摘要
支撑电容是轨道交通列车牵引变流器中的重要部件之一,其核心作用是稳定变流器的直流电压。支撑电容劣化将导致直流母线电压波动加剧、系统稳定性降低甚至出现爆炸等风险,严重危害列车的安全运行。为了解决该问题,需定期监视电容器的运行状态,电容器状态监视可通过电容参数辨识实现,因此,提出一种基于快速递归算法的变流器支撑电容参数辨识方法。首先,分析车载电容器的电路结构和运行工况,机车电容器的预充电过程无需安装额外传感器,具有非侵入的特点,对此进行研究并构建对应的参数系统;在此基础上,引进模型快速递归的算法,研究算法的递归推导和核心参数计算,并完成算法结合预充电模型进行电容参数估计的流程和关键参数计算方法;最后进行全面的实验验证,先通过不同采样步长和频率下的不同电容器组预充电实验验证该方法的鲁棒性和识别精度,实验结果表明该方法在不同采样频率下始终保持低识别误差,展现出良好的计算精度和鲁棒性。开展了该方法与经典参数辨识的最小二乘法对比试验,结果显示所提方法具有原理简单、计算复杂度低、结果稳定和误差率低等优点。方法为机车牵引变流器系统中的电容参数估算提供了一种高效、精准的解决方案,为系统的安全稳定运行提供了技术保障。
The DC-link capacitor is a crucial component in traction converters for railway vehicles,serving to stabilize the DC voltage.Capacitor degradation can exacerbate DC bus voltage fluctuations,reduce system stability,and even lead to catastrophic failures such as explosions,posing significant threats to train safety.To address this issue,regular monitoring of capacitor health is essential,which can be achieved through parameter identification.This study proposed a fast recursive algorithm-based method for identifying parameters of traction converter DC-link capacitors.The circuit structure and operating conditions of onboard capacitors were analyzed,leveraging the non-invasive pre-charging process that required no additional sensors to establish a parameter system.A fast recursive algorithm was introduced,with detailed exploration of its recursive derivation and key parameter calculations,followed by the integration of the algorithm with the pre-charging model for accurate capacitor parameter estimation.Comprehensive experiments were conducted,including pre-charging tests on capacitor groups under varying sampling steps and frequencies,demonstrating the method’s robustness and high accuracy with consistently low identification errors.Comparative analysis with the classical least squares method highlights the proposed method’s advantages,including simplicity,low computational complexity,stability,and reduced error rates.This method can provide an efficient and precise solution for capacitor parameter estimation in traction converter systems,offering robust technical support for safe and stable railway operations.
作者
侯招文
李凯迪
曾进辉
曹俊明
罗屿
HOU Zhaowen;LI Kaidi;ZENG Jinhui;CAO Junming;LUO Yu(Rail Transit Technology Center,Zhuzhou CRRC Times Electric Co.,Ltd.,Zhuzhou 412000,China;China Shenzhen Metro Group,Shenzhen 518035,China;School of Electrical and Information Engineering,Hunan University of Technology,Zhuzhou 412000,China;China Railway Guangzhou Bureau Group Co.,Ltd.,Changsha Locomotive Depot,Changsha 410000,China)
出处
《铁道科学与工程学报》
北大核心
2025年第9期4284-4294,共11页
Journal of Railway Science and Engineering
基金
湖南省科技创新计划资助项目(2024RC3269)。
关键词
支撑电容
快速递归算法
参数辨识
牵引变流器
状态监测
DC-link capacitor
fast recursive algorithm
parameter identification
traction converter
condition monitoring
