Title :
Controlled islanding strategy for power systems based on flexible semi-supervised spectral clustering
Author :
Azadian, Farshad ; Radzi, M.A.M. ; Wahab, Noor Izzri Abdul
Author_Institution :
Dept. of Electr. & Electron. Eng., Univ. Putra Malaysia, Serdang, Malaysia
Abstract :
Islanding strategy is a self-healing method applying to power systems as the final step to prevent wide area blackouts. To suggest an effective islanding method, several parameters should be considered. Without those considerations, there is high possibility that generated islands will not be stable enough and will collapse soon. In this paper, a new method based on flexible semi-supervised spectral clustering is proposed to handle multiple constraints in order to guarantee the stability of generated islands. This method is also capable of handling and tackling various satisfaction degree of each constraint. Hence, the best solutions are achieved with respect to each defined constraint and its level of importance. The efficiency of this method for power systems is shown by applying it on the IEEE 39-buses network.
Keywords :
IEEE standards; distributed power generation; power distribution faults; power generation control; power generation reliability; power system stability; IEEE 39-bus network; controlled islanding strategy; flexible semisupervised spectral clustering; islanding stability; power system blackout prevention; self-healing method; Algorithm design and analysis; Clustering algorithms; Generators; Islanding; Load flow; Power system stability; Thermal stability; Controlled islanding; blackout prevention; flexible spectral clustering;
Conference_Titel :
Research and Development (SCOReD), 2013 IEEE Student Conference on
Conference_Location :
Putrajaya
DOI :
10.1109/SCOReD.2013.7002630