Health Status Assessment Of Mountain Transmission Towers And Lines Under Wind And Ice Load

Transmission lines have double characteristics of high towers and large span distribution,which are composed of continuous transmission conductors connecting different transmission towers.Under the action of natural disasters such as wind and ice,damage and collapse are easy to occur,especially in the hilly geographical environment affected by special local climate and meteorology,the stress characteristics of transmission lines have complex relationships with altitude,slope orientation,terrain morphology and local climate.At present,the analysis of the health of mountain transmission tower lines mainly includes model analysis and measurement techniques,but simple model analysis can not reflect the special working conditions of transmission tower lines,and detection data are not easy to obtain the structural state of transmission tower lines.This paper combines mechanism model and data model to study the health status of mountain transmission lines.The main research contents are as follows:Firstly,the difficulties of the current research on health assessment of mountain transmission lines are analyzed,based on the mechanism model of tower-line structure of mountain transmission lines.The sensitive parameters such as wind-ice load,elevation difference and non-uniform span are integrated to establish the nonlinear dynamic analysis model of the coupled tower-line structure by finite element method.The safety of the transmission line is analyzed from the perspective of steady state and transient state,revealing the action law of external environment on the tower-line structure in the mountainous area with undulating terrain.In the steady state,the axial force of the transmission tower and the contribution of wind-ice load to the axial force are extracted,and the ultimate load value of the tension tower is completed by considering the bending effect and the critical initial defect.In the transient study,the material and geometric nonlinearity are further considered to analyze the impact of wind-ice load on the model,such as the tension,stress and tower-line displacement of the transmission line,and then the force transmission process of the straight tower is completed to provide the mechanism model basis for the subsequent prediction model of mountain transmission tower and health assessment model of transmission line.(2)In order to more accurately estimate the icing thickness of transmission lines,feature parameters related to it are selected from the monitoring parameters and an icing thickness prediction model is established.However,due to the specific correlation between the input feature parameters and the high dimensionality of the data,the modeling becomes complex and the generalization function is weak.Therefore,in order to further improve the accuracy of the estimation,SDAE(stacked denoising autoencoder)is used for feature compression in order to achieve better prediction results.A BiLSTM-based icing thickness prediction model was trained using the feature parameters extracted from the SDAE hidden layer in order to accurately estimate the standard icing thickness of the transmission line.First,in order to evaluate the advantages and generalization of the model,the artificial ice-viewing data was used for verification again.For further verification of the accuracy improvement of the model,RNN(Recurrent Neural Network),LSTM,BiLSTM,PCA-BiLSTM(Principal Component Analysis-Bidirectional Long Short-Term Memory Network)model methods were selected to compare with the SDAEBiLSTM prediction model of this paper,and a melting experiment was established to verify the effectiveness and accuracy of the prediction model of this paper.KNN,RF,XGBoost and BP models were selected for comparison and analysis to verify that the overall effect of the SDAE-BiLSTM in this experiment is better than other prediction models in short-term prediction.(3)Select multiple indicators to represent the health status of transmission lines,and use the method of hierarchical analysis and fuzzy evaluation to evaluate the health trend of transmission lines.The evaluation index of transmission lines is divided into three layers,with the safety trend warning map established by the mechanism model as the weight index of part of the parameters,and the weight calculation of the transmission lines is carried out by the combination of subjective and objective methods.Finally,the fuzzy evaluation model of transmission line health trend is evaluated by the improved 3εn criterion cloud model,and the applicability of the integrated mechanism model of transmission line hierarchical fuzzy evaluation model is verified by taking the actual working condition of the transmission line as an example.