| Title |
Design of Artificial Neural Network (ANN) Controller using Deep Learning to Damp out Inter-area Oscillations in Power Systems |
| DOI |
http://dx.doi.org/10.5207/JIEIE.2018.32.8.023 |
| Keywords |
Artificial Neural Network ; Deep Learning ; Low-frequency Oscillation ; Lead-lag Controller ; Power System |
| Abstract |
The ANN (Artificial Neural Network) has been harnessed in artificial intelligence as well as image processing and classification because of its versatility. A lead-lag controller is mainly used to damp out low-frequency oscillations and utilized in power systems as a PSS (Power System Stabilizer) and the ancillary controller of FACTS (Flexible AC Transmission System). However, there are some limitations incurred by the variable and unpredictable power system states, so that fixed controller parameters cannot damp out these oscillations properly. Moreover, in case of inter-area oscillations, the PSS at each generator will not satisfactorily work. In order to overcome these inherent drawbacks and solve the nonlinear system's problems, many researchers attempted to incorporate ANNs into the PSS of a generator. Although the sigmoid function is usually used as the activation function of a PSS, the saturation range may act as a disturbance in deep learning. In this paper, we investigated the control features of electromechanical oscillations in power grids using two different activation functions in an ANN controller. |