海底观测网数据传输系统构架设计、性能分析与故障定位方法研究

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 水下机器人研究室

题名	海底观测网数据传输系统构架设计、性能分析与故障定位方法研究
作者	孙凯
学位类别	博士
答辩日期	2017-12-06
授予单位	中国科学院沈阳自动化研究所
授予地点	沈阳
导师	王晓辉
关键词	海底观测网，有缆数据传输系统建模仿真，无线数据传输系统设计及信道分析，无线通信系统接口设计，故障定位方法
其他题名	Structural Design, Performance Analysis and Fault Location Method of the Data Transmission System of Subsea Observatories
学位专业	机械电子工程
中文摘要	本文以从国家大科学工程之国家海底科学观测系统为使用背景，以构建满足技术需求和目标要求的观测网海底光电缆和无线数据相结合的数据传输系统为目标，研究网络的构成、参数设置、接口设计、智能化等管理等网络建设的基本问题，主要包括：有缆数据传输系统构架设计及测试、有缆数据传输系统传输性能建模和分析、无线数据传输系统接口设计、调制方法研究、网络故障分析和定位方法。 (1) 首先研究海底观测网有缆数据传输系统构架和拓扑结构。主要包括两方面内容，一是数据传输网络采用什么样的拓扑结构形式，二是主干传送网络基于何种传送技术。拓扑结构是数据传送网络的基本要素之一，拓扑结构影响着网络的传送能力、鲁棒性、数据吞吐量、建设成本、网络管理、网络优化、故障定位等诸多重要指标和技术层面。海洋环境下，网络节点的增减相比于陆地网络困难许多，水下网络运行和维护多数通过远程监控操作实现，因此，合理的网络拓扑结构设计，可以从网络资源储备和硬件结构上提高网络可靠性，降低运行维护成本。传送技术直接影响着网络的传送能力、传送带宽、业务类型、管理能力、灵活性适应性等数据传送层面的指标。随着数据量的不断增大及传送业务类型的不断变化，传送技术也经历了多次演进。本文深入分析了已建成海底观测网传送网应用的技术，并结合目前电信行业主流传送技术的发展趋势，提出使用OTN（Optical Transport Network，光传送网）构架作为主干传送网骨干层传送技术、SDH（Synchronous Digital Hierarchy，同步数字体系）作为汇聚层传送技术的思想，建立新型的海底观测有缆数据传输系统构架。该构架构建了三层结构的海底观测网有缆数据传输系统，可实现从最底层的海洋传感器数据采集到海底数据集中汇聚、监控、路由再到岸基站实现信息集中可视化、分类及储存等服务。网络构架可实现基于因特网的数据服务，即搭建从因特网任一节点到海洋传感器之间的直接连接。该架构可提供面向网络服务的通用基础支撑环境，实现传感、存储和计算服务的聚合与共享，从而支持全国（全球）范围内的科研社区和虚拟实验环境，为后续的大数据处理、科学技术创新和工程应用奠定基础。通过实际设备进行了吞吐量、丢包率、时延、背靠背等指标的测试。依据相关行业标准，测试结果均高于指标规定值，验证了设计的合理性。(2) OTN光纤传输网络传输性能的优劣，与网络关键参数计算和配置直接相关。而网络关键参数的计算的前提条件是传输性能模型的建立。网络传输性能关键参数的建模，与网络具体结构、层次紧密相关。因此，本文针对海底观测网网络拓扑结构设计，建立传输功率、光信噪比、光放大器等关键指标和器件的模型，通过这些模型，研究传输距离与传输功率的关系，研究光信噪比约束下，网络的有效传输距离以及光放大器如何使用等关键问题，提出用于网络设计的数学模型，为网络设计提供了依据。本文提出了海底观测网无放大器传输性能模型，根据该模型，获得了该条件下的最大传输距离，为海底观测网主节点之间的距离设计提供了依据；本文研究了不同放大器的数学模型，并根据放大器模型，得到了增加不同类型放大器之后，数据传输距离的增益值，为海底观测网放大器的配置和对传输距离的改进提供了计算依据。本文建立了海底观测网拓扑结构下，光信噪比的数学模型，根据该模型，可以获得在一定信噪比要求条件下的数据有效传输距离，从而进一步明确是否需要增加提高信噪比设备已达到传输距离的要求。(3) AUV等移动平台和海底观测网进行数据交换，可用声学光学等多种技术手段。相比于声学数据传输，光学数据传输视具体使用条件的不同可达到兆级甚至千兆级的通信速率。海底观测网和移动平台的数据交换，从通信的建立到通信的保持都十分困难，因此，期望更高的通信速率以期降低缩短通信时间，尽快完成既定量数据的有效传输，本文研究基于水下无线光通信的海底观测网和移动平台之间数据传输技术问题。本文的研究具有鲜明的针对性，不研究水下无线光通信的基本原理，主要研究无线/有限光通信系统之间的接口问题，主要包括光源选择、主要调制技术分析和选择。针对研究背景及研究内容针对的使用环境，本文建立了海洋环境下无线光通信信道模型。通过该模型，在已知海水水文相关参数的前提下，能够得到某一发射功率下光的有效传播距离。传播距离这一数据对于海底观测网和AUV等移动平台在无人自主条件下建立有效通信十分重要。通过该模型，AUV等移动平台如果可以预先获得这一数据，将直接影响AUV自主接近海底观测网基站的路径规划算法问题。无线光通信调制方式关系到无线通信系统的功率消耗、传输容量等关键参数。本文对几种主要的调试方式进行理论分析，并根据应用背景选取了适合海底观测网和移动平台应用的无线光通信调制方式。(4) 研究数据传输系统的故障分析定位方法。本章提出基于卷积神经网络（CNN）算法的海底观测网数据传输系统故障分析和定位算法。算法将所有传感器统一看成一个故障对象，同时同步进行搜索，并深度结合海底观测网的拓扑结构，有针对性的进行故障的分析和定位。算法采用了自下而上的故障分析定位思想，以数量最多、数据量最大的传感器作为突破口，与海底观测网网络设备的自带功能相结合，考虑海底观测网拓扑和网络结构，层层推进算法，最终实现包含所有传感器和网络设备在内的故障分析和定位，探索适合海底观测网的故障定位方法。算法可实现不同目标函数的故障定位，比如故障发生时间（即某段时刻内有没有发生故障）、故障产生方位的群特性、故障产生的具体传感器等。多角度、全方位的故障分析定位技术，有助于统筹、全面的洞悉整个观测网设备运行健康情况及网络传输健康情况，真正从海底观测网建设维护者角度考虑故障定位问题，而不仅仅是独立的网络设备提供商或者传感器提供商的个体视角，这也是本文重要思想和观点之一。后续，作者也将沿着这一思想继续开展深入研究，以数据为对象和根本，研究海底观测网自身安全问题以及结合海洋传感器，探索其背后的海量数据挖掘技术。借助三亚海底观测网示范网积累的数据，以三亚南海海底观测网现有数据库数据为训练和测试样本，进行算法的训练和测试，以故障定性分析、故障源分析定位、故障传感器分析定位三种方式分别检验了算法的有效性，取得良好效果。
英文摘要	In this paper, the National Seafloor Observatory System of the Huge Scientific Project is used as the background, and construction of the data transmission system that combines optical data transmission system and the wireless data transmission system of the seafloor observation network is taken as the goal, we study its basic problems, mainly including: the design of the architecture of cabled data transmission system, the performance analysis of cable data transmission, the design of the interface of wireless data transmission system, the research of modulation method and transmission performance analysis. (1) Study cable data transmission system architecture of Seafloor Observation Network from the shore base station to the interfaces of the marine sensor. Establish a new type of cable data transmission system framework of Seafloor Observation Network. The framework consists of a three-layer Seafloor Observation Network with cable data transmission system, which can realize the centralized storage, management, integration and mining of information from the bottom of the marine sensor data collection to the submarine data convergence layer, and then to the shore base station for monitoring, routing, visualization, and other computing and information processing services. The architecture network could realize data services based on Internet, that is to build direct connections between any node from the Internet to the marine sensors. The architecture provides a common infrastructure for network services to realize the aggregation and sharing of sensing, storage and computing services, so as to support the nationwide (worldwide) research communities and virtual experimental environments, and lay the foundation for the following big data processing, scientific and technological innovation and engineering applications. Through the simulation of the model and the actual test of the miniature version, the principle of the architecture, the feasibility of the technology and the performance of key parameters such as XX are verified. (2) Analyze transmission performance of the physical layer based on the improved SSFM. We Model the light impairment at the physical layer, and the modified model considers not only the linear impairment, but also the non-linear effects such as dispersion and polarization. The model assumes that the power loss of the optical signal is exactly compensated by the optical amplifier, while the ASE noise generated by the EDFA can be filtered out at the receiver. Therefore, the entire optical path can be regarded as the series connection of network nodes and optical fibers. The main transmission impairments are node crosstalk, chromatic dispersion, PMD and nonlinear effects. First of all, the transmission of light is described by the nonlinear Schrodinger equation. Based on the type of optical fiber used by the Seafloor Observation Network, some basic parameters such as Kerr coefficient and fiber cross-sectional area are taken, and then the Fourier transform is performed. Considering the actual resource consumption and computational efficiency, the fast Fourier transform (FFT) is improved and the computational efficiency is improved. (3) Mobile platforms, such as AUV could exchange data with Seafloor Observatory Network by use of acoustic and optical technologies. Compared with the acoustic data transmission, the optical data transmission can reach the mega-level or even gigabit-level communication rate depending on the specific conditions of use. Therefore, a higher communication rate is expected in order to reduce the communication time and accomplish the effective transmission of a given amount of data as soon as possible. This paper studies technical problems of the data exchange between the Seafloor Observatory Network and the mobile platform by the means of wireless optical communication. The research in this paper has distinct pertinence. It does not study the basic principle of underwater wireless optical communication. It mainly studies the interface problems between wireless/wired optical communication systems, including the selection of light source, the analysis and selection of the main modulation techniques. In view of the research background and research environment, the paper establishes a wireless optical communication model in the marine environment. The effective propagation distance of light under a certain transmitting power can be obtained under the precondition of known parameters of seawater through this model. Propagation distance is important for establishing effective communication under unmanned conditions for seafloor observation network and mobile platform, such as AUV. If data can be obtained in advance, mobile platforms, such as AUV, could directly affect path planning algorithm for AUV to autonomous approach to the base station of seafloor observation network by using this model. Modulation of wireless optical communication related to the power consumption, transmission capacity and other key parameters of the wireless communication system. In this paper, we carry on the theoretical analysis to several main modulation methods, and chooses the wireless optical communication modulation mode suitable for the application of the seafloor observation network and the mobile platform according to the application background. (4) Research on fault analysis and location method of cable data transmission system. The data transmission system of Seafloor observation Network can be divided into circular trunk and vertical distribution of the communication branch. Different from ordinary metropolitan area network, the Seafloor observation Network is based on a large number of marine sensing devices mounted on the network as the basic service object. Therefore, the traditional positioning method for the backbone ring only does not suitable for Seafloor observation Network. Many of the ocean observing equipment attached to the Seafloor observation Network require continuity of data, and data lose their meaning if they lose continuity. This is also the significance of continuous in-situ observation. Therefore, our researches focus on the matrix algorithm of the fault location based on the combination of monitoring ring and monitoring track. It uses the detection loop to locate the main faults of the track and the monitoring road to locate the longitudinal faults. The fault location algorithm combined with the monitoring loop and the monitoring track can greatly reduce the computational complexity and improve the corresponding time of the algorithm to meet the real-time requirements of Seafloor observation Network to the maximum extent. Therefore, the Seafloor observation Network is different from the ordinary optical network. Taking into account the reliability and redundancy, both transmission channels and transmission equipment are basically taken redundant backup. The main transmission path adopts the dual-ring fiber channel. Due to adopting the wavelength division multiplexing, optical paths are actually virtual optical channels constructed by wavelength. Therefore, the concept of a monitoring ring cannot be limited to the physical optical fiber ring but based on virtual optical path composed of different wavelengths. On each ring light path, it can be assumed that various devices are vertically mounted lots of equipment, including transmission equipment and marine monitoring equipment, each of which may become a point of failure (Figure 20). The ultimate goal of the fault location of data transmission in Seafloor observation Network is to locate a certain device or a node on a transmission ring. Therefore, this paper proposes a fault location matrix algorithm based on the monitoring ring and the monitoring channel.
语种	中文
产权排序	1
页码	94页
内容类型	学位论文
源URL	[http://ir.sia.cn/handle/173321/21270]
专题	沈阳自动化研究所_水下机器人研究室
推荐引用方式 GB/T 7714	孙凯. 海底观测网数据传输系统构架设计、性能分析与故障定位方法研究[D]. 沈阳. 中国科学院沈阳自动化研究所. 2017.