面向无人驾驶的智能车系统平台研究与应用

发布时间：2018-06-28 20:15

本文选题：环境感知服务 + 分布式处理　；参考：《浙江大学》2017年硕士论文

【摘要】：近年来,全球每年都有百万级数量的人死于车祸,其中90%的车祸都是由于驾驶员的错误驾驶所导致。这引起了国内外愈来愈多的机构和个人对无人驾驶智能车的研究。本文以面向无人驾驶的智能车系统平台的研究与应用为主要内容,对面向无人驾驶的智能车系统平台进行详细的设计与实现。本文所研究的面向无人驾驶的智能车系统平台的主要内容包括:(1)基于分布式的环境感知服务设计与实现。随着行车路况的日益复杂化,感知数据的处理复杂度不断增加,本课题的环境感知服务借鉴了机器人操作系统(Robot Operating System,ROS)的分布式设计理念,将不同的感知信息处理分布在不同的计算节点上进行分布式处理,处理结果信息以服务的方式提供给其他计算节点。本文的感知服务实现了感知硬件设备的虚拟化管理和感知软件服务。(2)基于AUTOSAR的车辆控制服务的设计与实现。AUTOSAR(AUTomotive Open System Architecture)标准化了电子控制单元(Electric Control Unit)的软件层次结构,使得汽车软件的开发更加规范,提高了汽车电子软件的可重用性和可移植性。本文的车辆控制服务基于AUTOSAR的RTE(Runtime Environment)层进行设计与实现,对车辆的被控对象进行了分类,并为每一类被控对象设计并实现了控制操作方法。(3)无人驾驶智能车系统平台服务访问接口的设计与实现。感知服务访问接口实现过程中,利用本地化缓冲和实时更新技术,保证了感知服务信息的实时性和有效性。车辆控制访问接口的设计考虑了底层VCU(Vehicle Control Unit)的控制协议差异性,该部分接口实现减小了行车策略层和底层控制协议的耦合度。(4)无人驾驶智能车系统平台实验验证。将无人驾驶智能车系统平台进行应用实验,并对实验结果进行分析,给出了实验结论。
[Abstract]:In recent years, millions of people worldwide have died in accidents every year, 90 percent of which are caused by the driver's wrong driving. This has caused more and more institutions and individuals at home and abroad to study driverless smart vehicles. In this paper, the research and application of driverless intelligent vehicle system platform is taken as the main content, and the design and implementation of the driverless intelligent vehicle system platform is carried out in detail. The main contents of this paper are as follows: (1) Design and implementation of distributed environment aware service. With the increasing complexity of traffic conditions and the increasing complexity of processing perceptual data, the distributed design concept of Robot operating system Ros (Robot operating system Ros) is used for reference in this paper. Different perceptual information processing is distributed on different computing nodes for distributed processing, and the processing result information is provided to other computing nodes in a service manner. The perception service in this paper realizes the virtualization management of hardware devices and the sensing software service. (2) the design and implementation of vehicle control service based on AUTOSAR. AUTOSAR (Auto Tomotive Open system Architecture) standardizes the software hierarchy of Electric Control Unit. It makes the development of automobile software more standardized and improves the reusability and portability of automotive electronic software. The vehicle control service in this paper is designed and implemented based on the RTE (Runtime Environment) layer of AUTOSAR. Control operation methods are designed and implemented for each class of controlled objects. (3) Design and implementation of service access interface for unmanned intelligent vehicle system platform. In the process of implementing the aware service access interface, the localization buffering and real-time updating technology are used to ensure the real-time and validity of the aware service information. The design of vehicle control access interface takes into account the differences of control protocols of the underlying VCU (vehicle Control Unit). This part of the interface reduces the coupling degree between the driving strategy layer and the bottom control protocol. (4) the platform experiment verification of driverless intelligent vehicle system. The application experiment is carried out on the platform of unmanned intelligent vehicle, and the experimental results are analyzed, and the experimental results are given.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U463.6;TP311.52

【参考文献】