基于模糊支持向量机的智能站继电保护设备隐性故障检测方法 |
点此下载全文 |
引用本文:朱林.基于模糊支持向量机的智能站继电保护设备隐性故障检测方法[J].计算技术与自动化,2023,(4):53-58 |
摘要点击次数: 148 |
全文下载次数: 0 |
|
|
中文摘要:针对智能站继电保护设备运行工况不稳定,误动率和拒动率较高的问题,提出了基于模糊支持向量机的智能站继电保护设备隐性故障检测方法。采用多小波变换方法处理采集的继电保护设备信号数据噪声,将处理后的信号输入模糊支持向量机网络中,计算隐性故障样本类别隶属度函数以及训练该网络,建立模糊支持向量机模型;通过组合二类分类器,分类检测继电保护设备隐性故障;利用采用布谷鸟算法优化模型的隶属度函数和惩罚函数,提升故障检测精度。测试结果显示:隐性故障检测相对误差结果均低于0.2,检测效果较好,并且应用后保护的误动率和拒动率均低于0.3%;能够可靠完成继电保护设备隐性故障运行工况下的故障类别检测。 |
中文关键词:模糊支持向量机 智能站 继电保护设备 隐性故障检测 隶属度函数 信号降噪 |
|
Hidden Fault Detection Method of Relay Protection Equipment in Intelligent Station Based on Fuzzy Support Vector Machine |
|
|
Abstract:In order to solve the problems of unstable operating conditions, high false operation rate and high failure rate of relay protection equipment in intelligent station, a hidden fault detection method based on fuzzy support vector machine is proposed. The multi wavelet transform method is used to process the collected signal data noise of relay protection equipment, the processed signal is input into the fuzzy support vector machine network, the membership function of hidden fault samples is calculated and the network is trained, and the fuzzy support vector machine model is established. By combining two class classifiers, hidden faults of relay protection equipment are classified and detected. Using the cuckoo algorithm to optimize the membership function and penalty function of the model, the fault detection accuracy is improved. The test results show that the relative error results of hidden fault detection are lower than 0.2, the detection effect is good, and the maloperation rate and refusal rate of protection after application are lower than 0.3%. It can reliably complete the fault category detection of relay protection equipment under hidden fault operating conditions. |
keywords:fuzzy support vector machine intelligent station relay protection equipment hidden fault detection membership function signal noise reduction |
查看全文 查看/发表评论 下载pdf阅读器 |