Optimization Model for Whole Life Cycle Selection of Distribution Transformers Based on Load Characteristics Fault Risk Dual Factor Coupling

In response to the dual challenges of intensified load spatiotemporal fluctuations and rising fault risk costs in the selection of distribution transformers under the background of new power systems, this paper proposes a full life cycle selection optimization model for distribution transformers under new power systems. It uses the K-means++ algorithm to extract load spatiotemporal characteristics and constructs a load feature map. Based on the Arrhenius equation, it establishes a dynamic load rate driven insulation aging fault probability model to quantify fault risk costs. An improved NSGA-II algorithm is designed to optimize the life cycle cost (LCC) function, integrating initial investment, operating losses, and failure risk costs. Case analysis shows that this model reduces LCC by 12.7%, improves load rate balance by 18.3%, and speeds up optimal solution convergence by 34.6% compared to traditional methods. It provides a practical tool for selecting distribution equipment and improving grid asset management.