In recent years pattern recognition of power quality (PQ) disturbances in smart grids has developed into crucial topic for system equipments and end-users. Undoubtedly analyzing the PQ disturbances develop and maintain smart grids effectiveness. Voltage sags are the most common events that affect power quality. These faults are also the most costly. This paper represents performance comparisons of different computer intelligence methods for voltage sag identification. PQube Analyzer which is installed in Ondokuz Mayis University Computer Laboratory for collecting real time disturbances data for each three phases in order to test for proposed algorithms. Firstly, we used Hilbert Huang Transform to genarate Instantaneous Amplitude (IA) feature signal. Then Characteristic features are attained from IA. The 4 features, mean, standard deviation, skewness, kurtosis of IA are calculated. Support Vector Machines (SVMs) and C4.5 Decision Tree methods are conducted for classification of the disturbance. Secondly we used Fishers Discriminant Ratio for selecting statistical features such as mean, standard deviation, skewness and kurtosis of the normal and voltage sag signals for this part K Means Clustering Method were performed for classification of the disturbance. Consecuently, SVMs, C4.5 Decision Tree and K Means Clustering Methods were performed also their achievements were matched for error rates and CPU timing.