Research on fault detection of fire monitoring sensor based on multi-strategy improved PCA method

Artificial intelligence (AI) technology has significant potential in the field of industrial sensor fault detection, particularly in critical fire monitoring scenarios. To address the performance degradation of Principal Component Analysis (PCA) under complex conditions due to the neglect of data distribution characteristics and noise, the study innovatively proposes a multi-strategy fault detection method that integrates Density-Based Spatial Clustering of Applications with Noise (DBSCAN), the Savitzky-Golay Smoothing Algorithm (SG), and an improved PCA. This approach aims to enhance the detection accuracy and robustness of fire sensors. Experimental results show that this method achieves a true positive rate of 96% and a false positive rate of only 3% in both real and simulated datasets, with a minimum detectable bias of 0.5℃, significantly outperforming traditional PCA and single-improvement methods. This research provides a high-precision, low-bias fault detection solution for fire monitoring.