Research On The Prediction Of Trans Mission Line Icing And Ice Melting Technology

Because of the variability and uncontrollability of the natural environment,the safe and stable operation of the transmission line is challenged.Icing fault of transmission line caused by ice weather is one of them.In case of serious icing,tower will have a series of faults such as flashover,disconnection and collapse,which will bring unnecessary economic loss to the society.However,at present,the research on the formation process and prediction model of transmission line icing is not perfect,and the prediction accuracy of the model needs to be improved.Therefore,it is urgent to study a prediction method to accurately grasp the ice fault thickness of transmission lines.The main research work of this paper is as follows:(1)Based on the analysis and comparison of the classical calculation models of the ice thickness of the conductor,and considering various factors of the ice thickness of the line,the theoretical calculation model of the equivalent ice thickness of the line is derived from the ice type of the line and the classical equation.Using the historical ice cover data of the line,a single series ice cover thickness prediction model is established based on LSTM(long and short-term memory neural network),and the accuracy of the prediction model is verified by the actual ice cover data of the line.(2)In order to determine the influence of micro meteorological factors on the line icing,grey relation analysis(GRA)is carried out one by one for the natural conditions such as ambient temperature,humidity,wind speed and atmospheric pressure that may affect the ice thickness of transmission line.It is verified that the influence of ambient temperature,atmospheric pressure,ambient wind speed and ambient humidity on the ice thickness of the line is in order from large to small,and the rationality of using it as the input variable of the model is verified.(3)In view of the problems of single sequence prediction model,the loss function and network model are improved.A multi variable olstm(improvedlong and short-term memory neural network)prediction model is established based on a variety of meteorological factors,and the accuracy of the model is verified by the actual line icing data.The experimental results show that the method proposed in this paper has higher prediction accuracy and smaller error than other traditional methods,and can meet the actual engineering needs.(4)Based on the SVG reactive power device,a model of ice melting by controlling the reactive power transmitted by the line is built.By adjusting the line load,using SVG device to provide the corresponding reactive power compensation,increasing the current of the ice melting line,and realizing the ice melting function.And through MATLAB / Simulink software,the device is simulated.The results show that the device can track the change of reactive current in time and provide the current value needed for line ice melting,which has a certain engineering application value.In this thesis,the ice thickness prediction model and the ice melting model are discussed,and the accuracy and feasibility of the model are verified.In the future research work,we can combine more actual icing cases to further improve the model and improve the generalization ability of the model.