基于1DCNN-BiLSTM-BiGRU的电能质量扰动分类方法OA
The classification method of power quality disturbance based on 1DCNN-BiLSTM-BiGRU
为了应对电能质量扰动(PQD)识别中噪声干扰导致的识别率下降问题,本文提出一种基于一维卷积神经网络(1DCNN)-双向长短期记忆(BiLSTM)网络-双向门控循环单元(BiGRU)的 PQD 分类方法.该方法首先借助 1DCNN 有效地提取原始信号的浅层局部特征,然后通过BiLSTM和BiGRU组合模块对时序信息和上下文关系进行深入处理,从而实现深层时序特征的提取.最后,将所提取的特征经分类模块用于PQD识别.仿真结果表明,与传统方法相比,本文所提方法在准确性方面更具优势,且抗噪声能力更强.
To address the issue of reduced recognition accuracy in identifying power quality disturbance(PQD)due to noise interference,this paper introduces a PQD classification method based on one-dimensional convolutional neural network(1DCNN)-bidirectional long short-term memory(BiLSTM)-bidirectional gated recurrent unit(BiGRU).This method initially utilizes 1DCNN to effectively extract shallow local features from the raw signals.Subsequently,it employs a combination of BiLSTM and BiGRU modules to delve deeper into temporal information and contextual relationships,facilitating the extraction of deep temporal features.Finally,the extracted features are input to the classification module for PQD recognition.Simulation results show that the proposed method has better accuracy and stronger noise resistance.
王立辉;柯泳;苏如开
佛山科学技术学院机电工程与自动化学院,广东 佛山 528000
电能质量一维卷积神经网络(1DCNN)双向长短期记忆(BiLSTM)网络双向门控循环单元(BiGRU)
power qualityone-dimensional convolutional neural network(1DCNN)bidirectional long short-term memory(BiLSTM)bidirectional gated recurrent unit(BiGRU)
《电气技术》 2024 (005)
51-56,64 / 7
国家自然科学基金项目(62271199)
评论