数据挖掘的船舶电力电路短路识别被引量：1

Short circuit recognition of ship power circuit based on data mining

在线阅读下载PDF

导出

摘要电路短路类型多样复杂,当前方法的电路短路识别错误概率高,为降低电路短路识别的错误概率,设计了基于数据挖掘的船舶电力电路短路识别方法。首先采用传感器采集船舶电力电路工作状态信号,并采用小波包对船舶电力电路工作状态信号进行多层分解,提取相应的小波包能量熵,将其作为船舶电力电路短路识别的特征,然后采用数据挖掘技术对特征向量和船舶电力电路短路类型间的变化关系进行建模,设计船舶电力电路短路识别模型,最后在Matlab 2017平台进行了船舶电力电路短路识别实验,本文船舶电力电路短路识别正确率超过95%,而对比方法的船舶电力电路短路识别率低于90%,本文方法不仅大幅度降低电路短路识别的错误概率,而且船舶电力电路短路识别效率更高,能够用于实际的船舶电力系统管理。 There are many types and complexity of short circuit in the circuit,and the error probability of short circuit identification in current methods is high.In order to reduce the error probability of short circuit identification,a short circuit identification method of marine power circuit based on data mining is designed.Firstly,the sensor is used to collect the working state signal of the ship power circuit,and the wavelet packet is used to decompose the working state signal of the ship power circuit.The energy entropy of the corresponding wavelet packet is extracted and used as the feature of the shortcircuit knowledge of the ship power circuit.Then,the data mining technology is used to model the relationship between the feature vector and the short-circuit type of the ship power circuit.The short-circuit identification model of ship power circuit is established.Finally,the short-circuit identification experiment of ship power circuit is carried out on Matlab 2017 platform.The correct rate of short-circuit identification of ship power circuit in this paper is more than 95%,while that of ship power circuit in comparison method is less than 90%.This method greatly reduces the error probability of short-circuit identification of ship power circuit,and the efficiency of short-circuit identification of ship power circuit is higher.It can be used in the actual management of ship power system.

作者张红丽王春红 ZHANG Hong-li;WANG Chun-hong(The Department of Electrical Power Engineering,Zhengzhou Electric Power Vocational and Technical College,Zhengzhou 451450,China)

机构地区郑州电力职业技术学院电力工程系

出处《舰船科学技术》北大核心 2019年第12期85-87,共3页 Ship Science and Technology

关键词船舶电力系统小波包能量熵数据挖掘短路识别 marine power system wavelet packet energy entropy data mining short-circuit identification

分类号 U665 [交通运输工程—船舶及航道工程]

作者简介张红丽(1981-),女,硕士,讲师/电气工程师,研究方向为电力系统电气工程。