Research on Cost Forecasting Method of Power Grid Transmission Project

Transmission engineering is an important part of power engineering construction, and accurate cost prediction can effectively promote project evaluation, budget audit and construction. This study presents an integrated model forecasting method based on Stacking. SVM, BP neural network and KNN machine learning algorithms are selected as base learners to establish a transmission engineering cost prediction model based on Stacking integrated learning. The results show that compared with a single forecasting model, the integrated learning model can accurately predict the engineering cost.