基于GRU-Dropout网络的避雷器阻性电流预测研究
Study of Lightning Arrester Resistive Current Prediction Based on GRU-Dropout Network
刘谱辉 1周国平 1周逸鹏1
作者信息
- 1. 南京林业大学 信息科学技术学院,南京 210018;南京林业大学 人工智能学院,南京 210018
- 折叠
摘要
MOA(氧化锌避雷器)阻性电流在设备过电压时具有引导、分流和保护的作用,通过对其进行准确预测能够有效保证电力系统的安全与稳定.针对部分单一网络过拟合、泛化能力不足等弊端提出了一种基于GRU(门控循环单元)结合Dropout(随机失活)的MOA阻性电流预测方法.结合变电站实际电流数据,并设置合适的超参数对网络进行训练和分析.实验结果表明,采用GRU-Dropout网络相较于LSTM提高了泛化能力,减少了过拟合;相较于RNN和BP神经网络提高了精度,其最优平均绝对误差和均方根误差为1.98%和2.73%.在实际应用中能够较为准确地预测MOA对地绝缘趋势,具有一定的应用价值.
Abstract
MOA(metal oxide arrester)resistive current has the role of guiding,shunting and protecting during equip-ment overvoltage,and its accurate prediction can effectively ensure the safety and stability of the power system.A MOA resistive current prediction method based on GRU(gated recurrent unit)combined with Dropout(stochastic deac-tivation)is proposed to address the drawbacks of overfitting and insufficient generalization ability of some single net-works.The network is trained and analyzed by combining the actual current data of the substation and setting ap-propriate hyperparameters.The experimental results show that the use of GRU-Dropout network improves the general-ization ability and reduces overfitting compared with LSTM.Improves the accuracy compared with RNN and BP neu-ral networks,with its optimal average absolute error and root mean square error of 1.98%and 2.73%.It can predict the trend of MOA insulation to ground more accurately in practical applications,and has certain application value.
关键词
氧化锌避雷器/阻性电流/门控神经网络/随机失活Key words
metal oxide arrester(MOA)/resistive current/gated neural network/stochastic deactivatio引用本文复制引用
基金项目
国家自然科学基金资助项目(32171788)
出版年
2024
NETL
NSTL
万方数据