基于MSSA-SVM的电缆隧道故障预警系统设计  

Design of cable tunnel fault warning system based on MSSA-SVM

作　　者：纪超 王亮 王孝敬 李小兵 曹雯[1] JI Chao;WANG Liang;WANG Xiao-jing;LI Xiao-bing;CAO Wen(School of Electronic Information,Xi'an Polytechnic University,Xi'an 710600,China;Xi'an Jinpower Electrical Co.,Ltd.,Xi'an 710075,China)

机构地区：[1]西安工程大学电子信息学院,陕西西安710600 [2]西安金源电气股份有限公司,陕西西安710075

出　　处：《工程设计学报》2023年第1期109-116,共8页Chinese Journal of Engineering Design

基　　金：国家自然科学基金资助项目(51707141);西安市科技计划项目(22GXFW0041);金属挤压与锻造装备技术国家重点实验室开放课题(S2208100.W03)。

摘　　要：为了实现电缆隧道环境的在线监测和故障报警,提高电缆隧道监测系统的智能化水平,提出了一种基于多特征麻雀搜索算法(multi-feature modified sparrow search algorithm, MSSA)优化支持向量机(support vector machines, SVM)的故障预警系统。首先,对故障数据集进行归一化预处理;其次,建立多分类SVM模型,用MSSA对SVM进行参数寻优,从而建立MSSA-SVM模型,并将训练好的MSSA-SVM模型嵌入故障预警系统的数据库服务器中,对实时采集的数据进行在线监测、诊断,并及时报警;最后,通过实验验证了MSSA-SVM模型的有效性,并将其与麻雀搜索算法(sparrow search algorithm, SSA)、灰狼优化算法(grey wolf optimization, GWO)和粒子群算法(particle swarm optimization, PSO)进行对照实验,实验结果表明,MSSA-SVM模型的故障识别准确率最高,其识别准确率可达95%。研究结果为有效提高电缆隧道在线监测的智能性和准确性提供了参考。In order to realize online monitoring and fault alarm of cable tunnel environment and improve the intelligent level of cable tunnel monitoring system, a fault warning system based on multi-feature sparrow search algorithm(MSSA) optimized support vector machines(SVM) was proposed. Firstly, the fault data set was preprocessed normalized;secondly, a multi-class SVM model was established, and MSSA was used to optimize the parameters of the SVM, so as to establish the MSSA-SVM model. The trained MSSA-SVM model was embedded in the database server of the fault warning system, and the real-time collected data was monitored and diagnosed online, and the alarm was given in time;finally,the effectiveness of MSSA-SVM model was verified by experiments, and it was compared with sparrow search algorithm(SSA), grey wolf optimization(GWO) and particle swarm optimization(PSO). The experimental results showed that MSSA-SVM model has the highest fault recognition accuracy, and its recognition accuracy could reach 95%. The research result provides a reference for effectively improving the intelligence and accuracy of online monitoring of cable tunnels.

关 键 词：电缆隧道 监测系统 支持向量机 故障诊断 多特征麻雀搜索算法 

分 类 号：TM712[电气工程—电力系统及自动化]