A hybrid model for state estimation prediction composed of neural network and PSO algorithm for Iraqi national super grid system

This paper demonstrates the importance of the use of power system State Estimation (SE) to determine the real value of bus voltages and their angles in one line operation states. This is done to create a robust power system network that is able to monitor and analyse operational conditions and predict any emergency that may occur on the Power System Network due to human error, strong system component failure and other factors. A new hybrid (particle swarm optimization-Neural Network) algorithm is proposed to estimate the angles and voltages of buses in electrical system networks. All the results obtained from this study were compared with Phase Measurement Units (PMUs), Weighted Least Square (WLS) and Neural Network. A graph theoretic procedure algorithm (GTPA) was used to make the PMU's placements. The WLS, NN, and PMU and the new algorithm were tested on the 24 bus Iraqi Grid system. A MATLAB program was used to develop WLS codes and a hybrid (particle swarm optimization-Neural Network) algorithm. The MATLAB toolbox (PSAT and NN) were used for PMUs and Neural Network respectively. The simulation results showed that the algorithms were a good visualization and powerful system state estimation tools. The new algorithm was faster analyser than the conventional algorithm (WLS, NN, and PMU), it requires less parameter tuning. The results obtained proved the validation, acceptance and accuracy of the hybrid particle swarm optimization-Neural Network algorithm's state estimation (SE).