Implementation of the neural network for tracing of spot welders

Detection of flicker sources is the first step to mitigate the effect of flicker in power system. In this literature, existence of several flicker sources is studied and proposes a technique for detecting all existing tones in voltage envelope. This technique is based on the d-q transformation. Half wave rectifier is improved by using d-q transformation, to calculate amplitudes of all flicker tones. These amplitudes are considered as index of flicker sources detection. And in this paper, in order to reduce the number of measurement devices a neural network is train by using acquired indexes to identify the place of flicker sources. For validation, the 6-bus network is simulated and algorithm for flicker sources detection is tested. The simulations results show that by using the proposed algorithm, all flicker sources in a power system can be detected correctly.