硅酸盐学报

基于多尺度特征增强和时序Transformer的SiC外延生长浓度预测模型

PDF下载

张忠义 ¹^,² , 王朗 ¹^,²^,³ , 芦伟立 ¹^,²^,³ , 李帅 ¹^,²^,³ , 杨建业 ¹^,²^,³ , 高楠 ¹^,²^,³ , 王波 ¹^,²^,³ , 潘国平 ¹^,²^,³ , 房玉龙 ¹^,²^,³

硅酸盐学报 | AI for Materials专题—研究论文 2026,54(1): 35-48

收起

硅酸盐学报 |AI for Materials专题—研究论文 2026 , 54 (1) : 35 -48

基于多尺度特征增强和时序Transformer的SiC外延生长浓度预测模型

全屏

张忠义¹^,², 王朗¹^,²^,³, 芦伟立¹^,²^,³, 李帅¹^,²^,³, 杨建业¹^,²^,³, 高楠¹^,²^,³, 王波¹^,²^,³, 潘国平¹^,²^,³, 房玉龙¹^,²^,³

作者信息

^1．中国电子科技集团公司第十三研究所，石家庄 050051

^2．河北省半导体智能制造创新中心，石家庄 050051

^3．石家庄市半导体人工智能技术重点实验室，石家庄 050051

通讯作者:

房玉龙（1986—），男，博士，研究员。

作者简介:

张忠义（2000—），男，助理工程师

王朗（1989—），男，工程师。

A SiC Epitaxial Growth Doping Concentration Prediction Model Based on Multi-Scale Feature Enhancement and Temporal Transformer

Zhongyi ZHANG¹^,², Lang WANG¹^,²^,³, Weili LU¹^,²^,³, Shuai LI¹^,²^,³, Jianye YANG¹^,²^,³, Nan GAO¹^,²^,³, Bo WANG¹^,²^,³, Guoping PAN¹^,²^,³, Yulong FANG¹^,²^,³

Affiliations

^1.The 13th Research Institute, CETC, Shijiazhuang 050051, China

^2.Hebei Province Semiconductor Intelligent Manufacturing Innovation Center, Shijiazhuang 050051, China

^3.Shijiazhuang Key Laboratory of Semiconductor Artificial Intelligence Technology, Shijiazhuang 050051, China

Corresponding Author:

FANG Yulong (1986-), male, Ph.D., Professor. E-mail: yvloong@163.com

ZHANG Zhongyi (2000-), male, Assistant Engineer. E-mail: 2841622099@qq.com

WANG Lang (1989-), male, Engineer. E-mail:wangl170326@163.com

出版时间: 2026-01-20 doi: 10.14062/j.issn.0454-5648.20250794

文章导航

摘要

收起

碳化硅（SiC）外延层掺杂浓度直接决定功率器件性能，现有调控依赖人工经验与离线测试，存在成本高、滞后性强的问题。为此，提出一种多尺度特征增强时序Transformer（Multi-scale Feature-enhanced Temporal Transformer Network, MFT-Net）模型，整合多尺度卷积、压缩和激励（SE）模块、Transformer及门控循环单元（GRU）模块，构建“当炉-下炉”双场景浓度预测体系。多尺度卷积捕捉毫秒至小时级参数动态，SE强化核心特征，Transformer建模全参数耦合，GRU传递跨炉状态。基于1200炉数据实验表明，模型当炉预测相对误差低至1.35%、决定系数R²达0.89，下炉预测相对误差为1.66%，R²达到0.87，显著优于传统模型。经统计，该模型预计可降低离线测试成本约30%、提升工艺稳定性约15%，为SiC外延智能化提供支撑。

关键词

碳化硅外延生长 / 核心层掺杂浓度 / 多尺度卷积 / Transformer / 工艺参数预测 / 深度学习

Abstract

收起

Introduction

Silicon carbide (SiC), as a representative third-generation wide-bandgap semiconductor material, exhibits superior performance over traditional silicon-based devices in high-voltage, high-power, and high-frequency applications. The doping concentration of the SiC epitaxial layer critically determines the performance of power devices. However, current doping control heavily relies on manual experience and offline testing, which incurs high costs and significant time delays. This study aims to address these limitations by proposing a data-driven model for accurate and real-time prediction of doping concentration in both in-situ and next-batch scenarios.

Methods

This paper proposes a Multi-scale Feature-enhanced Temporal Transformer Network (MFT-NET) for predicting the core-layer doping concentration during SiC epitaxial growth. The model integrates four key modules:

Multi-scale Convolution Module: Employs parallel convolutional branches with kernel sizes of [3,3], [9,3], and [21,5] to capture parameter dynamics across millisecond-to-hour scales.

Squeeze-and-Excitation (SE) Module: Enhances feature channels correlated with doping concentration and suppresses noise via channel-wise attention.

Transformer Module: Models global coupling relationships among all 75 process parameters using multi-head self-attention.

Gated Recurrent Unit (GRU) Module: Captures cross-batch state dependencies to account for residual dopant and equipment state drift between consecutive epitaxial runs.

The model was trained and evaluated on a dataset comprising 1200 epitaxial runs from a commercial SiC epitaxial reactor. Data preprocessing included Pearson correlation-based feature selection and min-max normalization. A weighted mean squared error loss function was adopted to prioritize mainstream process samples.

Results and Discussion

Experimental results demonstrate that MFT-NET significantly outperforms traditional models including Ridge Regression, Support Vector Regression (SVR), CNN, CNN+LSTM, Transformer, and CNN+Transformer. In the in-situ prediction scenario, MFT-NET achieved a mean absolute percentage error (MAPE) of 1.35%, a root mean square error (RMSE) of 2.81×10¹⁴ cm^-3, and a coefficient of determination (R²) of 0.89. For next-batch prediction, it attained a MAPE of 1.66%, RMSE of 3.42×10¹⁴cm^-3, and R² of 0.87.

Ablation studies confirmed the contribution of each module: the multi-scale convolution and SE modules improved feature extraction and selection; the Transformer enabled global parameter interaction modeling; and the GRU was crucial for capturing cross-batch state transfer, reducing next-batch prediction RMSE by 20.8% compared to the model without GRU.

Hyperparameter analysis revealed that a 3-layer Transformer, the proposed multi-scale kernel combination, and a Dropout ratio of 0.2 yielded optimal performance.

Conclusions

The MFT-NET model effectively addresses the multi-scale dynamics, global parameter coupling, and cross-batch state transfer challenges in SiC epitaxial growth. It provides highly accurate doping concentration predictions for both in-situ monitoring and next-batch forecasting, with the potential to reduce offline testing costs by 30% and improve process stability by 15%. This work offers a robust data-driven framework for intelligent optimization of SiC epitaxial processes.

Key words

silicon carbide epitaxial growth / core-layer doping concentration / multi-scale convolution / Transformer / process parameter prediction / deep learning

引用本文

张忠义, 王朗, 芦伟立, 李帅, 杨建业, 高楠, 王波, 潘国平, 房玉龙. 基于多尺度特征增强和时序Transformer的SiC外延生长浓度预测模型. 硅酸盐学报, 2026 , 54 (1) : 35 -48 . DOI: 10.14062/j.issn.0454-5648.20250794

Zhongyi ZHANG, Lang WANG, Weili LU, Shuai LI, Jianye YANG, Nan GAO, Bo WANG, Guoping PAN, Yulong FANG. A SiC Epitaxial Growth Doping Concentration Prediction Model Based on Multi-Scale Feature Enhancement and Temporal Transformer[J]. Journal of the Chinese Ceramic Society, 2026 , 54 (1) : 35 -48 . DOI: 10.14062/j.issn.0454-5648.20250794

基金

收起

重点新材料研发及应用国家科技重大专项(2025ZD0619502)

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

2026年第54卷第1期

PDF下载

引用本文

BibTeX

文章信息

doi: 10.14062/j.issn.0454-5648.20250794

接收时间：2025-11-02
首发时间：2026-10-27
出版时间：2026-01-20

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2025-11-02
修回日期：2025-11-28

基金

重点新材料研发及应用国家科技重大专项(2025ZD0619502)

作者信息

^1．中国电子科技集团公司第十三研究所，石家庄 050051

^2．河北省半导体智能制造创新中心，石家庄 050051

^3．石家庄市半导体人工智能技术重点实验室，石家庄 050051

通讯作者:

房玉龙（1986—），男，博士，研究员。

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/gsyxb/CN/10.14062/j.issn.0454-5648.20250794

分享至

全文二维码

扫描看全文

引用本文

BibTeX

本文的引用情况

2种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT