Long-Range Raman Distributed Fiber Temperature Sensor With Early Warning Model for Fire Detection and Prevention

In the application of the fire prevention, the sensor is expected to alarm before the occurrence of fire. We propose and experimentally demonstrate a long-range Raman distributed fiber temperature sensor (RDFTS) and a temperature early warning model (TEWM) for fire detection and prevention. The TWEM includes first-order and second-order moving average method. And the second-order moving average method is suitable for prediction based on trends with significant temperature changes. The RDFTS achieved a spatial resolution of 3.0 and 28.6 m for the sensing range of 1.38 and 28.9 km, respectively. In experiment, the sensing fiber is placed in the temperature-controlled chamber and outdoor building (Boxue Building, Taiyuan University of Technology, China) for detecting and predicting the temperature. The experimental results show that the TEWM can predict the temperature trend in advance of 46 s. Simultaneously, the mean absolute errors of prediction temperature are 0.33 °C (temperature-controlled chamber) and 0.31 °C (Boxue Building). The experimental results are in accordance with the China-National Standard System for fire detection and prevention.