摘要
电压源换流器(VSC)是电压源换流器高压直流输电(VSC-HVDC)系统中的重要组成部分,其过压和过流能力差,容易发生各种故障,且受传输功率影响故障信号特征提取复杂、故障诊断困难。针对此问题,基于所建立的VSC-HVDC系统的PSCAD/EMTDC仿真模型,首先分析了系统在发生各种故障时所输出直流电压的基本特性,在充分考虑传输功率影响的条件下,根据直流电压故障信号幅值波动的范围,判断系统故障的性质,然后再通过小波分析方法提取特征向量,并结合人工神经网络方法实现系统故障的识别。仿真结果表明,这种方法能对VSC-HVDC的故障进行有效诊断和识别,且准确度不受传输功率的影响,具有良好的应用前景。
As one of important components in the VSC-HVDC system, voltage source converter（VSC） generates various faults easily because of its poor over voltage/current capacity. Furthermore, it is complicated to extract the feature of faulty signal and difficult to carry on the fault diagnosis. On the basis of the PSCAD/EMTDC simulation model of VSC-HVDC system, we analyzed the essential characteristics of output DC voltage. Then, fully considering the influence of transmission power, we analyzed the characters of system fault according to the amplitude＇s fluctuation range of DC voltage faulty signal. Moreover, the wavelet analysis method was adopted to extract the eigenvector of faulty signal, and combined with method of artificial neural network, the system fault was identified. The simulation results show that the method can realize the diagnosis and discrimination on the VSC- HVDC faults. Moreover, its accuracy is independent of the transmission power. Therefore, the method has a good application prospect.
出处
《高电压技术》
EI
CAS
CSCD
北大核心
2012年第6期1383-1390,共8页
High Voltage Engineering
基金
陕西省科技计划项目(2010k09-09)
陕西省教育厅科研计划项目(11JK1012)~~
