Search

Solar cells can operate at a lower efficiency after a certain temperature, which is caused by a negative thermal coefficient. Therefore, the temperature prediction of photovoltaic (PV) modules is critical to accurately evaluate the efficiency of photovoltaic devices. We propose and experimentally demonstrate a Fuzzy Temperature Difference Threshold Method (FTDTM) based on Raman Distributed Temperature Sensor (RDTS) system for the detection and prediction of PV module temperature. The FTDTM consists of four steps, i.e., division of the universe, establishment of fuzzy relationships, definition of relationship matrix and calculation of predicted temperature. The experimental results show that the proposed RDTS can detect and accurately predict temperature trends by using the FTDTM. When the system window base of the FTDTM is set to 2, the average absolute error of the predicted temperature is 1.08 °C, and the fluctuation range of prediction error is ±3.7 °C. In addition, the experiment studied some factors affecting temperature distribution characteristics of PV modules, including solar radiation intensity, surface dust and inclination angle. We provide a solution for large-scale PV module temperature detection and early warning through RDTS systems with FTDTM.

We propose and experimental demonstrate a Dynamic Threshold Identification Method (DTIM) based on the Raman distributed fiber sensor (RDFS) system for detecting the pipeline leak. The DTIM can detect the position of pipeline leak according to distributed temperature measured by the RDFS. The DTIM includes three steps, which are the space segmentation detection, $3\sigma $ criterion and identification of historical data. The experimental results show that the proposed RDFS can detect and accurately locate the leak position by using the DTIM, and the positioning accuracy is 1 m. In addition, the temperature change rate of the leak position performs a quadratic function relationship with the leak rate, and the leak rate can be estimated by this method. It provides an effective distributed monitoring method for the leak monitoring of pipeline.

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.