基于LSTM-ARIMA算法的发电机定子线棒出水温差预测  

Prediction of Outlet-water Temperature Difference of Generator Stator Bars based on LSTM-ARIMA Algorithm

作　　者：陈聪 王晓剑 徐俊元 胡磊 何天磊 梁辰 CHEN Cong;WANG Xiaojian;XU Junyuan;HU Lei;HE Tianlei;LIANG Chen(China Power Huachuang Electricity Technology Research Co.,Ltd.,Shanghai 200086,China;China Power Huachuang(Suzhou)Electricity Technology Research Co.,Ltd.,Suzhou 215123,China)

机构地区：[1]中电华创电力技术研究有限公司,上海200086 [2]中电华创(苏州)电力技术研究有限公司,江苏苏州215123

出　　处：《大电机技术》2023年第5期43-48,共6页Large Electric Machine and Hydraulic Turbine

基　　金：国家电力投资集团中国电力国际发展有限公司科技项目(2020-005-ZGDL-KJ-X)。

摘　　要：对定子线棒出水温度的最大温差进行预测,对于保障汽轮发电机的安全运行具有重要意义。但由于发电机运行过程中工况多变,温差时间序列变化模式复杂,因此趋势预测相对困难。本文使用长短时记忆(LSTM)神经网络对复杂的变化模式进行学习,并进一步融合了差分整合移动平均自回归(ARIMA)模型,用以弥补工况多变导致的训练不足的问题,从而对LSTM神经网络的预测结果进行修正。然后,在型号为QFSN-660-2-22的汽轮发电机运行数据上开展了实验,结果表明该方法预测效果优于单独的LSTM神经网络和ARIMA模型算法,并且可用于短期预警,准确率高于95%。The prediction of the maximum temperature difference of outlet-water of the stator bars is of great significance to ensure the safe operation of turbo-generator.However,the working conditions of turbo-generator are variable,resulting in the complex pattern of temperature difference sequence,which makes it difficult to predict the trend accurately.In this paper,long short-term memory(LSTM)neural network is used to learn complex patterns.And auto-regressive integrated moving average(ARIMA)model is further combined to make up for the lack of training caused by variable working conditions,so as to correct the prediction of LSTM neural network.Experiments are carried out on the operation data of QFSN-660-2-22 turbo-generator.The results show that the prediction effect of the combined method mentioned in this paper is better than the single LSTM network or ARIMA model algorithm.This method can be applied to short-term early warning,of which the accuracy is higher than 95%.

关 键 词：定子线棒 出水温差 温度预测 长短时记忆神经网络 差分整合移动平均自回归模型 时间序列分析 

分 类 号：TM311[电气工程—电机]