基于时间序列的发电机设备异常分析
Abnormal Analysis of Generator Equipment Based on Time Series
陆钊 1龙法宁 1陈国年1
作者信息
- 1. 玉林师范学院,广西 玉林 537000
- 折叠
摘要
为提高发电机组设备运行维护管理水平,提出一种基于PCA-Informer方法的发电设备故障预测技术.首先使用主成分分析(PCA)算法降低时间序列数据的特征维度;其次将数据分批次输入Encoder中,由Encoder执行蒸馏操作,提取长时间序列输入间的Long-Range依赖,通过蒸馏操作为重要特征赋予更高的权重,并在下一层生成聚焦的Self-Attention Feature Map;最后由Decoder通过一个正向过程一步生成长序列输出.通过实验验证,该方法能够有效地对发电设备的故障进行预测.
Abstract
To improve the level of maintenance and management of power generation equipment operation,a fault prediction technology for power generation equipment based on PCA-Informer method is proposed.Firstly,it uses Principal Component Analysis(PCA)algorithm to reduce the feature dimension of time series data.Secondly,the data is inputted into the Encoder in batches,and the Encoder performs distillation operations to extract Long-Range dependencies between long time series inputs.The important features are given higher weights through distillation operation,and a focused Self-Attention Feature Map is generated in the next layer.Finally,the Decoder generates a long sequence output by one-step reaction through a forward process.Experimental results show this method can effectively predict the faults of power generation equipment.
关键词
发电机设备/主成分分析/Informer/故障预测Key words
generator equipment/PCA/Informer/fault prediction引用本文复制引用
基金项目
玉林市科学研究与开发计划(玉市科20202925)
出版年
2024
国家科技期刊平台
NSTL
万方数据