• /
• 2003

본 논문에서는 퍼지제어를 이용하여 사람의 얼굴을 추적하는 카메라를 구동하는 시스템을 구현한다. 팬틸트(Pan Tilt)구조를 가진 카메라 시스템은 먼저 영상을 PC로 보내고 PC에서는 추적알고리즘에 의한 추적 좌표를 다시 카메라에 전송하면 카메라는 목표 얼굴을 실시간으로 추적하는 방식으로 되어 있다. 카메라를 구동하는 2축의 스텝모터는 PC에서 전송된 목표 좌표로 최대한 빠르고 또한 부드럽게 제어되어야 한다. 이를 위해서 본 논문에서는 퍼지제어기를 제안하여 구동용 가·감속 주파수를 만들고 두 축의 스텝모터를 빠르면서도 부드럽게 제어한다 그리고 본 제안 방식의 효율성을 검증하기 위하여 실험 장치를 제작하고 실험을 수행한다.

• 한국멀티미디어학회논문지
• /
• 제20권8호
• /
• pp.1271-1281
• /
• 2017

It is essential to collect and analyze target information around the vehicle for autonomous driving of the vehicle. Based on the analysis, environmental information such as location and direction should be analyzed in real time to control the vehicle. In particular, obstruction or cutting of objects in the image must be handled to provide accurate information about the vehicle environment and to facilitate safe operation. In this paper, we propose a method to simultaneously generate 2D and 3D bounding box proposals using LiDAR Edge generated by filtering LiDAR sensor information. We classify the classes of each proposal by connecting them with Region-based Fully-Covolutional Networks (R-FCN), which is an object classifier based on Deep Learning, which uses two-dimensional images as inputs. Each 3D box is rearranged by using the class label and the subcategory information of each class to finally complete the 3D bounding box corresponding to the object. Because 3D bounding boxes are created in 3D space, object information such as space coordinates and object size can be obtained at once, and 2D bounding boxes associated with 3D boxes do not have problems such as occlusion.

• 제어로봇시스템학회논문지
• /
• 제20권12호
• /
• pp.1181-1188
• /
• 2014

In this paper, we propose a novel concept foran unmanned delivery service using a cooperative heterogeneous unmanned system consisting of a self-driving car and an unmanned aerial vehicle (UAV). The proposed concept is suitable to deliver parcels in high-density and high-rise urban or residential areas. In order to achieve the proposed concept, we will develop acooperative heterogeneous unmanned system. Customers can order goods using a smartphone application and the order information, including the position of the customer and the order time, and the package is transported automatically by the unmanned systems. The system assigns the tasks suitable for each unmanned vehicle by analyzing it based on map information. Performance is validated by experiments consisting of autonomous driving and flight tests in a real environment. For more evaluation, the landing position error analysis is performed using circular error probability (CEP).

• 한국자동차공학회논문집
• /
• 제11권6호
• /
• pp.183-189
• /
• 2003

The study of the automotive noise reduction has been concentrated on the reduction of the automotive engine noise because the engine noise is the major cause of automotive noise. However, many studies of automotive engine noise led to the interest of the noise reduction of the exhaust and intake system. Recently, the active control method is used to reduce the noise of an automotive exhaust and intake system. It is mostly used the LMS(Least-Mean-Square) algorithm as an algorithm of active control because the LMS algorithm can easily obtain the complex transfer function in real-time. Especially, Filtered-X LMS (FXLMS) algorithm is applied to an Active Noise Control system. However, the convergence performance of LMS algorithm went bad when the FXLMS algorithm was applied to an active control of the induction noise under rapidly accelerated driving conditions. So, in order to solve this problem, the modified FXLMS algorithm is proposed. In this study, the improvement of the control performance using the modified FXLMS algorithm under rapidly and suddenly accelerated driving conditions was identified. Also, the performance of an active control using the LMS algorithm under rapidly accelerated driving conditions was evaluated through the theoretical derivation using a chirp signal to have similar characteristics with the induction noise signal.

• 한국인공지능학회지
• /
• 제6권1호
• /
• pp.11-15
• /
• 2018

The purpose of this study is to develop a pulsatile drowsiness detection system that can compensate the limitations of existing camera - based or breathing pressure sensor based Drowsiness driving prevention systems. A heart rate sensor mounted on the driver's finger and an alarm system that sounds when drowsiness is detected. The heart rate sensor was used to measure pulse changes in the wrist, and an alarm system based on the Arduino, which works in conjunction with the laptop, generates an audible alarm in the event of drowsiness. In this paper, we assume that the pulse rate of the drowsy state is 60 ~ 65 times / minute, which is the middle between the awake state and the sleep state. As a result of the experiment, the alarm sounded when the driver's pulse rate was in the drowsy pulse rate range. Based on these experiments, the drowsiness detection system was able to detect the drowsiness of the driver successfully in real time. A more effective drowsiness prevention system can be developed in the future by incorporating the results of the present study on a pulse-based drowsiness prevention system in an existing drowsiness prevention system.