基于TimeGAN的电液伺服阀性能评估模型  

Performance Evaluation Model of Electro-hydraulic Servo Valve Based on TimeGAN

在线阅读下载全文

作  者:余萍 孙嘉徽 谢周鹤 张超[2] YU Ping;SUN Jiahui;XIE Zhouhe;ZHANG Chao(Shanghai Marine Equipment Research Institute,Shanghai 200031;School of Automation Science and Electrical Engineering,Beihang University,Beijing 100191)

机构地区:[1]中国船舶重工集团公司第七○四研究所,上海200031 [2]北京航空航天大学自动化科学与电气工程学院,北京100191

出  处:《液压与气动》2024年第10期170-179,共10页Chinese Hydraulics & Pneumatics

摘  要:电液伺服阀容易受到油液污染的影响,当污染颗粒通过节流口时会对阀口壁面产生冲蚀磨损,使得电液伺服阀阀口过流断面变大,内泄漏量增加。针对现有仿真手段难以准确衡量电液伺服阀的性能退化,而试验样本有限等问题,提出一种基于时间序列生成对抗网络算法的评估模型用于评估电液伺服阀性能。首先分析了电液伺服阀的失效机理,然后建立了时间序列生成对抗网络算法的框架,利用维纳过程分析并得到了电液伺服阀的性能评估模型,最后设计并完成了污染磨损试验,利用时间序列生成对抗网络算法对试验数据扩充,通过参数估计的结果对比发现,利用时间序列生成对抗网络算法对试验数据扩充可以有效提升电液伺服阀性能评估模型的准确度。Electro-hydraulic servo valves are easily affected by oil pollution.When pollutants pass through the throttle,they will cause erosion and wear on the valve wall,resulting in a larger flow cross-section and increased internal leakage of the electro-hydraulic servo valve.A timeseries generative adversarial network(TimeGAN)algorithm based evaluation model is proposed to assess the performance degradation of electro-hydraulic servo valves,which is difficult to accurately measure with existing simulation methods and has limited experimental samples.Firstly,the failure mechanism of the electro-hydraulic servo valve is analyzed,and then the framework of the TimeGAN algorithm is established.The Wiener process is used to analyze and obtain the performance evaluation model of the electro-hydraulic servo valve.Finally,the pollution wear test is designed and completed.The TimeGAN algorithm is used to expand the test data.Through the comparison of parameter estimation results,it is found that using the TimeGAN algorithm to expand the test data can effectively improve the accuracy of the performance evaluation model of the electro-hydraulic servo valve.

关 键 词:电液伺服阀 冲蚀磨损 内泄漏 时间序列生成对抗网络 性能评估 

分 类 号:TH137[机械工程—机械制造及自动化]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象