A framework planning method for distribution networks considering the source-load correlation and uncertainty

At present， distribution network framework planning plays a critical role in enhancing grid reliability and stability. However， the correlation between source and load outputs is often neglected， leading to a high rate of solar power curtailment， significant investment in the distribution network framework， and low line utilization rates. To address these issues， this paper proposes a framework planning method for distribution networks that considers the correlation and uncertainty of source and load output. Firstly， historical data on source and load is analyzed， accounting for the temporal autocorrelation of photovoltaic （PV） and load outputs. Latin hypercube sampling （LHS） and Cholesky decomposition are employed to generate source-load scenario sets. By incorporating their correlation， a load reduction method under these scenarios is designed accordingly. Secondly， the uncertainty in source and load outputs is considered to determine typical source-load scenarios and their probabilities. A multi-objective optimization planning model is developed for the distribution network. This model aims to maximize expected PV consumption， minimize the expected annual investment and operational costs， and minimize the expected value of line overload probability， while considering relevant constraints. The improved NSGA-II algorithm is used to solve this model and generate a distribution network framework planning scheme. Finally， the differences among Pareto frontier solutions and the impact of various parameters on the planning results are discussed， providing planners with alternative options for decision-making.