摘要
针对智能预警系统无法快速准确预测出故障数据的问题,将金枪鱼群优化(TSO)算法与麻雀搜索算法(SSA)融合到电网智能预警系统,扩展BP神经网络模型的应用领域。在原有的神经网络模型中,引入SSA算法中的3种数学模型,通过不断地迭代以及系统更新,最终确定最佳权重。所提出的系统不仅克服了原有神经网络训练数据速度慢的问题,提高了模型的训练速度,还增加了系统故障预测的准确率。TSO算法在预警系统中的应用,有助于更好地优化模型参数,提高其在时序序列数据预测中的准确性。实验结果表明,使用SSA训练的神经模型相对于其他系统,故障数据的预测准确率提高了10.28%,从而提高了电网智能预警系统的准确性。
To solve the problem that the intelligent early warning system cannot predict the fault data quickly and accurately,the tuna swarm optimization(TSO)algorithm and sparrow search algorithm(SSA)are integrated into the intelligent early warning system of power grid.The application field of BP neural network model is expanded.In the original neural network model,three kinds of mathematical models in SSA are introduced,and the optimal weight is finally determined through continuous iteration and system update.The proposed system not only overcomes the problem of slow training data speed of the original neural network,improves the training speed of the model,but also increases the accuracy of the system fault prediction.The application of TSO algorithm in the early warning system is helpful to optimize the model parameters better and improve its accuracy in the prediction of time series data.The experimental results show that compared with other systems,the prediction accuracy of fault data of neural model trained by SSA is improved by 10.28%,thus improving the accuracy of power grid intelligent early warning system.
作者
王伟
韦怡
吴政
王凤祥
赵晟
WANG Wei;WEI Yi;WU Zheng;WANG Fengxiang;ZHAO Sheng(Longtan Hydropower Plant of Longtan Hydropower Development Co.,Ltd.,Nanning 530000,China)
出处
《微型电脑应用》
2025年第12期58-61,共4页
Microcomputer Applications
基金
2022年龙滩水电开发有限公司龙滩水力发电厂信息化项目(CDT-LTHPC-E-2395)。
