Research On The Structural Analysis And Monitoring Information Processing Of Mountain Power Transmission Towers And Lines

High voltage transmission line is the lifeblood of the national economy,and it is particularly important to maintain its normal operation.China's high voltage transmission lines spanning mountains,especially in Yunnan,are complicated in geology and changeable in climate and environment.La ndslide is one of the typical geological disasters.Monitoring and identifying the state of high voltage transmission lines is an important part of maintaining the safety of high voltage transmission towers in mountainous areas.In order to study the influ ence of landslide to the tower line and identify the state of the tower line through monitoring data,this paper first analyzes the position relation of the easy landslide and the tower and the stages of the landslide occurrence to determine each working c ondition and carry out the finite element analysis.Secondly,the state model of the tower line system,the optical fiber monitoring model and the parameters for determining the safety state of the tower line are established by comparing the results of the finite element analysis.Finally,the monitoring data is selected from the tower optical fiber monitoring system for data pre-processing and tower safety status identification.This paper uses the confidence distance to improve the classical fuzzy theory algorithm to preprocess the monitoring data,then uses the pre processed data to train the BP neural network and the GA-BP neural network,and uses the trained model to identify the tower line state.The main contents of this paper are as follows:(1)According to the stage of landslide occurrence and the relative position of the tower prone to landslide,four working conditions are obtained.The tower is located in the scope of landslide impact,resulting in severe landslides.The tower is located on the landslide body,and the landslide falls down unevenly.The tower is located on the landslide body,resulting in horizontal slip of the landslide.The tower is located on the landslide body,and the landslide body is sunk.The location,size,change rule and maximum displacement of the maximum stress of single tower under four working conditions are analyzed.There are four typical models for mountain high pressure tower line system erection.According to the actual situation of a certain line in Yunnan,one of them is selected for analysis.From the four working conditions,three common working conditions are selected to analyze the location,size and variation of the maximum stress load under the influence of landslide and the interaction of the tower lines.The similarities and differences of the stress load magnitude,location and transformation rule of single tower and tower line system under the action of landslide are analyzed.(2)According to the finite element analysis results of single tower and towe r line system and the inherent nature of the tower line,a safety state model of mountain tower line is established.The safety state model determines the relationship between the external physical conditions and the state of the tower line system when the landslide occurs.The optical fiber monitoring model for tower line in mountainous area is established,and the monitoring parameters for identifying tower line safety are determined.The data in the monitoring s ystem are selected according to the parameters needed to identify the state of the to wer.Preprocessing the monitori g data,eliminating data noise and data specification.The classical fuzzy theory algorithm is used to preprocess the monitoring data.In order to make up for the shortcoming of the data noise processing which is caused by the close degree of the classical algorithm,the confidence interval is used to improve the classical fuzzy theory for data preprocessing.(3)Training BP neural network with preprocessed data to identify tower line status.BP neural network is easy to fall into local optimum when it does not meet the required precision.In order to solve th e shortcomings of BP neural network,genetic algorithm is used to optimize it.The same data set is used to train and test the GA-BP neural network,and the accuracy,recall and time complexity of the two algorithms are compared,and a more suitable algori thm is selected to monitor the state of the tower line.