ID 原文 译文
42246 由于 Ge 材料具有较为活泼的化学性质，易与环境中的氧化剂发生化学反应，使 Ge单晶抛光片表面质量因表面化学组成的变化而出现恶化，降低交付可靠性。 Due to the relatively active chemical properties of Ge materials, they are easy to reactwith oxidants in the environment, which makes the surface quality of the Ge single crystal polished waferdeteriorate due to the change of surface chemical composition and reduces the delivery reliability.
42247 通过实验验证了在氧化性氛围中，Ge 单晶抛光片表面易产生雾，且随着时间延长而逐渐增多，直至表面完全被覆盖。 It is verified by experiments that the Ge single crystal polished wafer surface is prone to produce haze in theoxidation atmosphere and the haze increases with time, until the whole surface is completely covered withhaze.
42248 从反应机理分析了 Ge 单晶抛光片表面可能发生的化学反应。 The possible chemical reaction taken on the Ge single crystal polished wafer surface was analyzedfrom the reaction mechanism.
42249 通过 X 射线光电子能谱 ( XPS) 验证了 Ge 单晶抛光片表面的价态发生了变化，发现 Ge 单晶抛光片表面发生了氧化反应。 It is verified by the X-ray photoelectron spectroscopy (XPS) that the valence state of the Ge single crystal polished wafer surface has changed, and it can be seen that oxidationreaction occurs on the Ge single crystal polished wafer surface.
42250 从环境因素入手，分析了净化等级和化学氛围对 Ge 单晶抛光片表面质量的影响，确定了 Ge 单晶抛光片的存放环境，保证了 Ge 单晶抛光片的交付可靠性。 The influences of the cleanliness level andchemical ambience on the surface quality of the Ge single crystal polished wafer were analyzed, and the storage environment of the Ge single crystal polished wafer was determined to guarantee the delivery reliability of the Ge single crystal polished wafer.
42251 针对 IGBT 芯片被封装在模块内部，芯片结温无法直接测量的问题，提出了基于思维进化算法 ( MEA) 优化的反向传播 ( BP) ( MEA-BP) 神经网络算法的 IGBT 结温预测算法模型。 To solve the problem that the IGBT chip is encapsulated in the module, so that the chipjunction temperature cannot be measured directly, an IGBT junction temperature prediction algorithmmodel based on the mind evolutionary algorithm (MEA) optimized back-propagation (BP) (MEA-BP)neural network algorithm was proposed.
42252 首先，利用温敏电参数 ( TSEP) 法搭建 IGBT 模块饱和压降实验平台; Firstly, the saturation voltage drop experiment platform of theIGBT module was built by using the temperature sensitive electrical parameters (TSEPs) method.
42253 然后，从实验数据中提取 338 组饱和压降与集电极电流数据作为 TSEP，表征其与 IGBT 模块结温的关系; Then, 338 groups of saturation voltage drop and collector current data were extracted from the experimental dataas TSEPs to characterize their relationship with the junction temperature of the IGBT module.
42254 最后，利用MEA-BP 神经网络算法将提取出的电气参数建立结温预测模型，对结温进行预测。 Finally, thejunction temperature prediction model was established with the extracted electrical parameters by usingthe MEA-BP neural network algorithm to predict the junction temperature.
42255 实验结果表明，MEA-BP 神经网络算法的结温预测值平均绝对百分比误差在集电极电流小于临界电流时为0.114，在大于临界电流时为 0. 062， The test results show that theaverage absolute percent error of the junction temperature of the MEA-BP neural network algorithm is0.114 when the collector current is less than the critical current and 0.062 when it is greater than thecritical current.