| 基于NARX神经网络方法的汽轮机转子关键部位应力预测 |
| |
| 引用本文： | 赵翔,茹东恒,王鹏,吴昊,甘磊,仲政. 基于NARX神经网络方法的汽轮机转子关键部位应力预测[J]. 应用数学和力学, 2021, 42(8): 771-784. DOI: 10.21656/1000-0887.410372 |
| |
| 作者姓名： | 赵翔 茹东恒 王鹏 吴昊 甘磊 仲政 |
| |
| 作者单位： | 1. 同济大学 航空航天与力学学院， 上海 200092; |
| |
| 基金项目： | 国家自然科学基金（119320051197225511772106）；2019年人工智能创新发展专项（2019 RGZN 01059） |
| |
| 摘 要： | 汽轮机启动过程中,对转子进行应力预测具有重要意义.为满足在线预测国产某350 MW超临界汽轮机转子关键部位应力的需要,提出了一种基于具有外部输入的非线性自回归(NARX)神经网络的应力预测方法.根据转子实际尺寸建立二维轴对称有限元模型,确定了相应的边界条件,并对有限元计算结果进行验证,得到了转子在冷启动工况下的温度场和...
|
| 关 键 词： | 汽轮机转子 NARX神经网络 有限元分析 应力预测 |
| 收稿时间： | 2020-12-07 |
On the Stress Prediction of Key Components in Steam Turbine Rotors Based on the NARX Neural Network |
| |
| Affiliation: | 1. School of Aerospace Engineering and Applied Mechanics, Tongji University, Shanghai 200092, P.R.China;2. Turbine Plant, Shanghai Electric Power Generation Equipment Co. Ltd.,Shanghai 200240, P.R.China;3. School of Science, Harbin Institute of Technology(Shenzhen), Shenzhen, Guangdong 518055, P.R.China |
| |
| Abstract: | Stress prediction of steam turbine rotors during startup processes is of great significance. To predict the stresses of key components in a 350 MW supercritical steam turbine rotor, a NARX neural network-based method was proposed with a 2D axisymmetric finite element model established according to the actual dimensions of the rotor. Appropriate boundary conditions were applied to the model and the temperature and stress distributions under cold startup conditions were calculated. The simulated results were experimentally verified and the danger points of the rotor were then determined after 288 finite element calculations according to typical startup conditions. The stresses calculated near the danger points as well as several user-selected operating parameters were used to establish the neural network sample dataset. An effective NARX neural network was employed to estimate the stresses at the danger points. The results show that, the proposed method can accurately predict the stresses with their tendency. The stresses predicted by the NARX neural network are in good agreement with the finite element simulated results, and can meet the requirements for rotor stress monitoring. |
| |
| Keywords: | |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《应用数学和力学》浏览原始摘要信息 |
|
点击此处可从《应用数学和力学》下载免费的PDF全文 |
|
