PIN diode temperature characteristics prediction based on variational mode decomposition and autoencoder
-
摘要: 提出融合变分模态分解（VMD）和自编码器的预测方法，将温升特性曲线分解成若干个子信号分量，其中包含高频的波动量、中间量和低频的趋势量，然后利用自编码器对每个分量进行预测，最后将分量的预测值相加，从而实现对PIN二极管温升特性曲线的精准预测。通过与多种机器学习方法的对比验证了结合VMD分解可有效提升预测精度，同时也验证了自编码器在特性曲线拟合上的优势。Abstract: PIN diodes are critical devices for preventing damage from strong electromagnetic signals. Accurately predicting the temperature rise curve of the PIN diode has important guiding significance for selecting protective devices. Machine learning-based methods can effectively predict the characteristics of devices. However, the temperature rise characteristic curve of the PIN diode contains strong nonlinearity and small fluctuations, and traditional machine learning methods cannot predict accurately. To accurately predict the temperature rise characteristic curve of PIN diodes, this paper proposes a prediction method that combines variational mode decomposition (VMD) and autoencoder to decompose the temperature rise characteristics into sub-signals, which include high-frequency fluctuations, intermediate quantities, and low-frequency trend quantities. Then an autoencoder is used to predict each component. Finally, the predicted values of the components are added together, so as to accurately predict the temperature rise characteristic curve of the PIN diode. By comparing with various machine learning methods, it is verified that combining VMD decomposition can effectively improve prediction accuracy, and the advantage of autoencoder in fitting characteristic curves is also verified.
-
图 1 结合VMD分解和自动编码器的算法流程图
Figure 1. Flowchart of the proposed algorithm by combining decomposition and autoencoder
图 4 自编码器采用及没有采用VMD分解预测结果对比
Figure 4. Predicted curves comparison with and without VMD decomposition
表 1 实验结果
Table 1. Results of experiments
network possibility of R2 ＞ 0.9 without VMD with VMD autoencoder 0.76 0.91 1DCNN 0.73 0.80 SVM 0.73 0.75 MLP 0.68 0.73 GRP 0.82 0.86 
下载: 导出CSV
-
[1] 袁月乾, 陈自东, 马弘舸, 等. PIN限幅器的高功率微波单脉冲效应研究[J]. 强激光与粒子束, 2020, 32:063003 doi: 10.11884/HPLPB202032.190174Yuan Yueqian, Chen Zidong, Ma Hongge, et al. High power microwave effect of PIN limiter induced by single pulse[J]. High Power Laser and Particle Beams, 2020, 32: 063003 doi: 10.11884/HPLPB202032.190174 [2] 王明, 马弘舸. 组合脉冲内间隔对限幅器热损伤效应的影响[J]. 强激光与粒子束, 2018, 30:063002 doi: 10.11884/HPLPB201830.170426Wang Ming, Ma Hongge. Influence of pulse interval on thermal damage process of PIN limiter[J]. High Power Laser and Particle Beams, 2018, 30: 063002 doi: 10.11884/HPLPB201830.170426 [3] Bera S C, Bharadhwaj P S. Insight into PIN diode behaviour leads to improved control circuit[J]. IEEE Transactions on Circuits and Systems II:Express Briefs, 2005, 52(1): 1-4. doi: 10.1109/TCSII.2004.839537 [4] 张永战, 孟凡宝, 赵刚. Ⅰ层厚度对限幅器热损伤效应的影响[J]. 强激光与粒子束, 2017, 29:093002 doi: 10.11884/HPLPB201729.170087Zhang Yongzhan, Meng Fanbao, Zhao Gang. Influence of I layer thickness on thermal damage process of PIN limiter[J]. High Power Laser and Particle Beams, 2017, 29: 093002 doi: 10.11884/HPLPB201729.170087 [5] 赵振国, 周海京, 马弘舸, 等. 微波脉冲频率与重频对限幅器热损伤效应的影响[J]. 强激光与粒子束, 2015, 27:103239 doi: 10.11884/HPLPB201527.103239Zhao Zhenguo, Zhou Haijing, Ma Hongge, et al. Influence of frequency and microwave repetition rate on thermal damage process of PIN limiter[J]. High Power Laser and Particle Beams, 2015, 27: 103239 doi: 10.11884/HPLPB201527.103239 [6] Ko K, Lee J K, Kang M, et al. Prediction of process variation effect for ultrascaled GAA vertical FET devices using a machine learning approach[J]. IEEE Transactions on Electron Devices, 2019, 66(10): 4474-4477. doi: 10.1109/TED.2019.2937786 [7] Liang Wei, Yang Xuejiao, Miao Meng, et al. Novel ESD compact modeling methodology using machine learning techniques for snapback and non-snapback ESD devices[J]. IEEE Transactions on Device and Materials Reliability, 2021, 21(4): 455-464. doi: 10.1109/TDMR.2021.3116599 [8] Wang Jing, Kim Y H, Ryu J, et al. Artificial neural network-based compact modeling methodology for advanced transistors[J]. IEEE Transactions on Electron Devices, 2021, 68(3): 1318-1325. doi: 10.1109/TED.2020.3048918 [9] Yang Qihang, Qi Guodong, Gan Weizhuo, et al. Transistor compact model based on multigradient neural network and its application in SPICE circuit simulations for gate-all-around Si Cold source FETs[J]. IEEE Transactions on Electron Devices, 2021, 68(9): 4181-4188. doi: 10.1109/TED.2021.3093376 [10] Mehta K, Wong H Y. Prediction of FinFET current-voltage and capacitance-voltage curves using machine learning with Autoencoder[J]. IEEE Electron Device Letters, 2021, 42(2): 136-139. doi: 10.1109/LED.2020.3045064 -
点击查看大图
图(4) / 表(1)
计量
- 文章访问数: 89
- HTML全文浏览量: 47
- PDF下载量: 37
- 被引次数: 0
