Voltage security monitoring, prediction and control by neural networks

The authors present voltage collapse evaluation as an artificial neural network task with the aim of making the evaluation fast enough for online use. They describe the use of a neural network to approximate the complicated mathematical functions of the voltage collapse evaluation method. The approximation is achieved by a learning process in which the neural network is trained to associate the security level of a power system with its operating condition which is characterised by the system parameters. In addition to voltage security monitoring, the neural network can be exploited for contingency monitoring, security prediction, and voltage control. The IEEE 57 busbar network is used to demonstrate the application of the neural network.<
>