Deep level donor＇s ionization behavior of passive film formed on the surface of stainless steel was investigated by Mott-Schottky plots. It is indicated that transformation process of deep level donors＇ ionization b...Deep level donor＇s ionization behavior of passive film formed on the surface of stainless steel was investigated by Mott-Schottky plots. It is indicated that transformation process of deep level donors＇ ionization behavior of passive film on surface of stainless steel can be divided into 4 stages with rising immersion time. At the initial immersion stage （10 min）, Fe（II） located in the octahedral sites of the unit cell is not ionized and the deep level does not appear in Mott-Schottky plots. At the second stage （9-38 h）, Fe（II） located in the octahedral sites starts to be ionized, which results in deep level donors＇ generation and density of deep level donors almost is constant with augmenting immersion time but the thickness of space charge layer is more and more thicker with rising immersion time. At the third stage （48 h-12 d）, density of deep level donors rises with increasing immersion time and the thickness of passive films space charge layer decreases. At last stage （above 23 d）, both the space charge layer＇s thickness and density of deep level donors are no longer changed with increasing immersion time. In the overall immersion stage, the shallow level donors＇ density is invariable all the time. The mechanism of deep level donor＇s ionization can be the generation of metal vacancies, which results in crystal lattice＇s aberration and the aberration energy urges the ionization of Fe（ II ） in octahedral sites.展开更多
中点钳位(neutral point clamped,NPC)型三电平逆变器并网工作环境恶劣,IGBT面临单管与双管同时故障的挑战,这使得故障特征之间的差异变得非常微弱,进而导致双管故障的识别精度难以有效提升。为此,提出了一种新的故障诊断方法,该方法结...中点钳位(neutral point clamped,NPC)型三电平逆变器并网工作环境恶劣,IGBT面临单管与双管同时故障的挑战,这使得故障特征之间的差异变得非常微弱,进而导致双管故障的识别精度难以有效提升。为此,提出了一种新的故障诊断方法,该方法结合了多通道的二维递归融合图和轻量化多尺度残差(lightweightmultiscale convolutional residuals,LMCR)网络。首先,通过仿真获取三相电流信号作为故障信号;再利用递归图(recurrence plot,RP)将三相电流信号分别转化为二维图并进行多通道融合,以捕捉时间序列中的周期性、突变点和趋势等特征;最后,将递归融合图作为输入,输入到LMCR模型中进行故障识别,LMCR模型整合多级Inception结构和残差网络,用于提取不同尺度的特征并融合这些特征,从而保证网络的梯度消失和爆炸。实验结果显示,该方法在IGBT故障识别中表现出色,无噪声环境下平均识别准确率达100%,噪声环境中也达到了92.53%,充分证明了该方法具有较强的特征提取能力和优异的抗噪性能。展开更多
基金Foundation item: Projects(50571059, 50615024 ) supported by the National Natural Science Foundation of ChinaProject(NCET-07-0536) supported by Program for New Century Excellent Talents in UniversityProject(IRT0739) supported by Program for Innovative Research Team in University
文摘Deep level donor＇s ionization behavior of passive film formed on the surface of stainless steel was investigated by Mott-Schottky plots. It is indicated that transformation process of deep level donors＇ ionization behavior of passive film on surface of stainless steel can be divided into 4 stages with rising immersion time. At the initial immersion stage （10 min）, Fe（II） located in the octahedral sites of the unit cell is not ionized and the deep level does not appear in Mott-Schottky plots. At the second stage （9-38 h）, Fe（II） located in the octahedral sites starts to be ionized, which results in deep level donors＇ generation and density of deep level donors almost is constant with augmenting immersion time but the thickness of space charge layer is more and more thicker with rising immersion time. At the third stage （48 h-12 d）, density of deep level donors rises with increasing immersion time and the thickness of passive films space charge layer decreases. At last stage （above 23 d）, both the space charge layer＇s thickness and density of deep level donors are no longer changed with increasing immersion time. In the overall immersion stage, the shallow level donors＇ density is invariable all the time. The mechanism of deep level donor＇s ionization can be the generation of metal vacancies, which results in crystal lattice＇s aberration and the aberration energy urges the ionization of Fe（ II ） in octahedral sites.
