摘要
为准确提取ZPW-2000A型无绝缘轨道电路调谐区和补偿电容故障特征信息,提出一种以包络熵为优化目标的遗传算法(Genetic Algorithm, GA)优化变分模态分解(Variational Mode Decomposition, VMD)的故障特征提取方法 .首先,采用传输线理论建立无绝缘轨道电路模型,仿真分析调谐区和补偿电容不同故障的分路电流曲线.其次,以VMD方法中的模态分量的最小包络熵为目标函数,通过遗传算法搜索目标函数的最小值来确定最佳组合参数,原始故障信号经最佳参数组合下的VMD分解成若干个本征模态分量,并对算法的有效性进行了验证.最后,通过与总体经验模态分解(Ensemble Empirical Mode Decomposition, EEMD)、传统VMD方法相对比.研究结果表明:GAVMD方法具有更强的故障信息提取能力,避免了EEMD方法出现的模态混叠问题和传统VMD方法在分解信号过程中出现的过分解、欠分解问题,GA-VMD方法实现了复杂信号的准确提取.
To accurately extract the fault feature information from the tuning area and compensation ca⁃pacitor of ZPW-2000A track circuit,a fault feature extraction method is proposed based on optimiza⁃tion Variational Mode Decomposition(VMD)of Genetic Algorithm(GA)with envelope entropy as the optimization objective.First,a jointless track circuit model is established using the transmission line theory,and the shunt current curves of different faults in the tuning area and compensation capacitor are simulated and analyzed.Second,the minimum envelope entropy of modal component in the VMD method is taken as the objective function,and the optimal combination parameters are determined through genetic algorithm by searching the minimum value of the objective function.The original fault signal is decomposed into several intrinsic modal components by the VMD under the optimal combina⁃tion to verify the effectiveness of the algorithm.Finally,a comparison is made between the proposed method,the overall Empirical Mode Decomposition(EEMD),and the traditional VMD approach.The research results show that the GA-VMD method proposed in this paper has stronger ability to ex⁃tract fault information.It avoids the mode aliasing problem of the EEMD method and the overdecomposition and under-decomposition problems of the traditional VMD method during signal de⁃composition,achieving the accurate extraction of complex signals.
作者
焦梅梅
石磊
刘雅芝
JIAO Meimei;SHI Lei;LIU Yazhi(School of automation and electrical engineering,Lanzhou Jiaotong University,Lanzhou 730070)
出处
《北京交通大学学报》
CAS
CSCD
北大核心
2023年第3期149-158,共10页
JOURNAL OF BEIJING JIAOTONG UNIVERSITY
基金
甘肃省自然科学基金(21JR7RA292)。
