Prediction of cascading collapse occurrence due to the effect of hidden failure protection system using different training algorithms feed-forward neural network / N. H. Idris ...[et al.]

• Main Authors: Idris, N. H., Salim, N. A., Othman, M. M., Yasin, Z. M.
• Format: Article
• Language: English
• Published: UiTM Press 2017
• Subjects: Neural networks (Computer science)
• Online Access: https://ir.uitm.edu.my/id/eprint/63015/1/63015.pdf; https://ir.uitm.edu.my/id/eprint/63015/; https://jeesr.uitm.edu.my/v1/

Protection system plays a significant role in power system and operation of electrical networks especially in transmission system. The outage in transmission line that causes from hidden failure in protection system should be avoided. Artificial Neural Network (ANN) is one of the problem solver with variety of training algorithms that helps to predict the cascading collapse occurrence due to the hidden failure effect. The historical data obtained from NERC report is analyzed and being used in ANN for prediction purposed. This paper compares the supervised training algorithms of feed-forward neural network with backpropagation include Lavenberg- Marquadt (LM), Scale Conjugate Gradient (SCG) and Quasi Newton Backpropagation (BFG). IEEE 14 bus system is used as a case study. The performance of the training algorithms is analyzed based on Correlation Coefficient (R) and Mean Square Error (MSE)