基于RBF神经网络的装备预防性维修时间间隔预测

Preventive maintenance interval prediction of equipment based on RBF neural network

针对装备系统维修工作的时间间隔确定问题,提出了基于RBF神经网络分析模型的装备预防性维修时间间隔的方法。通过构建以预防性维修时间、故障修复性维修时间、系统可靠度为输入,以维修时间间隔为输出的RBF神经网络模型,可准确地实现维修时间间隔的确定。将该方法结果与基于双参数威布尔分布规律以及基于最大可用度的预测结果进行了对比,结果表明:所提方法具有更高的预测精度,在针对装备的预防性维修中具有重大意义。

To solve the problem of determining the time interval of equipment system maintenance work,the method of equipment preventive maintenance time interval based on RBF neural network analysis model was proposed.The determination of maintenance intervals can be accurately achieved by constructing an RBF neural network model with preventive maintenance time,fault repair maintenance time and system reliability as inputs and maintenance intervals as outputs.In comparison with those based on the two-parameter Weibull distribution pattern and those based on maximum availabi-lity,the results of this method indicate that its higher prediction accuracy is of great significance in preventive maintenance for equipment.