ID 原文 译文
40936 该模型首先通过长短期记忆网络(LSTM)捕获云虚拟机性能指标的时序特征，同时利用Dropout以防止数据过拟合; Firstly, the model captures the timing characteristics of cloud virtual machine performance indicators through long short term memory LSTM), and uses Dropout to prevent data overfitting.
40937 其次利用高斯玻尔兹曼机(GRBM)得出自由能; Secondly, the free energy is obtained by using The Gaussian Boltzmann machine(GRBM).
40938 最后利用得出的自由能与训练阶段得到的参数基准模型进行对比，判断云虚拟机是否出现异常。 Finally, the free energy obtained is compared with the parameter benchmark model obtained in the training stage to judge whether there are abnormalities in the cloud virtual machine.
40939 实验表明该模型能够对云虚拟机进行异常检测，且准确率有较大的提升。 The experiment shows that the model can detect the abnormity of the cloud virtual machine, and the accuracy is improved greatly.
40940 为了降低DC/DC电源芯片过流期间的功耗，本文设计了一种可频率迁移的高精度振荡器，芯片出现过流情况时，振荡器频率可降至100 kHz以下。 In order to reduce the power consumption during the over-current of the DC/DC power supply chip, a high-precision oscillator with frequency shift is designed in this paper.When an overcurrent condition occurs in the chip, the oscillator frequency can be reduced to below 100 kHz.
40941 基于CSMC 0.25μm 30 V-60 V BCD工艺完成了电路设计、仿真。 The circuit is simulated based on CSMC 0.25 μm 30 V-60 V BCD process.
40942 该振荡器常温下可以得到信号幅值为0.6 V～3.2 V、频率为260 kHz的锯齿波。 A sawtooth wave with a signal amplitude of 0.6 V～3.2 V and a frequency of 260 kHz can be obtained at normal temperature.
40943 温度偏差在SymbolqB@4.23%以内，电源灵敏偏差在SymbolqB@1.6%以内。 Temperature deviation is within SymbolqB@4.23%, power supply sensitivity deviation is within SymbolqB@1.6%.
40944 出现工艺角偏差时，采用熔丝修调对振荡器的频率加以修正。 When the process corner deviation occurs, the frequency of the oscillator is corrected by fuse trimming.
40945 视网膜血管图像分割在疾病的自动诊断中起着重要的作用，是早期诊断和手术规划的关键步骤。 Retinal vascular image segmentation plays an important role in the automatic diagnosis of diseases, and is a key step in early diagnosis and surgical planning.