Adaptive Load Frequency Control of Nigerian Hydrothermal System Using Unsupervised and Supervised Learning Neural Networks
This document has been relocated to http://scholarsmine.mst.edu/ele_comeng_facwork/1951
There were 17 downloads as of 29 Jun 2016.
Abstract
This work presents a novel load frequency control design approach for a two-area power system that relies on unsupervised and supervised learning neural network structure. Central to this approach is the prediction of the load disturbance of each area at every minute interval that is uniquely assigned to a cluster via unsupervised learning process. The controller feedback gains corresponding to each cluster center are determined using modal control technique. Thereafter, supervised learning neural network (SLNN) is employed to learn the mapping between each cluster center and its feedback gains. A real time load disturbance in either or both areas activates the appropriate SLNN to generate the corresponding feedback gains. The effectiveness of the control framework is evaluated on the Nigerian hydrothermal system. Several far-reaching simulation results obtained from the test system are presented and discussed to highlight the advantages of the proposed approach.
