摘要
随着移动电源技术的快速发展,移动电源运行管理的智能化日益受到重视。文章针对传统电源故障事后维修的问题,提出了基于相似组推荐深度神经网络的电源故障预测方法。首先,基于电源参数时序信息特征,构建了相似组推荐深度神经学习网络;其次,构建了故障的相似性矩阵、故障置信图以及故障类别预测模块,实现了电源故障分组的语义推荐;最终,基于知识数据库,利用分组与融合操作预测电源健康状态,实现了电源故障的有效智能预警。
With the rapid development of mobile power supply technology,the intelligence of mobile power supply operation and management has been paid more and more attention.In this paper,a power supply fault prediction method based on similar group recommendation deep neural network is proposed for the problem of traditional power supply fault maintenance afterwards.Firstly,a similar group recommendation deep neural learning network is constructed based on the transformed power supply parameter timing information features;secondly,the similarity matrix of faults,fault confidence map and fault category prediction module are constructed to realize the semantic recommendation of power supply fault grouping.Finally,based on the knowledge database,the grouping and fusion operations are used to predict the power supply health status and realize the effective intelligent warning of power supply faults.
作者
李渭
刘少明
余锐
王学超
景元辉
LI Wei;LIU Shaoming;YU Rui;WANG Xuechao;JING Yuanhui
出处
《今日自动化》
2023年第6期130-133,共4页
Automation Today
基金
江西省自然基金青年项目(20212BAB212012)。
关键词
极坐标变换
相似组推荐
深度网络
电源健康管理
polar coordinate transformation
similar group recommendation
deep network
power supply health assessment
