A probabilistic planning method for hybrid high-voltage AC/DC distribution networks under multi-dimensional uncertainties

In the context of constructing a new-type power system， distribution networks face the impact of multi-timescale uncertainties during the planning period， which significantly affects the feasibility and economic efficiency of the final planning scheme. To address this， multi-dimensional uncertainties are modeled as a multi-stage hybrid scenario tree. Based on this model， a multi-stage probabilistic-stochastic planning model is proposed for hybrid high-voltage AC/DC distribution networks， taking into account large-scale new energy consumption and system reliability requirements. To resolve curse of dimensionality caused by the massive number of scenarios in the large-scale mixed integer programming model， an improved Benders decomposition algorithm is proposed to decouple and efficiently solve the original model. Finally， the effectiveness and advantages of the proposed planning model are validated through case studies of a high-voltage distribution system. Compared to existing planning methods， the proposed method enhances the system′s adaptability to multi-dimensional uncertainties， improves new energy consumption levels， and reduces total planning costs.