NEURAL NETWORKS BASED LOAD FLOW ANALYSIS OF RADIAL DISTRIBUTION NETWORKS

load flow studies are performed to determine the distribution system static states at each node to find the steady state operating condition of a distribution system. Load flow study is the most frequently carried out study used by distribution companies at all the stages of system planning, operation and control. In this paper Radial Bias Function Networks (RBFN) and multilayer feed forward network with back propagation algorithm (MFFNN with BPA) are proposed to solve load flow problem under different loading conditions for computing node voltage magnitudes of the distribution system. The off-line load flow calculation is carried out by network topology based three phase distribution load flow. The results of network topology based three phase distribution load flow calculations are used to train the Radial Bias Function and multilayer feed forward network with back propagation algorithm.