基于动态神经网络的装备技术状态预测模型

装备的技术状态受多种因素的影响,诸多影响因素之间是一种多变量、强耦合、非线性的关系,同时这种关系还是动态的。根据装备技术状态的特性,使其在相空间里重构,然后利用动态神经网络建立装备技术状态预测模型,并以装备振动信号预测为例进行案例研究,验证了利用动态神经网络进行预测的可行性和优越性。

Forecasting Model of Equipment Technique Condition Based on Dynamic Neural Network

The technique condition of military equipment is influenced by various factors.The relation of these factors is variable,strong coupling,non-linear and dynamic.The forecasting of equipment technique condition is an important area of condition-based maintenance study,which provides an important basis for equipment maintenance.According to its characteristics,the equipment technique condition is reconstructed in the phase space and then the forecasting model is built up with dynamic neural network.The case study of equipment vibration signal forecasting shows the feasibility and superiority of forecasting with dynamic neural network.