Distribution network engineering project construction period prediction method based on feature selection

Abstract

The invention discloses a distribution network engineering project construction period prediction method based on feature selection, and the method comprises the steps: selecting factors affecting theprogress of a project and unique factors of a distribution network engineering project, and constructing a factor set; training an extreme learning machine classifier by using all factors in the factor set; using 10-fold cross validation for obtaining classification precision p, randomly and temporarily removing a certain factor in the factor set, retraining the extreme learning machine classifier, using 10-fold cross validation for obtaining classification precision p', determining calculation factors through comparison and screening, and establishing a construction period prediction model for construction period prediction calculation. A multi-step dimension reduction method is designed by using an extreme learning machine. Main factors of a distribution network engineering project datasample are extracted to obtain a small number of main factors with high prediction capability, and the main factors are used for establishing a prediction model of a distribution network engineeringproject, so that the possibility of project overdue completion is greatly reduced, the accuracy of an engineering plan is enhanced, and the project overdue risk is effectively reduced.

Description

technical field [0001] The invention relates to the field of engineering prediction, in particular to a feature selection-based method for predicting the duration of a distribution network engineering project. Background technique [0002] According to data, at present, power supply enterprises have difficulties in the process of project progress control and real-time adjustment of project plans for distribution network projects. The promotion has its own characteristics, and the time consumption of each step in the project is very different. It is difficult to use the prescribed standard limits, which leads the company to basically rely on experience to judge the progress of the project in the management and control, and it is difficult to realize the lean management and control of the whole process of the distribution network project. At the same time, due to the inability to use accurate data to quantitatively analyze the lag or advance of the progress of each stage of th...

AI Technical Summary

Problems solved by technology

[0005] The present invention mainly solves the technical problem that the original construction period prediction calculation takes too long and may lead to prediction failure, provides a method for forecasting the construction period of distribution network engineering projects based on feature selection, and designs a multi-step reduction method using an extreme learning machine. Dimensional method, extract the main factors of the data samples of the distribution network engineering project, and get the main factors with a small number but strong predictive ability, which are used to establish the prediction model of the distribution network engineering project, provide data support for the real-time change of the engineering plan, and greatly reduce the project cost. The possibility of overdue completion can effectively improve the lean management and control ability of the whole project process, strengthen the accuracy of the project plan, and effectively reduce the risk of project overdue

Images