Iterative Direction Expansion of Power Flow and Convergence Analysis Based on Deep Learning

Power flow calculation is essential for optimizing control actions and analyzing the running state of power systems. The development of smart grids has led to more complex power systems, challenging the convergence and computational efficiency of conventional algorithms. This study focuses on the limitations of current power flow algorithms in terms of iteration direction and proposed a method to extend the iteration direction in power flow calculation. Moreover, this method incorporated deep learning techniques to ease the computational burden while improving the convergence. This study aims to provide a useful reference for the analysis of complex power flows in power systems.