Abstract—Ice load generally occurs cold and heavy fog ambient conditions, and it causes major faults in overhead lines. Hence ice load detection studies are important to prevent ice load effect, and image processing is very useful method for ice load detection studies. It is indicated in literature, dark and heavy fog ambient conditions are major problem for ice load detection. These problems are solved by using multilevel threshold-optimization method, but local optimum falling problems occurs due to high threshold level. Hence threshold level must be reduced to reduce number of local optimum falling problems. In this study, multilevel threshold- optimization methods are used, and threshold level is reduced to reduce local optimum falling number. Gravitational Search Algorithm (GSA) is used as optimization method, and Ramesh and Otsu methods are used as multilevel threshold method. When 7 level GSA-Ramesh Method is used, its accuracy rate is 99.43%, also number of local optimum falling is reduced.
Index Terms—GSA, ice load, optimization method, ramesh method, otsu method.
Bahadır Akbal is with the Department of Electrical and Electronics Engineering, Selçuk University Konya, Turkey (e-mail: bakbal@selcuk.edu.tr).
[PDF]
Cite: Bahadır Akbal, "Comparison of GSA-Ramesh and GSA-Otsu Method to Detect Ice Load in Real Ambient Conditions," International Journal of Engineering and Technology vol. 9, no. 1, pp. 40-44, 2017.