Research On Effect Evaluation And Investment Forecast Of Six Defense Technical Transformation Project For Overhead Lines

The six defense technical transformation projects for bird damage,and pollution flashover,and wind damage,and external damage,and lightning damage,and ice damage are of great significance to the risk prevention of overhead lines and the safe and stable operation of the system.The investment in the six defense technical transformation is an important part of the assets of the power grid company.At present,the existing evaluation methods for the effect of production and technical transformation do not fully consider the economic benefits of overhead line operation and risk attribute,also,the evaluation index system is not comprehensive and hard to measure,which makes it hard to evaluate the effect of technical transformation,and the lack of effective prediction methods for technical transformation investment leads to the lack of supporting basis for investment decision-making.In view of the above problems,this dissertation has completed the systematic research,and the main work is as follows:The effect evaluation system of six defense technical transformation project is constructed.This dissertation combs the principles and measures of six defense.This dissertation studies the line risk attributes corresponding to the six defense measures,and relates the line operation risk,and introduces the concept of line risk value.By introducing the concept of life cycle cost,the index of unit investment loss reduction benefit is built as the benefit evaluation index.The technical transformation project income table is compiled and the basic data collection is explained.The evaluation system of six kinds of single measure technical transformation project is constructed by synthesizing the operation index with risk index and benefit index of the technical transformation project.The specific evaluation index of each level is determined,and the index calculation method is elaborated,so as to effectively realize the quantitative evaluation of the technical transformation technical effect and economic benefit.Based on subjective and objective weighting,the effect evaluation method of six defense technical transformation project is proposed.AHP is used to determine the subjective weight of the indicators in the evaluation system of technical transformation effect.According to the basic data of the samples of the technical transformation projects,the variance weight with information entropy weight and gray correlation weight of the indicators are calculated respectively,and the average of the three is determined as the objective weight of the indicators,and the objective weight is more balanced than the subjective weight.The subjective and objective weights are integrated and normalized to subjective and objective weights to avoid the subjectivity and randomness of the evaluation process.The index score and subjective and objective weight are linear weighted to determine the comprehensive evaluation results of the technical transformation project.This dissertation has completed the evaluation of six defense technical transformation samples to verify the comprehensiveness and rationality of the evaluation model.A combinational forecasting method based on machine learning for overhead line technical transformation investment demand is proposed.Based on the first mock exam,this dissertation extracts historical data of technical transformation projects,and then selects the main factors based on data collection and preprocessing.Then this dissertation uses principal component analysis to cut down the dimension of the features and completes the frame of the model input.Then this dissertation uses XGBoost and SVR separately to predict the investment demand by single model,and uses the best combination weight method with the error matrix.The Lagrange multiplier method is introduced to determine the weight coefficient of two single prediction models,and the combined prediction model is established in this way.Finally,the sample calculation example shows that the combined model can improve the prediction accuracy with efficiency.