﻿ 舰船装备电子电路故障区域节点跟踪定位方法
 舰船科学技术  2023, Vol. 45 Issue (12): 152-155    DOI: 10.3404/j.issn.1672-7619.2023.12.030 PDF

Tracking and locating method of electronic circuit fault area node in ship equipment
MEI Jie, GAO Feng
Hubei University of Technology Engineering and Technology College, Wuhan 430068, China
Abstract: This paper studies the fault area node tracking and locating method of ship equipment electronic circuit, improves the tracking and locating effect of fault area node, and provides guarantee for ship safe navigation. The least squares support vector machine is used to process the original data of the electronic circuits of ship equipment in different states. The empirical mode decomposition algorithm is used to extract the fault characteristics of electronic circuits in different states from the data after extended processing. By introducing weights in the Mahalanobis distance, an improved Mahalanobis distance algorithm is obtained. The improved Mahalanobis distance algorithm is used to calculate and extract the Mahalanobis distance between the fault feature and the known fault set. Experimental results show that the proposed method can effectively extract the fault characteristics of ship equipment electronic circuits under different states. This method can track and locate the fault area node of electronic circuit effectively, and the tracking and positioning accuracy is high.
Key words: ship's equipment     electronic circuit     fault area node     tracking and positioning     support vector machine     mahalanobis distance
0 引　言

1 电子电路故障区域节点跟踪定位

1.1 基于EMD的舰船装备电子电路故障特征提取

1) 确定全部舰船装备电子电路原始数据的全部局部极值点 $a\left( t \right)$ ，共包含2种类型，分别是极大、小值点，记作 ${a_{\max }}\left( t \right)$ ${a_{\min }}\left( t \right)$

2) 通过三次样条线分别连接 ${a_{\max }}\left( t \right)$ ${a_{\min }}\left( t \right)$ ，获取上、下包络线 ${q_{up}}\left( t \right)$ ${q_{down}}\left( t \right)$ ${q_{up}}\left( t \right)$ ${q_{down}}\left( t \right)$ 内包含全部舰船装备电子电路原始数据点。令 ${q_{up}}\left( t \right)$ ${q_{down}}\left( t \right)$ 的均值是 ${q_1}\left( t \right)$

3) 在 $x\left( t \right)$ 内去掉 ${q_1}\left( t \right)$ 获取新的数据 ${z_1}\left( t \right)$ ，公式如下：

 ${z_1}\left( t \right) = x\left( t \right) - {q_1}\left( t \right) 。$ (1)

4) 若 ${z_1}\left( t \right)$ 不符合IMF条件，则 ${z_1}\left( t \right)$ 以为原始数据，返回步骤1，获取新的均值 ${q_{11}}\left( t \right)$ ，在 ${z_1}\left( t \right)$ 内去掉 ${q_{11}}\left( t \right)$ 得到新的数据 ${z_{11}}\left( t \right)$ ，公式如下：

 ${z_{11}}\left( t \right) = {z_1}\left( t \right) - {q_{11}}\left( t \right) ，$ (2)

 ${z_{1k}}\left( t \right) = {z_{1\left( {k - 1} \right)}}\left( t \right) - {q_{1k}}\left( t \right)。$ (3)

5) 在 $x\left( t \right)$ 内，分离 ${c_1}\left( t \right)$ ，获取：

 ${r_1}\left( t \right) = x\left( t \right) - {c_1}\left( t \right) 。$ (4)

 $\left\{ \begin{gathered} {r_2}\left( t \right) = {r_1}\left( t \right) - {c_2}\left( t \right) ，\\ {r_3}\left( t \right) = {r_2}\left( t \right) - {c_3}\left( t \right)，\\ \mathop {}\nolimits_{} \mathop {}\nolimits_{} \mathop {}\nolimits_{} \mathop {}\nolimits_{} \vdots \\ {r_n}\left( t \right) = {r_{n - 1}}\left( t \right) - {c_n}\left( t \right)。\\ \end{gathered} \right.$ (5)

${r_n}\left( t \right)$ 无法提取符合IMF条件的分量时，结束操作，获取：

 $x\left( t \right) = \sum\limits_{i = 1}^n {{c_i}\left( t \right) + {r_n}\left( t \right)} ，$ (6)

6) 计算各IMF分量的能量 ${E_i}$ ，公式如下：

 ${E_i} = \sum\limits_{j = 1}^{{M_i}} {{{\left| {c_i^j\left( t \right)} \right|}^2}} 。$ (7)

7) 以能量熵 $H$ 为舰船装备电子电路故障特征，公式如下：

 $H = - \sum\limits_{i = 1}^n {\frac{{{E_i}}}{E}\ln } \frac{{{E_i}}}{E}。$ (8)

 $f\left( x \right) = \sum\limits_{l = 1}^\eta {{\alpha _l}} K\left( {{x_l},{x_\beta }} \right) + b 。$ (9)

 $x_l^*\left( t \right) = f\left( {{x_{\eta - 1}},{x_{\eta - 2}}, \cdots ,{x_{\eta - \varepsilon }}} \right) 。$ (10)

1.2 基于改进马氏距离的电子电路故障跟踪定位

 $d\left( h \right) = \sqrt {{{\left( {h - \bar Y} \right)}^{\rm{T}}}\sum _Y^{ - 1}\left( {h - \bar Y} \right)}。$ (11)

$\bar Y$ $\displaystyle\sum _Y^{}$ 的计算公式如下：

 $\bar Y = \frac{{\displaystyle\sum\limits_{i' = 1}^g {{Y_{i'}}} }}{g}，$ (12)
 $\sum _Y^{} = \frac{{\displaystyle\sum\limits_{i' = 1}^g {{{\left( {{Y_{i'}} - \bar Y} \right)}^2}} }}{{g - 1}} 。$ (13)

 $d\left( h \right) = \sqrt {{{\left( {h - \bar Y} \right)}^{\rm{T}}}\omega \sum _Y^{ - 1}\omega \left( {h - \bar Y} \right)} ，$ (14)

$j'$ 个故障特征样本的权值 ${\omega _{j'}}$ 为：

 ${\omega _{j'}} = {\left( {\frac{{{o_{j'}}}}{{\displaystyle\sum\limits_{j' = 1}^{N'} {{o_{j'}}} }}} \right)^2} 。$ (15)

2 实验结果分析

 图 1 延拓处理结果以及故障特征提取结果 Fig. 1 Results of extension processing and fault feature extraction

3 结　语

 [1] 朱文昌, 李振璧, 姜媛媛. 联合VMD与ISSA-ELM的电力电子电路软故障诊断[J]. 电子测量与仪器学报, 2022, 36(5): 223-233. DOI:10.13382/j.jemi.B2104972 [2] 谢李为, 李勇, 罗隆福, 等. 基于极点对称分解的多分支线路故障定位方法[J]. 中国电机工程学报, 2021, 41(21): 7326-7339. DOI:10.13334/J.0258-8013.PCSEE.201320 [3] 胡创业, 张振华, 刘东昆, 等. 核辐射探测器模拟电路混合故障智能诊断与定位[J]. 核电子学与探测技术, 2020, 40(3): 417-421. HU Chuangye, ZHANG Zhenhua, LIU Dongkun, et al. Intelligent Diagnosis and Location of Hybrid Fault in Analog Circuit of Nuclear Radiation Detector[J]. Nuclear Electronics & Detection Technology, 2020, 40(3): 417-421. DOI:10.3969/j.issn.0258-0934.2020.03.008 [4] 高校平, 黄文焘, 邰能灵, 等. 基于暂态电流的MMC-LVDC双极短路故障定位方法[J]. 电力系统自动化, 2020, 44(17): 127-135. GAO Xiaoping, HUANG Wentao, TAI Nengling, et al. Transient Current Based Fault Location Method of Pole-to-pole Short-circuit for Modular Multilevel Converter Low-voltage Direct Current[J]. Automation of Electric Power Systems, 2020, 44(17): 127-135. DOI:10.7500/AEPS20200106004 [5] 赵师兵, 张志明. 基于时域信号特征和卷积神经网络的模拟电路故障诊断算法[J]. 计算机应用, 2022, 42(S2): 320-326. ZHAO Shibing, ZHANG Zhiming. Fault diagnosis algorithm of analog circuits based on time domain signal features and convolutional neural network[J]. Journal of Computer Applications, 2022, 42(S2): 320-326. [6] 马垠飞, 王力. 融合D-S证据理论的DBN电路故障诊断算法[J]. 辽宁工程技术大学学报(自然科学版), 2021, 40(5): 448-453. DOI:10.11956/j.issn.1008-0562.2021.05.011 [7] 邓勇, 胡徐松. 基于EMD和SPS的容差模拟电路故障诊断[J]. 电子测量与仪器学报, 2020, 34(2): 67-72. DENG Yong, HU Xusong. Fault diagnosis of tolerance analog circuit based on EMD and SPS method[J]. Journal of Electronic Measurement and Instrumentation, 2020, 34(2): 67-72. DOI:10.13382/j.jemi.B1902175