|
重庆邮电大学学报(自然科学版) 2012
Analysis of transmission line icing detection and prediction based on neural network
|
Abstract:
The line icing is one main factor to threat the safety of transmission lines. Based on real-time experimental data and the research of mechanics, this paper proposes a BP neural network model for predicting the thickness and weight of ice. This model's inputs are the temperature, humidity, wind direction of the place where the transmission line lies, while its output is ice thickness, and the number of network hidden layer units and the center vector adopting the orthogonal least squares (OLS). On the basis of this model, we can send out forecasting and warning information according to expert software to analyze ice conditions. Simulation results show that it was in line with authorities expectation, which also had a certain reference value and actual value to prevent ice storm and to protect the safety of transmission lines.