4th International Conference on Electric Power Equipment– Switching Technology (ICEPE-2017)

518 / 2017-05-11 12:25:32

A fast Current Zeroes Estimation Algorithm for Controlled Interruption Based on an Improved BP Neural Network

13664,6357,14014,14015

Predicting zeroes precisely and rapidly after the fault initiation is the basis of controlled interruption. However, none available algorithms could predict current zeroes within several milliseconds. The objective of this paper is to propose a fast estimation algorithm for current zeroes based on an improved BP neural network. Both waveform of the fault current and initial phase angel of the fault acquired by wavelet transform are set as an input of the neural network. The first current zero is directly set as an output. The neural network was trained by over 10,000 waveforms of fault current with different parameters acquired by the short fault circuit simulation. Results show that the first current zero is estimated within 3 ms after the fault initiation with an error of ±0.5 ms.