Research On Prediction Technology Of Transmission Line Icing Based On Micro-meteorology

The icing on the transmission line is extremely destructive to the safe operation of the transmission network.The bad micro-meteorological environment is the main reason for the formation of icing on the transmission line,and the micro-meteorological environment is unchanged in the local area.Therefore,this thesis predicts the transmission line icing thickness from the micro-meteorological point of view.In the natural state,the icing on the transmission line is in a non-linear growth state,and the micro-meteorological factors are numerous.It is difficult to predict the short-term thickness of icing by a simple fitting function.In view of the above problems,this thesis introduces the BP neural network with strong nonlinear approximation ability,strong generalization ability and strong adaptive ability to predict the short-term transmission line icing thickness.BP neural network is easy to fall into the local optimization,over-fitting,low stability,and other shortcomings,and mind evolutionary algorithm(MEA)has a strong global search ability.So using the mind evolutionary algorithm to optimize the BP neural network,establish a MEABP neural network model based on micro-meteorology.Micro-meteorological influencing factors such as temperature,humidity,wind speed,wind direction,atmospheric pressure,and light intensity are used as input vectors to train a high-precision short-term icing thickness prediction model.The GABP neural network prediction model is established for comparison.The experimental results show the MEABP short-term prediction model can predict the icing thickness more accurate and stable.Taking into account the complex topography and geography of China,large water areas such as lakes,rivers,and reservoirs within 200 m have an impact on the surrounding air humidity.Therefore,this thesis establishes an icing thickness correction model and uses a geographic location correction value to correct the predicted value of the transmission line ice thickness.Combining the MEABP neural network model with the geographic location correction model,the experimental results show that the geographically corrected short-term prediction model based on micro-meteorology information has higher prediction accuracy and wider application range.Finally,based on the well-trained icing thickness prediction model,a short-term prediction system for icing thickness of transmission lines based on micro-meteorology is designed and implemented.