基于MI-GA-BP和误差统计分析的水电机组健康评估模型研究 |
点此下载全文 |
引用本文:傅质馨1,2,曹延1,朱俊澎1,2,袁越1,2.基于MI-GA-BP和误差统计分析的水电机组健康评估模型研究[J].电网与清洁能源,2021,37(7):97~106 |
摘要点击次数: 108 |
全文下载次数: 0 |
|
基金项目:国家自然科学基金青年项目(51807051); 江苏省自然科学基金青年项目(BK20180507) |
|
中文摘要:针对水电机组运行期间故障样本较少,难以构建故障诊断模型的问题,提出基于MI-GA-BP和误差统计分析的方法构建水电机组健康评估模型。首先通过互信息理论选择相关的多个工况参数,进而采用GA-BP神经网络建立水电机组的振动预测模型,确定健康评估模型的基准值。然后通过对振动相对预测误差进行统计分析,采用非参数核密度估计法和正态分布估计法分别拟合其概率密度函数,计算在一定置信水平下的置信区间,采用熵权法对2种方法得到的区间进行组合得到最优区间,得到健康评估模型的界限值。通过仿真分析表明,所构建的健康评估模型能够实时反映机组的健康状态,判断其是否处于异常状态。 |
中文关键词:水电机组 健康评估 互信息 GA-BP 误差分析 |
|
Research on Hydropower Unit Health Assessment Model Based on MI-GA-BP and Error Statistical Analysis |
|
|
Abstract:Given that there are few fault samples during the operation of hydropower generating units and it is difficult to construct a fault diagnosis model, a method based on MI-GA-BP and error statistical analysis is proposed to construct a health assessment model for hydropower generating units. First, multiple relevant working condition parameters are selected through mutual information theory, and then GA-BP neural network is used to establish the vibration prediction model of the hydropower unit to determine the reference value of the health assessment model. Second, through the statistical analysis of the relative vibration prediction error, the non-parametric kernel density estimation method and the normal distribution estimation method are respectively fitted to the probability density function, and the confidence interval under a certain confidence level is calculated. The entropy method is used to combine the intervals obtained by the two methods to obtain the optimal interval, and then obtain the limit value of the health assessment model. Simulation analysis shows that the constructed health assessment model can reflect the health status of the unit in real time and judge whether it is in an abnormal state. |
keywords:hydropower unit health assessment mutual information GA-BP error analysis |
查看全文 查看/发表评论 下载PDF阅读器 |