Research On Tree Barrier Early Warning Model Of Airborne Transmission Line Spectrum Analysis And Multi-objective Optimization

With the implementation of the policy of returning farmland to forest and the construction of ecosystems,the maintenance of transmission line vegetation has become one of the important tasks of the daily operation and maintenance of the power system.Compared with the traditional manual line patrol,which is time-consuming,labor-intensive and inefficient,using a predictive model and a telemetry system to establish a targeted optimization model can reduce the meaningless artificial waste while strengthening the management plan of line vegetation maintenance In order to manage vegetation around the transmission line more scientifically and effectively,this paper uses hyperspectral to identify the distribution of vegetation in power transmission corridors,screens out high-threatening fast-growing trees according to the recognition results,and reconstructs the growth model to establish a growth model suitable for vegetation early warning.Revise the vegetation maintenance reliability objective function,and establish an optimization model that includes maintenance reliability and economy.The individual parameters in the model are determined by combining the spectral identification and the tree height growth model.The specific work is as follows:Firstly,due to the complex terrain of the transmission corridor,it is not easy for people to enter,and the maintenance staff’s professional restrictions,etc.,it is difficult to effectively understand the overall distribution of high-threat trees and bamboos,resulting in the possibility of missing or inaccurate transmission line maintenance.In order to overcome this situation,this paper effectively classifies the spectral lines by performing spectral preprocessing,spectral line transformation,feature extraction and other methods on the hyperspectral images taken by the drone to identify the types of trees and bamboo in the transmission corridor.The experimental results verify the feasibility of identifying the specific tree species distribution in the transmission corridor by airborne hyperspectral.Secondly,by analyzing the traditional growth model,combined with the actual application of the transmission line tree barrier warning,it is found that the traditional growth model uses tree age or breast diameter as the independent variable.This information is not easy to obtain,and there is a large error in using the existing tree height to infer.To solve this problem,the model is reconstructed,the existing tree height is used for prediction,and the influence factor is added to modify the model.Using the data of measured fast-growing high-threat trees for model verification,the effectiveness of the reconstructed model is proved.Thirdly,establish a vegatation maintenance model for high threat tree species in the transmission corridor,combine the spectral recognition results with the tree growth reconstruction model to determine individual parameters.Modify the reliability function of the transmission line tree barrier maintenance,use the revised line reliability and economy as the objective function,establish a multi-objective optimization model according to the binary coding method,and ensure the maintenance of the transmission line reliability while reducing economic expenditure as much as possible.Through the comparison of actual cases,it is proved that the revised tree maintenance optimization model can provide a more superior maintenance plan,and the final maintenance plan is given to provide the minimum economic expenditure plan under the premise of equal reliability,to avoid unnecessary economic waste.