Development of Adaptive Artificial Neural Network Security Assessment Schema for Malaysian Power Grids

Abstract

The mission of the power system operator has become more complicated than before due to increasing load demand, which causes power systems to operate near their security limits. The deregulation of electricity markets, which requires independent system operation driven by economic considerations, is still an essential requirement of modern power systems. This study presents an enhanced model of developed adaptive artificial neural network (AANN) technique for security enhancement of Malaysian power grids, inclusive of a remedial action (generation redispatch/load shedding) at any scale of system operation. Automatic data knowledge generation systems for AANN inputs and data selection and extraction methods are developed. Results show that the proposed AANN can provide the required amount of generation redispatch and load shedding accurately and promptly for computing large sample data.

Keywords

• Security assessment
• artificial neural network (ANN)
• backpropagation
• remedial action
• contingency analysis