Proper

Abstract

Three statistical methods, Generalized Additive Model (GAM), Generalized Linear Model (GLM) and Linear Mixed Effects Model (LME) are used to analyze the relationship between the electric pole vibration and the weather conditions. All the models were fitted individually to the respective weather conditions such as temperature, humidity, wind speed and wind direction. All the information from the sensors are processed and analyzed, where the pitch and the roll of the electric pole reveals the influence of the temperature over the respective data. Therefore, the model is fitted with the respect to the weather conditions obtained from different source and platform. In order to fit the model accurately, all three models implemented to pitch and roll, along with the weather conditions. The result shows that the best model among the three is Generalized Additive Model, which is identified using AIC value, BIC value and the deviance explained. For more deep understanding and clearance, the residual fit is performed and the model validation is tested for normality using the Kolmogorov-Smirnov normality test. With the best-fit model, the risk assessment becomes more reliable, with either, minor or major causalities.