基于增强支持向量机的电力隧道多状态全过程监控方法 |
点此下载全文 |
引用本文:刘滨,刘春,邵必飞,杨郭明,苟军,黄贵武.基于增强支持向量机的电力隧道多状态全过程监控方法[J].计算技术与自动化,2023,(4):41-46 |
摘要点击次数: 148 |
全文下载次数: 0 |
|
|
中文摘要:提出基于增强支持向量机的电力隧道多状态全过程监控方法,精准监控电力隧道多种状态的全过程并直观展现监控结果。通过离散小波提取电力隧道动态数据流的低频近似因数重新组合数据并提取特征,再融合低频因数,得到电力隧道多状态混合特征;深入学习支持向量机(SVM),访问未获取类别的样本,通过标有“对”与“错”分类结果的访问点来优化分类器,以此增强支持向量机;运用粒子群优化(PSO)算法优化增强支持向量机参数,得到基于PSO-SVM的电力隧道多状态识别模型,通过模型的训练和测试,识别电力隧道的多状态,实现电力隧道多状态全过程监控。实验证明:该方法能够及时、精确地监控电力隧道内各指标的多种状态,对异常情况及时预警,可实现电力隧道多状态全过程的监控。 |
中文关键词:电力隧道 支持向量机 全过程监控 多特征混合 分类器 离散小波分解 |
|
Multi-state Whole-process Monitoring Method of Power Tunnel Based on Enhanced Support Vector Machine |
|
|
Abstract:A multi-state whole-process monitoring method of power tunnel based on enhanced support vector machine is proposed to accurately monitor the whole process of multi-state power tunnel and visually display the monitoring results. The low-frequency approximate factors of the dynamic data flow of the power tunnel are extracted by discrete wavelet, the data are recombined and the features are extracted, and the low-frequency factors are fused again to obtain the multi-state mixed characteristics of the power tunnel. In-depth study of support vector machine (SVM), access to the samples of unacquired categories, through the access points marked with “true” and “false” classification results to optimize the classifier, so as to enhance the support vector machine. Particle swarm optimization (PSO) algorithm was used to optimize and enhance the parameters of support vector machine, and a power tunnel multi-state recognition model based on PSO-SVM was obtained. Through the training and testing of the model, the multi-state of power tunnel was identified, and the whole process monitoring of power tunnel multi-state was realized. Experimental results show that this method can monitor various states of various indicators in power tunnel in time and accurately, give early warning to abnormal situations in time, and realize the whole process of multi-state monitoring of power tunnel. |
keywords:power tunnel support vector machine whole-process monitoring multi-feature mixing classifier discrete wavelet decomposition |
查看全文 查看/发表评论 下载pdf阅读器 |