基于RBF神经网络的导弹备件需求量预测仿真

Simulation on Missile Spares Demands Prediction Based on RBF Neural Network

针对导弹部署后期与备件有关的各项保障数据相对较多的情况,提出采用径向基函数(Radial Basis Function,RBF)网络方法。以某型导弹的某备件为预测对象,对导弹维修备件需求影响因素进行分析,介绍RBF网络的结构、工作原理及预测步骤和流程图,并进行仿真结果分析。分析结果表明,该方法比普通前向网络训练省时,能解决备件需求量的预测问题。

There are plenty of factors affecting a missile fittings demands, introduce the RBF method. Adopt the certain type missile as the forecasting object, and analyze the factors influencing missile maintenance spare part, introduce the structure, working principle, forecasting steps and flow chart of RBF network, and analyze the simulation result. The analyzing result shows that the method can use less time to solve the forecasting method of spare part requirement than ordinary RBF network.