<![CDATA[dispersed demands, insufficient utilization of renewable resources, weak infrastructure, and differences in economic development stages. To address these issues, this study proposes a dynamic planning framework that integrates the macroeconomic development stage classification and improves the multi-objective pelican optimization algorithm. The algorithm dynamically adjusts the target weights and constraint boundaries based on the macroeconomic development stage. The algorithm initializes the population using the Sobol sequence combined with weak Gaussian perturbation and introduces an adaptive dynamic factor embedded with stage information to jointly adjust the search step size. A two-stage stochastic programming model with the objective of minimizing the total annual equivalent cost and maximizing the annual net carbon reduction is constructed. Simulation examples using typical rural areas in northern China show that the proposed algorithm outperforms the comparison algorithms in terms of convergence and distribution of solution sets. The Pareto solution set distribution index is the best at 0.082. The average annual total cost of the optimization scheme is 14.856 million yuan, which is 2.5% to 4.9% lower than the comparison algorithm, and the annual net carbon reduction is 634.2 tons, with a comprehensive energy efficiency of 76.8%. This framework converts the a priori knowledge of macroeconomic stages into adaptive parameters of the algorithm, effectively connecting planning preferences with search strategies, and providing a dynamic decision-making tool that balances economy, low carbon, and reliability for rural areas at different development levels. It has theoretical and application value for the coordinated advancement of rural energy transition and rural revitalization.]]>
