State space modeling and control of power plant electrical faults with neural networks for diagnosis
Takialddin Al Smadi, Khalaf S. Gaeid, Ayad Tareq Mahmood, Rawaa J. Hussein, Yaseen Al-Husban
Abstract
Samarra power station is an operating power station of at least 1260 megawatts (MW) in Samarra, Salah al-Din, Iraq. It is also known as Salah al-Din. This paper introduces a fault location method for transmission lines, utilizing an artificial neural network (ANN) technique to analyze voltage and current data from both terminals. The method specifically targets line-to-line, double-line-to-ground, and single line-to-ground faults that cause system downtime. The study focuses on applying ANN as an intelligent tool depending on binary NN (BNN) for diagnosing faults in electrical power facilities. The selection of ANN is based on its good performance in pattern recognition, classification, matching, prediction, decision-making, and control and optimized with genetic algorithm (GA). Different modeling validation algorithms for the proposed system, such as ARX, transfer function, and state space model, are used. The fit to estimation data and it's also the prediction of the accuracy are 70.39%, 98.18%, and 99.84% according to the identification toolbox of the Matlab 2020 respectively. Also, the results of the ANN algorithm, such as fault location, status, and time, are compared to the real corresponding terms in the electrical system of Samarra city in Iraq. The difference between the experimental cases inside the Samarra power station and ANN is 0.1%. The future development of the Samarra thermal power station which may powered by solar panels is also introduced or using the Internet of Things (IOT) in the next step to control the systems equipment's.