Transmission tower bird damage risk prediction method based on random forest

Abstract

The invention discloses a transmission tower bird damage risk prediction method based on a random forest, and the method comprises the steps of collecting the transmission tower bird damage historicaldata, and constructing an original data set; dividing the bird damage risk levels in combination with the bird damage historical data to obtain a sample set; preprocessing the bird damage data in thesample set to obtain a training set and a test set; initializing a random forest classifier; training the bird damage data in the training set by using a random forest algorithm to generate a randomforest classification model; determining the bird damage risk level of a transmission tower. According to the invention, the random forest algorithm in the machine learning is adopted to predict the bird damage risk level of the power transmission tower, the high-dimensional data can be well processed, the feature filtering is not needed, and the training process can be highly parallelized. The method is quite friendly to the large sample data training, the model training speed is high, the over-fitting is not likely to happen, the complex mathematical operation like an analytic hierarchy process is not needed, the deviation caused by the subjective weighting can be reduced, and the solving precision is improved.

Description

technical field [0001] The invention relates to the technical field of machine learning, in particular to a random forest-based method for predicting bird hazard risk on power transmission towers. Background technique [0002] In recent years, transmission line faults caused by bird damage have occurred frequently, for example, line tripping and power failure caused by bird damage has become one of the main faults in the transmission network. The existing prevention and control of bird damage and monitoring mostly rely on manual inspection, which is blind, and the bird prevention is not timely, and the effect of bird prevention is not obvious. Because of its many types, large numbers, wide distribution and rapid changes, bird damage is easily affected by geography and climate. It is difficult for power operation and maintenance personnel to grasp the status of bird damage in a timely and effective manner, which wastes huge manpower and material resources, but it is difficult...

AI Technical Summary

Problems solved by technology

[0003] At present, relevant experts and scholars mainly use the AHP to establish the bird damage risk assessment model, but the traditional AHP has some defects in the establishment of the model. Judgment matrix often requires a lot of data modification, which takes a lot of time. Secondly, when analyzing problems, the analytic hierarchy process has many qualitative components and strong subjectivity, and the analysis results are not convincing.

Images