Lasso-GBDT-based Investment Forecasting for Distribution Transformer Replacement Projects

With the continuous development of China's economy, the requirements for safety and reliability of power supply are greatly increased, and the investment in technical transformation of power grid companies is increasing. As an important project of the technical transformation project, the replacement of 10kV distribution transformers occupies a large proportion of the project investment. It is beneficial to realize the controllable and controllable cost by analyzing the influencing factors of cost and improving the accuracy of cost forecast. Based on the actual engineering data, this paper constructs a 10kV distribution transformer replacement investment prediction model based on Lasso and GBDT algorithms. Firstly, the Lasso regression model is used to screen multiple features of the data, and secondly, the screened features are subjected to GBDT modelling. The prediction results show that the prediction accuracy and stability of the model used in this paper are better than the benchmark model, which is suitable for investment predictions for the power distribution transformer replacement project, and can contribute to the economic and social benefits of power grid companies.