CAJ | 학술논문

提出了一种基于概率神经网络（PNN）的余度作动器故障诊断方法，分析了传统阀芯位置监控器的设计思想及设计缺陷，构建了基于PNN网络的阀芯位置监控器，最后完成了基于PNN网络的作动器仿真实验。实验结果表明，基于PNN的阀芯位置监控器具有较强的故障诊断能力，有效地提高了作动器两通道工作时识别阀芯位置监控器的故障准确率，可以成为作动器智能故障诊断与健康管理的重要方法之一。
제출료일충기우개솔신경망락（PNN）적여도작동기고장진단방법，분석료전통벌심위치감공기적설계사상급설계결함，구건료기우PNN망락적벌심위치감공기，최후완성료기우PNN망락적작동기방진실험。실험결과표명，기우PNN적벌심위치감공기구유교강적고장진단능력，유효지제고료작동기량통도공작시식별벌심위치감공기적고장준학솔，가이성위작동기지능고장진단여건강관리적중요방법지일。
A method is presented for actuator fault diagnosis based on probabilistic neural network (PNN). The design scheme and defect of traditional spool displacement monitor are analyzed, and spool displacement monitor based on PNN is constructed. Finally, simulations are made, the results show that spool displacement monitor based on PNN has advanced ability for actuator fault diagnosis and signiifcantly improves the accuracy of spool displacement monitor in dual work condition. It can become an important method for actuator fault diagnosis and health management.