Dynamic Signal Adaptive Reconstruction and Multitask Knowledge Distillation for Industrial Surveillance of Hazardous Chemicals Using Gas Sensor

This is the authors' abstract. We don't add key points for this paper.

Ensuring safety in hazardous chemical production relies heavily on robust gas monitoring systems for industrial reliability. Integrating dynamic temperature modulation signal processing and artificial intelligence with semiconductor gas sensors advances gas sensing capabilities significantly. Challenges include manual signal parameterization for various gases and complex network deployment on resource-limited devices. To tackle these, we propose the lightweight and precise dynamic signal adaptive reconstruction multitask knowledge distillation network teacher–student (DSAR-MTKDNetT-S. DSAR enhances response signals from individual gas sensors through adaptive reconstruction, while multitask knowledge distillation compresses an long short term memory networks (LSTM)-based teacher model (MTKDNetT) into a streamlined 1D-CNN-based student model (MTKDNetS), boosting the student model's prediction accuracy. Experimental findings on datasets including toluene, acetone, and acetic anhydride demonstrate DSAR-MTKDNetT-S’s superior performance over existing methods. This demonstrates the superior practicality and versatility of the approach in enhancing gas monitoring for hazardous chemical production environments.