Observability Analysis for Large-Scale Power Systems Using Factor Graphs

The state estimation algorithm estimates the values of variables based on measurement model described as system equations. Prior to applying algorithm, existence and uniqueness solution underlying equations is determined through observability analysis. If a unique does not exist, analysis defines observable islands further an additional set (measurements) needed determine solution. For first time, we utilise factor graphs Gaussian belief propagation define novel approach. placement measurements restore are identified by following evolution variances across iterations over graph. Due sparsity power network, resulting method has linear computational complexity (assuming constant number iterations) making it particularly suitable for solving large-scale systems. can be flexibly matched distributed resources, allowing determination restoration in fashion. Finally, discuss performances proposed using systems whose size ranges between 1354 70 000 buses.