利用consensus滤波器诊断HVDC系统故障

Fault Diagnosis for HVDC System via Consensus Filter

高压直流输电(HVDC)系统对运行时的稳定性要求比较严格,出现故障时应能及时分辨故障类型并快速恢复。传统的利用神经网络诊断HVDC故障一般都是将电压电流信号输入网路,没有实际测量过程中随机噪声的干扰。为此,针对长输电线路中实际测量的直流电压信号易引入随机噪声干扰的特点,提出了一种分布式故障诊断算法并研究了consensus滤波器在滤除直流电压信号噪声中的应用。最后的仿真结果表明,consensus滤波器可有效滤除测量噪声,从而可有效检测出HVDC系统中的故障。

Different from the traditional neural-network-based fault diagnosis approach, the authors propose a novel fault diagnosis scheme based on the consensus filter, which takes into account the sensor＇s measurements noise. Firstly, the basic theories and theorems about consensus filter are introduced briefly. For the convenient of study, a mathematical model of HVDC system is established by setting some appropriate parameters according to the first benchmark model under PSCAD（Power system CAD）. Accordingly, we use consensus filter to filter d.c. voltage signals which are corrupted by random noise. Then, a fault detection filter is constructed by using the outputs of consensus filter to detect the HVDC system faults in the current DC transmission line. The residuals are generated using the outputs of the fault detection filter and real HVDC system de voltages. According to the relationship between the generated residuals and selected threshold, the HVDC system fault can be detected effectively. Finally, different kinds of faults are simulated by using the PSCAD software. Simulation results are provided to show the efficiency of the proposed approach.