基于LS-SVM的装甲装备故障预测模型研究

LS-SVM-Based Research on Fault Prognostics Model of Armored Equipment

针对装甲装备故障信息不足、故障预测困难等问题,通过对装甲装备的故障特点和支持向量机回归算法的分析,利用最小二乘支持向量机LS-SVM建立故障预测模型,并利用某型装甲发动机进气门盘部局部断裂故障数据对故障预测模型进行了验证.事实证明,基于最小二乘支持向量机建立故障预测模型能够较好地对装甲装备故障的趋势进行预测.

To solve the problems like insufficient information on armored equipment fault and difficulty in predicting the equipment fault, this paper analyses the fault characteristics of the armored equipment and the support vector machine regression algorithm and establishes a fault prediction model by using the least squares support vector machine （LS-SVM）, and the fault prediction model is verified by using local fracture failure data of a certain type of armored equipment inlet valve plate. It is proved that the proposed model can well predict the trend of the armored equipment fault.