基于FRFT-RBF网络的船舶电网故障定位方法  

Fault Location Method for Ship Grid based on FRFT-RBF network

作　　者：王高阳 刘峰 WANG Gao-yang;LIU Feng(Wuhan Institute of Shipbuilding Technology,Wuhan 430050,China;Guangzhou Marine Engineering Design and Research Institute,Guangzhou 510250,China)

机构地区：[1]武汉船舶职业技术学院,湖北武汉430050 [2]中国船舶集团有限公司第六○五研究院,广东广州510250

出　　处：《武汉船舶职业技术学院学报》2023年第6期77-84,共8页Journal of Wuhan Institute of Shipbuilding Technology

基　　金：武汉船舶职业技术学院“船舶电网短路故障识别与定位方法研究”(课题编号:2021y19)。

摘　　要：为了解决大规模船舶电网在运行中的单相接地故障无法快速精准定位问题,提高船舶电网运行可靠性和生命力,针对船舶电网故障信号具有非平稳性的特点,通过对分数阶傅立叶变换(FRFT)基本理论、最优阶次确定方法、滤波去噪方法,以及功率谱和能量率理论的研究,提出了基于FRFT的电网故障特征提取方法和实现步骤;并针对船舶电网故障特性具有非规律性特点,通过对径向基神经网络算法(RBF)的研究,结合故障特征提取的方法,提出了基于FRFT-RBF网络的船舶电网故障定位方法和实现步骤。通过开展电网故障定位的仿真实验,验证了方法的可行性,为解决大规模船舶电网故障定位问题提供了可行的工程应用方法。The single-phase grounding fault of large-scale ship power system cannot be quickly and accurately located during operation.In order to solve this problem and improve the reliability and vitality of ship grid operation,a method for extracting fault characteristics based on Fractional Fourier Transform(FRFT)has been proposed due to the non-station-ary nature of fault signals in ship power systems,and this method involves the study of the basic theory of FRFT,determination of optimal orders,filtering and de-noising tech-niques,as well as power spectrum and energy rate theory.The non-regular characteristics of ship power system faults have led to the proposal of a fault localization method based on the FRFT-RBF network.This method combines research on Radial Basis Function Neural Network(RBF)algorithms with fault feature extraction techniques to achieve accurate fault localization in ship power systems.Through conducting simulation experiments on power grid fault localization,the feasibility of the method has been verified,providing a viable en-gineering application approach to address the issue of large-scale ship power system fault localization.

关 键 词：船舶电网 故障定位 FRFT 能量率 RBF 

分 类 号：U665[交通运输工程—船舶及航道工程]