A New Method for Identification of Voltage Sags According to Cause

Voltage sags are probably one of the most important power quality problems because of its impact on malfunctioning electrical equipment in industrial and commercial installations and its high frequency. This fact highlights the need for an effective technique of detection, evaluation and identification for the sags problems. This paper proposed a voltage sags identification method according to cause, based on an improved phase-located-loop (PLL) and radial basis function (RBF) neural network. Characteristics of voltage sag such as magnitude, duration, phase jump, and amplitude shape, were detected out with the proposed PLL. Then, through a data dealing with detecting outputs from the PLL, a set of features were extracted for identification of voltage sags. Finally, a RBF neural network was developed for voltage sags identification. The prototype of the system is built and tested using 300 simulation voltage waveforms which covers five types of sags events widely spread in power system. The test results obtained prove that the method enables accurate identification of these type sags. The identification accuracy even reached 100% without any wrong or rejected identification. Potential applications of the proposed system in power system are also described.