Integrating Classification and Regression Tree Algorithm for Operational Optimization in AC-DC Hybrid Power System Planning: A Novel Approach

The increasing demand for electricity and the imperatives of climate change have made the optimization of power system planning critical for the energy transition and grid efficiency. This study presents an innovative planning method for inter-regional AC-DC hybrid power systems, leveraging the Classification and Regression Tree (CART) algorithm to optimize the operational characteristics of direct current (DC) channels. By designing a closed-loop iteration, precise operational constraints are considered by the CART algorithm, which immerged into the planning model to achieve safe and economic optimization. Based on the empirical analysis of the HRP-38 system, this study concludes that the CART algorithm offers a constructive approach to managing the operational complexities of modern power grids. By optimizing and refining DC operational characteristics based on actual system requirements, the algorithm contributes to improvements in safety, economic efficiency, and environmental sustainability within the confines of the HRP-38 node system. Consequently, the effectiveness of the CART optimization approach could be corroborated. Meanwhile, this study also acknowledges the limitations in generalizing these results to other power grid configurations and the need for further exploration in developing environmentally conscious planning methods.