面向涡轮的PCA-POA-LSTM数据驱动建模及故障预警方法  

PCA-POA-LSTM data-driven modeling and fault warning method for turbine systems

作　　者：刘斌[1] 白红艳 何璐瑶[1] 张晓北 田野 杨理践[1,2] Liu Bin;Bai Hongyan;He Luyao;Zhang Xiaobei;Tian Ye;Yang Lijian(School of Information Science and Engineering,Shenyang University of Technology,Shenyang 110870,China;Shenyang Lujie Pipeline Inspection Co.,Ltd,Shenyang 110027,China;Pipeline Group(Xinjiang)Joint Pipeline Co.,Ltd,Urumqi 830011,China)

机构地区：[1]沈阳工业大学信息科学与工程学院,沈阳110870 [2]沈阳路杰管道检测有限公司,沈阳110027 [3]管网集团(新疆)联合管道有限责任公司,乌鲁木齐830011

出　　处：《电子测量与仪器学报》2025年第1期145-155,共11页Journal of Electronic Measurement and Instrumentation

基　　金：国家自然科学基金(62371315);国家自然科学基金(62301341);辽宁省自然科学基金计划(2023-BS-126);辽宁省应用基础研究计划(2022JH2/101300244)项目资助。

摘　　要：针对传统LSTM数据驱动模型存在输入参数规模过大导致运算负担过大、超参数选择不当和涡轮系统故障发生频率、运维成本高的问题,提出一种基于PCA-POA-LSTM的涡轮数据驱动建模方法,并结合滑动窗口法实现了涡轮故障预警。首先,应用PCA降维技术,减少输入数据维度;其次,采用POA参数寻优方法选出最优超参数组合;然后,利用LSTM算法预测涡轮的输出参数;最后,在PCA-POA-LSTM涡轮数据驱动模型预测结果的基础上,结合滑动窗口法对涡轮故障进行预警,通过窗口内标准差定义报警阈值,攻克了涡轮故障预警的难题。结果表明,以PCA-POA-LSTM为基础的涡轮数据驱动建模实现了较高的精确度,平均绝对百分比误差均在0.396以下,平均绝对误差均在0.809以下,平均方根误差均在1.387以下。并且故障预警方法,至少可提前173个监测点发出故障预警信号,实现了对涡轮故障预警的目的,为未来开展涡轮健康管理提供了理论依据和技术支持。In response to the issues such as the overly large scale of input parameters in the traditional LSTM data-driven model,which leads to an excessive computational burden,improper selection of hyperparameters,high frequency of turbine system failures,and high operation and maintenance costs,a turbine data-driven modeling approach based on PCA-POA-LSTM is proposed,and the turbine fault early warning is achieved by combining with the sliding window method.Firstly,the PCA dimensionality reduction technique is applied to reduce the dimension of the input data.Secondly,the POA parameter optimization method is adopted to select the optimal combination of hyperparameters.Then,the LSTM algorithm is utilized to predict the output parameters of the turbine.Finally,based on the prediction results of the PCA-POA-LSTM turbine data-driven model,the turbine faults are warned by combining with the sliding window method,and the alarm threshold is defined by the standard deviation within the window,thus conquering the difficulty of turbine fault early warning.The results indicate that the turbine data-driven modeling based on PCA-POA-LSTM achieves a relatively high accuracy,with the average absolute percentage error all below 0.396,the average absolute error all below 0.809,and the average root mean square error all below 1.387.Moreover,the fault early warning method can issue a fault early warning signal at least 173 monitoring points in advance,achieving the purpose of turbine fault early warning and providing theoretical basis and technical support for the future development of turbine health management.

关 键 词：涡轮 鹈鹕优化算法 长短期记忆网络 主成分分析 数据驱动 

分 类 号：V235.13[航空宇航科学与技术—航空宇航推进理论与工程] V240.2[电子电信—物理电子学] TN06