自适应神经模糊推理系统在电力故障重现中的应用

将电力故障录波数据重现为实际波形对于继电保护测试和保护动作行为分析等具有重要意义。文章将自适应神经模糊推理系统和数字闭环修正技术应用于电力系统故障重现装置中,实现了整体数字域内的闭环控制,利用输出端回采数据与原始数据进行比较并修正信号源的方法极大地减小了故障重现的非线性误差。Matlab仿真和基于该算法的故障重现装置的实际应用证明了自适应神经模糊推理在故障重现中应用的可行性和有效性。

Application of Adaptive Neuro-Fuzzy Inference System in Power System Fault Recurrence

It is of significance for the testing and action analysis of protection devices to reproduce the actual waveforms by use of fault record data of power system. Applying the adaptive neuro-fuzzy inference system and digital closed-loop modification technology to power system fault recurrence devices, the closed-loop control in whole numeric area is implemented; by use of comparing the data sampled at output terminal with original data and modifying the signal source, the nonlinear error of the fault recurrence is greatly reduced. The simulation results by Matlab and the results from the practical applications of the fault recurrence devices based on this algorithm prove that it is feasible and effective to apply adaptive neuro-fuzzy inference in fault recurrence.