pp. 1217-1230
S&M3593 Research Paper of Special Issue https://doi.org/10.18494/SAM4784 Published: March 29, 2024 Prediction Model of Residual Current Based on Grey Association and Neural Network [PDF] Guoyu Sun (Received November 30, 2023; Accepted March 6, 2024) Keywords: electrical fire warning, grey association, neural network, prediction model
To enhance early electrical fire warning in power IoT systems, we propose a residual current modeling method combining grey correlation and neural networks. By analyzing 27985 sets of data from an intelligent fire monitoring system, effective data collection and processing with advanced sensor technology in an IoT context were demonstrated. The model, derived from correlation analysis and grey prediction algorithms, uses a trained neural network for predicting residual current. This method not only augments the efficiency and accuracy of data processing in IoT but also underscores the significance of sensor technology in electrical monitoring and fire prevention. The comparative analysis of predicted and actual residual currents, showing an error range of 0.18 to 3.21%, validates the accuracy of the model and the utility of sensor-driven methods in IoT applications.
Corresponding author: Guoyu SunThis work is licensed under a Creative Commons Attribution 4.0 International License. Cite this article Guoyu Sun, Prediction Model of Residual Current Based on Grey Association and Neural Network, Sens. Mater., Vol. 36, No. 3, 2024, p. 1217-1230. |