基于预充电模型与低频采样的直流支撑电容器电容辨识方法  

Pre-charging model based capacitance estimation method for DClink capacitors with low sampling frequency

作　　者：伍珣 田睿 李凯迪 于天剑[1] 成庶[1] 陈春阳[1] WU Xun;TIAN Rui;LI Kaidi;YU Tianjian;CHENG Shu;CHEN Chunyang(School of Traffic&Transportation Engineering,Central South University,Changsha 410075,China;State Grid Hunan Extra High Voltage Substation Company,Changsha 410004,China;Shenzhen Metro Operation Group Co.Ltd.,Shenzhen 518040,China)

机构地区：[1]中南大学交通运输工程学院,湖南长沙410075 [2]国网湖南省超高压变电公司,湖南长沙410004 [3]深圳地铁集团,广东深圳518040

出　　处：《中南大学学报（自然科学版）》2023年第8期3325-3334,共10页Journal of Central South University:Science and Technology

基　　金：国家自然科学基金资助项目(52072414)。

摘　　要：直流支撑电容器是列车牵引变流器的重要组成部分,主要用于整流器的输出电压滤波与稳压,并保证逆变器的高效运行。电容是衡量电容器剩余使用寿命的重要指标。在恶劣的工作环境中,电容的减小速率往往比预期更快,准确、有效地对电容进行估计是直流电容器状态监测的必要条件。然而,在轨道交通应用中,中间直流环节电压传感器的信号噪声波动幅值与纹波分量波动幅值几乎相同,导致现有电容器状态辨识方法对电容的辨识误差明显上升。针对此问题,提出一种基于预充电模型与低频采样的直流电容器电容估计方法。首先,对预充电过程进行分析,建立离散数学模型以减少噪声的影响。在此模型基础上,推导电容器电容的计算公式,采用带噪声估计的RELS算法进一步减小噪声影响并获得模型参数的准确估值,进而对直流支撑电容器电容进行精确辨识。最后,在dSPACE平台上验证该方法的快速性和鲁棒性。研究结果表明:该方法可以在较低的采样频率下实现电容的准确计算;与现有估计方法相比,该方法在地铁等轨道交通场合具有较大优越性。DC-link capacitor is a significant part of traction converter in railway applications.It mainly filters the output voltage of rectifier and ensures the efficient operation of inverter.The capacitance is an important indicator for the residual service life of the capacitors,and it tends to degrade fast in harsh working environment.Accurate capacitance estimation is necessary for the condition monitoring of DC-link capacitors.However,the noise fluctuation of voltage sensor signals is nearly the same as that of the ripple components in railway applications,which leads to the increase of errors in existing methods.Therefore,a pre-charging model based capacitance estimation method for DC-link capacitors was proposed.The pre-charging process was analyzed and a discrete model was built to reduce the impact of noise.Based on the model,RELS algorithm with noise evaluation was utilized to further lessen the noise effect and obtain accurate capacitance estimation results.The accuracy and robustness of the proposed method were verified on a dSPACE platform.The results show that the accurate calcutation can be achieved with a low sampling frequency.The comparison results with existing estimation methods show the superiority of the proposed method.

关 键 词：电容估计 直流支撑电容器 预充电 噪声 RELS算法 

分 类 号：TM464[电气工程—电器]