• International Journal of Automotive Technology
• /
• 제3권2호
• /
• pp.71-78
• /
• 2002

A new control method far vision-based autonomous vehicle is proposed to determine navigation direction by analyzing lane information from a camera and to navigate a vehicle. In this paper, characteristic featured data points are extracted from lane images using a lane recognition algorithm. Then the vehicle is controlled using new Levenberg-Marquardt neural network algorithm. To verify the usefulness of the algorithm, another algorithm, which utilizes the geometric relation of a camera and vehicle, is introduced. The second one involves transformation from an image coordinate to a vehicle coordinate, then steering is determined from Ackermann angle. The steering scheme using Ackermann angle is heavily depends on the correct geometric data of a vehicle and a camera. Meanwhile, the proposed neural network algorithm does not need geometric relations and it depends on the driving style of human driver. The proposed method is superior than other referenced neural network algorithms such as conjugate gradient method or gradient decent one in autonomous lateral control .

• 센서학회지
• /
• 제10권5호
• /
• pp.329-336
• /
• 2001

This paper describes the steering control and geomagnetism cancellation for an autonomous vehicle using an MR sensor. The magneto-resistive (MR) sensor obtains the vector summation of the magnetic fields from embedded magnets and the Earth. The vehicle is controlled by the magnetic fields from embedded magnets. So, geomagnetism is the disturbance in the steering control system. In this paper, we propose a new method of the sensor arrangement in order to remove the geomagnetism and vehicle body interference. The proposed method uses two MR sensors located in a level plane and the steering controller has been developed. The controller has three input variables ($dB_x$, $dB_y$, $dB_z$) using the measured magnetic field difference, and an output variable (the steering angle). A simulation program was developed to acquire the data to teach the neural network, in order to test the ability of a neural network to learn the steering control process. Also, the computer simulation of the vehicle (including vehicle dynamics and steering) was used to verify the steering performance of the vehicle controller using the neural network. From the simulation and field test, good result was obtained and we confirmed the robustness of the neural network controller in a real autonomous vehicle.

• 대한임베디드공학회논문지
• /
• 제8권2호
• /
• pp.69-78
• /
• 2013

As vehicles become more intelligent for safety and convenience of drivers, in-vehicle networking systems such as controller are network (CAN) have been widely used due to increasing number of electronic control unit (ECU). Recently, FlexRay was developed to replace CAN protocol in chassis networking systems, to remedy the shortage of transmission capacity and unsatisfactory real-time transmission delay of conventional CAN. However, it is difficult for vehicle network designers to calculate platform configuration registers (PCR) and determine a base cycle or slot length of FlexRay. To assist vehicle network designers for designing FlexRay cluster, this paper presents automatic field bus exchange format (FIBEX) generation algorithm from CANdb information, which is de-facto standard database format for CAN. To design this program, structures of FIBEX, CANdb and relationship among PCR variables are analyzed.

• 디지털산업정보학회논문지
• /
• 제19권4호
• /
• pp.127-136
• /
• 2023

Many research results on the traffic enforcement of illegal driving of twowheeled motor vehicles using license plate recognition are introduced. Deep learning convolutional neural networks can be used for character and word recognition of license plates because of better generalization capability compared to traditional Backpropagation neural networks. In the plates of twowheeled motor vehicles, the interdependent government and city words are included. If we implement the mutually independent word recognizers using error correction rules for two word recognition results, efficient license plate recognition results can be derived. The CPU based convolutional neural network without library under real time processing has an advantage of low cost real application compared to GPU based convolutional neural network with library. In this paper twowheeled motor vehicle license plate recognition algorithm is introduced using CPU based deep-learning convolutional neural network. The experimental results show that the proposed plate recognizer has 96.2% success rate for outdoor twowheeled motor vehicle images in real time.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.711-716
• /
• 2003

This study considers the implementation issues of the inter-vehicle communication system for the vehicle platoon experiments via a testbed. The testbed, which consists of three scale vehicles and one RCS(remote control station), is developed as a tool for functions evaluation between simulation studies and full-sized vehicle researches in the previous study. The cooperative communication of the vehicle-to-vehicle or the vehicle-to-roadside plays a key role for keeping the relative spacing of vehicles small in a vehicle platoon. The static platoon control, where the number of vehicles remains constant, is sufficient for the information to be transmitted in the suitably fixed interval, while the dynamic platoon control such as merge or split requires more flexible network architecture for the dynamical coordination of the communication sequence. In this study, the wireless communication device and the reliable protocol of the flexible network architecture are implemented for our testbed, using the low-cost, ISM band transceiver and the 8-bit microcontroller.

• 한국전자통신학회논문지
• /
• 제15권2호
• /
• pp.299-306
• /
• 2020

신경망을 이용한 영상인식은 여러 분야에 널리 사용되고 있다. 본 연구에서는 차량 번호 인식 및 특정 구역 입출 시 통제에 필요한 인가/비인가 차량 인식 시스템을 연구하였다. 이 시스템은 영상을 인식하는 기능을 갖추고 있어 차량 번호에 대한 모든 정보를 확인하고, 차량 번호판을 정확히 인식할 수 있는 기능을 추가하였다. 그 밖에 신경망을 이용하여 좀 더 빠르게 차량번호를 확인할 수 있도록 하였다.

• 한국통신학회논문지
• /
• 제42권2호
• /
• pp.424-431
• /
• 2017

요즘 차량 네비게이션 시스템은 큰 관심 분야이다. GNSS(Global Navigation Satellite System)은 실외 측위를 위한 기술 중 핵심적인 기술이다. 그러나 GNSS는 높은 정확도와 신뢰도를 제공하지 못한다. 이러한 이유로, 우리는 차량의 GNSS 성능의 정확도를 향상시키기 위하여 Network-RTK를 적용하였다. 이 Network-RTK 모드에서 GNSS 에러는 급격히 감소하게 된다. 본 논문에서 우리는 ntrip client 프로그램을 설명하고 다양한 환경에서의 실험 결과를 보여준다.

• 한국정밀공학회지
• /
• 제26권10호
• /
• pp.40-46
• /
• 2009

Recently CAN (Controller Area Network) is widely used as an in-vehicle networking protocol for intelligent vehicle. The identifier field (ID) of CAN is used not only to differentiate the messages but also to give different priorities to access the bus. This paper presents a dynamic 10 allocation algorithm in order to enhance the real-time quality-of-service (QoS) performance. When the network traffic is increased, this algorithm can allocate a network resource to lower priority message without degradation of the real-time QoS performance of higher priority message. In order to demonstrate the algorithm's feasibility, message transmission delays have been measured with and without the algorithm on an experimental network test bed.

• 한국항행학회논문지
• /
• 제11권4호
• /
• pp.401-407
• /
• 2007

본 논문에서는 차량 내 편의장치를 제어하는"body"영역과 멀티미디어 데이터를 전송하는 "멀티미디어" 영역을 위해서 이더넷 기반의 차량 내 네트워크 프로토콜을 제안한다. 이더넷 기반의 차량 내 네트워크 프로토콜은 해당 응용 분야의 요구를 만족시키기 위해 각 노드에서 트래픽 제어를 수행하는데 이를 위해서 2계층과 네트워크 토폴로지를 수정하였다. 링 토폴로지를 갖는 기존의 MOST 프로토콜과 본 논문에서 제안하는 링 토폴로지의 이더넷 기반의 프로토콜을 시뮬레이션하여 그 성능을 비교 분석하였다. 그 결과 본 논문에서 제안하는 토폴로지 형태에서 15% 성능 개선을 확인하였다. 또한 리눅스 기반에서 구현한 시스템을 통해 실시간 멀티미디어 패킷이 베스트 에포트 패킷과 공존 시에 QoS를 보장 받는 것을 확인하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.131-134
• /
• 1997

This paper presents a new approach to the design of intelligent control system for track vehicle system using fuzzy logic based on neural network. The proposed control scheme uses a Gaussian function as a unit function in the neural network-fuzzy, and back propagation algorithm to train the fuzzy-neural network controller in the framework of the specialized learning architecture. It is proposed a learning controller consisting of two neural network-fuzzy based on independent reasoning and a connection net with fixed weights to simply the neural networks-fuzzy. The performance of the proposed controller is illustrated by simulation for trajectory tracking of track vehicle speed.