Shuaijie Zhao 교수

The need for power modules has been promoted by the emergence of electric vehicles. The requirements of small volume, high integrity, and high reliability for the next-generation power module lead to the change of encapsulation method from previous gel encapsulation to epoxy encapsulation. Efforts have been made to enhance epoxy materials, and commercialized high-end epoxy has shown excellent properties, but power module failures still exist. The possible reason may lie in the interfacial intera

High reliability is critical for semiconductors, especially power semiconductors, which usually work in high temperature conditions. One important reliability issue for power semiconductors is the copper diffusion into polymers. Copper diffusion accelerated polymer decomposition. Also, severe copper diffusion even causes short circuits. Although copper diffusion has been identified, the reason for the initiation of copper diffusion is still not well understood. Here, we encapsulated copper subst

Gel encapsulation is gradually replaced by epoxy encapsulation for power modules due to the miniature of epoxy encapsulation. The epoxy and the copper substrate form a dissimilar bond. Generally, the mismatch of the coefficient of thermal expansion (CTE) is thought as the reason for the bond failure. Here, we investigated the influence of copper/epoxy interface interaction on reliability under various reliability tests. It shows that besides the mismatch of CTE, the interfacial interaction also

Driver fatigue detection is one of the crucial methods to ensure driving safety. This paper proposes a driver fatigue detection model through multimodal fusion based on convolutional neural networks (CNN) and gated transformer networks (GTN). Specifically, the driver’s facial features are extracted using CNN and combined with the driver’s electrocardiogram (ECG) and vehicle state features to form multivariate time series. Then, the GTN employs self-attention and masking mechanisms to capture the