Bird Damage Risk Warning Model And Software Realization For Overhead Transmission Lines

In recent years,the frequent occurrence of transmission line bird damage accidents has aroused great attention of the State Grid Corporation of China,so a large number of human and material resources are invested in bird damage prevention and control every year,but the effect is not obvious.Many experts and scholars have also carried out a large number of studies on the problem of bird damage.The research directions mainly include two kinds: one is to improve the design of bird prevention devices;the other is to predict bird damage by combining various factors affecting bird damage.For the prediction of bird damage risk,most thesis are using Comprehensive Evaluation Method,the Entropy Weight Method and analytic hierarchy process construct the sound prediction model,but these prediction methods are subjectivity is stronger,the real significance is not big problem,and these thesis will be set for all eigenvalues corresponding weights,but for the size of the weight assignment is no unified standard system,So the difference is large.In addition to the sound of data statistics,tend to be ideal,in the actual eccn company might lacks for a detailed record of the sound,for example,some current power grid company for the sound data collection is only the birds have the detailed record tripping caused by accident,and for not only records of the time line accident happened for birds,There is no indication of bird droppings falling,large birds flying over or nesting on towers,so it is impossible to determine the level of risk associated with such data.Therefore,for such unlabeled data,this thesis adopts the clustering analysis algorithm to build a bird damage risk warning model.The main idea is to use the relationship between data themselves to classify bird damage risk levels,reduce the influence of human subjective factors and reduce the requirements of the model on original data.In this thesis,the main content is based on machine learning algorithms developed a set of transmission line of the sound risk early warning model,and based on the Vue front frame design to develop a set of sound map software,make the power grid operations staff more clearly see the sound risk distribution,and targeted to develop the related prevention measures,enhances the working efficiency of the anti bird,better guarantee smooth operation of the transmission line.The research of this thesis is as follows:（1）Based on the collected bird damage fault data of a power supply bureau in Guangdong from 2016 to 2020,and combined with relevant research data,the data were supplemented and sorted out to analyze the relationship between the occurrence of bird damage and voltage level,tower type and surrounding environment,and construct the original data set based on voltage level,tower type and surrounding environment information.（2）In view of the problem that human factors have a great influence on the construction of bird damage risk warning model,this thesis uses K-Means clustering,DBSCAN clustering and spectrum clustering in unsupervised machine learning algorithm to build the model.The elbow rule and contour coefficient method are used to improve the K-Means clustering algorithm,which has the randomness of the clustering number,and the principal component analysis method is used to reduce the dimension of the data.The results show that the data clustering effect after dimensionality reduction is better.（3）SSE index and CHI index were used to compare the models constructed by the improved K-Means clustering algorithm,DBSCAN clustering algorithm and spectral clustering algorithm,and the results showed that the improved K-Means clustering algorithm had better clustering effect.（4）Aiming at the problem of low efficiency of the current of the sound control,this thesis built the sound risk early warning model of the sound risk rating projections for transmission line tower,and based on the vue front frame design to develop a set of sound map software,can make the power grid operations staff at all levels more intuitive to see the sound distribution of targeted for bird prevention measures for prevention and control of the sound,Improve work efficiency.