Collaborative optimization of distribution and microgrid based on combined forecasting and improved analytical target cascading method

The stochasticity and volatility of renewable energy generation pose significant challenges to the secure and economic operation of power grids. To address the issues of insufficient prediction accuracy and the collaborative optimization of distribution networks and microgrids (MGs), this study conducts a comprehensive investigation. First, a hybrid prediction model based on Ensemble Empirical Mode Decomposition (EEMD), Principal Component Analysis, and Bidirectional Long Short-Term Memory is constructed. EEMD is utilized to decompose renewable power sequences, which are subsequently filtered according to their contribution values to enhance prediction accuracy at the source. Second, operational models for the distribution network and microgrids are established. In the subsequent solution of the coordinated dispatch model, an improved Analytical Target Cascading method incorporating a balance coefficient is designed to eliminate the influence of improper initial penalty multiplier selection on model convergence. Simulation results based on an IEEE 33-node system integrated with microgrids demonstrate that the proposed method effectively reduces network losses and voltage deviation rates while further enhancing the overall economic efficiency of the system.