Research On Transmission Line Icing Laws And Forecasting Model

Overhead transmission line icing often causes the line over load, conductorgalloping and other hazards, which resulting the breakdown of electrical tower or thegrid failure. The transmission line icing has a serious impact on the reliability of thetransmission grid and the security and stability of the power system. The ice coatingwas mainly affected by micro-meteorological conditions, micro-topographyconditions and climate. The laws and the prediction model of the transmission lineicing were discussed from these three aspects in this paper.Micro-meteorological conditions including temperature, humidity, rainfall and soon, which played a vital role in the transmission line icing. Differentmicro-meteorological conditions were simulated in the artificial climate chamber, andthe growth laws of different types of wire icing in these conditions were got,meanwhile the difference of wire charged icing and uncharged icing was compared,and different types of wire icing models were establish.In Hunan region has a lot of complex mountainous and hilly terrain, thesemicro-meteorological environment have a significant impact on the transmission lineicing. The areas with typical micro-topographical features were selected formonitoring, summarizing the law of transmission line icing under the influence ofdifferent micro-topography, and the corresponding mechanism has explained.With the abnormal climate change and the frequent occurrence of extremeweather, geographical location of Hunan makes severe weather like thelow-temperature freezing rain break out easier than other places in winter, this glazeweather will cause transmission lines of Hunan power grid got large-scale ice coatingphenomenon. Based on the historical data of Hunan power grid ice coating andatmosphere data in the winter before, large-scale transmission line icing in winter ofHunan power grid stepwise regression prediction model and support vector machineprediction model was established by correlation analysis, stepwise regression, fuzzypattern recognition methods, artificial neural networks and particle swarmoptimization.The study of transmission line icing under micro-meteorological conditions andmicro-topographic conditions has significant meanings to the reason of transmissionline getting ice coating in different areas of Hunan. Compared to traditional stepwise regression model the model base on the fuzzy pattern recognition methods andsupport vector machine has the better capability to catch the complex relationshipbetween the prediction object with predictor, which make the prediction result oftransmission lines of Hunan power grid got large-scale ice coating phenomenon betterand more accurate.