电光与控制, 2014, 21 (5): 78, 网络出版: 2014-05-22
导弹退化故障预测方法研究
A Method for Missile Degradation Fault Prediction
导弹 故障预测 状态监测 最小二乘支持向量机 missile fault prediction condition monitoring least square support vector machine
摘要
对导弹退化故障预测方法进行了研究, 对测试数据进行统计推断, 确定了测试数据的分布规律, 并针对导弹测试数据分布参数存在小样本、非线性等特点, 应用最小二乘支持向量机预测算法对测试数据的分布参数进行预测, 确定了测试数据未来某一时刻的分布函数, 进而建立了导弹退化故障预测模型,得出了导弹未来一段时间内的退化故障概率。通过实例分析, 验证了退化故障预测模型的合理性。
Abstract
The method for missile degradation fault prediction is studied.The distribution regularity of test data is obtained by statistical inference.Considering that the distribution parameters of missile test data have the features of small sample and nonlinearitythe prediction algorithm of least square support vector machine is used to predict the distribution parameters of test data in order to get the future distribution function.Further morethe degradation fault prediction model of missile is established to get the future degradation fault probability of missile.The results of example analysis validate the rationality of degradation fault prediction model.
丛林虎, 徐廷学, 杨继坤, 董琪. 导弹退化故障预测方法研究[J]. 电光与控制, 2014, 21(5): 78. CONG Lin-hu, XU Ting-xue, YANG Ji-kun, DONG Qi. A Method for Missile Degradation Fault Prediction[J]. Electronics Optics & Control, 2014, 21(5): 78.