Skip to Main Content
Fault detection and classification (FDC) has been recognized as an integral component of the advanced process control (APC) framework in the semiconductor industry, as it helps to improve overall equipment efficiency (OEE). However, some unique characteristics of semiconductor manufacturing processes have posed challenges for FDC applications, such as nonlinearity in most batch processes, and multimodal batch trajectories due to product mix. To explicitly account for these unique characteristics, a pattern recognition based fault detection method utilizing the k-nearest-neighbor rule (FD-kNN) was previously developed. In FD-kNN, historical data are used directly as the reference of normal process operation to determine whether a new measurement is a fault. Therefore, for processes with a large number of variables, it can be computation and storage intensive, and may be difficult for online process monitoring. To address this difficulty, we propose a fast pattern recognition based fault detection method, termed principal component-based kNN (PC-kNN), which takes advantages of both principal component analysis (PCA) for dimensionality reduction and FD-kNN for nonlinearity and multimode handling. Two simulation examples and an industrial example are used to demonstrate the performance of the proposed PC-kNN method in fault detection.