舰船维护中机械潜在故障智能预测方法
|
摘要
舰船机械部件是一个非线性系统,舰船机械部件出现故障概率相当高,当前故障预测方法无法描述舰船机械部件故障的不确性,因此舰船机械部件故障预测精度低,为了提高舰船机械部件故障预测精度,克服当前舰船机械部件故障预测方法的缺陷,设计了一种舰船维护中机械潜在故障智能预测方法。首先提取描述舰船机械部件故障类别的特征信息,然后采用BP神经网络对舰船机械部件故障特征信息进行学习,确定相对应的舰船机械部件故障类别,并解决BP神经网络参数确定问题,最后与其他方法进行了对比实验。结果表明,本文方法的舰船机械部件故障预测精度超过95%,远远高于对比方法的舰船机械部件故障预测精度,改善了舰船机械部件故障诊断速度,具有十分广泛的应用前景。
Marine mechanical components are a non-linear system. The failure probability of ship mechanical components is quite high. When the fault prediction method can not describe the uncertainty of ship mechanical components, the prediction accuracy of ship mechanical components is low. In order to improve the prediction accuracy of ship mechanical components and overcome the shortcomings of current prediction methods for ship mechanical components, a design is made based on intelligent prediction method for potential mechanical failure in ship maintenance. Firstly, the feature information describing the fault types of ship mechanical components is extracted, then the fault feature information of ship mechanical components is learned by using BP neural network, the corresponding fault types of ship mechanical components are determined, and the problem of determining the parameters of BP neural network is solved. Finally, the comparison experiment of ship mechanical components fault prediction with other methods is carried out. The accuracy of fault prediction of mechanical components is over 95%, which is much higher than that of comparison method. It also improves the speed of fault diagnosis of marine mechanical components, and has a very wide application prospect.
Marine mechanical components are a non-linear system. The failure probability of ship mechanical components is quite high. When the fault prediction method can not describe the uncertainty of ship mechanical components, the prediction accuracy of ship mechanical components is low. In order to improve the prediction accuracy of ship mechanical components and overcome the shortcomings of current prediction methods for ship mechanical components, a design is made based on intelligent prediction method for potential mechanical failure in ship maintenance. Firstly, the feature information describing the fault types of ship mechanical components is extracted, then the fault feature information of ship mechanical components is learned by using BP neural network, the corresponding fault types of ship mechanical components are determined, and the problem of determining the parameters of BP neural network is solved. Finally, the comparison experiment of ship mechanical components fault prediction with other methods is carried out. The accuracy of fault prediction of mechanical components is over 95%, which is much higher than that of comparison method. It also improves the speed of fault diagnosis of marine mechanical components, and has a very wide application prospect.
引文
[1]祁贝贝,王延涛.大数据分析在船舶机械故障诊断系统中的应用[J].舰船科学技术, 2019, 41(8):202–204.
[2]朱煜,熊小虎.基于振动分析的船舶旋转机械故障诊断研究[J].江苏船舶, 2016, 33(5):11–14.
[3]陆洋,吴智文,韩文玉,韩莉.共振解调技术在船舶机械故障诊断中的应用[J].船舶工程, 2009, 31(3):31–34.
[2]朱煜,熊小虎.基于振动分析的船舶旋转机械故障诊断研究[J].江苏船舶, 2016, 33(5):11–14.
[3]陆洋,吴智文,韩文玉,韩莉.共振解调技术在船舶机械故障诊断中的应用[J].船舶工程, 2009, 31(3):31–34.