摘要
为了实时、准确、可靠地预测轨姿控发动机试验中推力矢量的变化情况,本文在分析支持向量机(SVM)预测算法缺陷的基础之上,提出并建立了一种自适应能力较强的故障预测模型——ARVM(Adaptive Relevance Vector Machine),并将其应用于某型轨姿控发动机高模试验推力矢量参数预测中.研究结果表明,ARVM预测模型在稀疏性和算法精度方面均体现出较高的优越性,能够很好地预测轨姿控发动机试验推力矢量的变化趋势.
In order to predict the variation trend of thrust vector in orbit and attitude engine test, a new prediction algorithm mode, which is named Adaptive Relevance Vector Machine （ARVM）, was pro- posed and set up based on the analysis of SVM algorithm. This ARVM mode was applied to the thrust vector prediction in a certain type of orbit and attitude engine test. The research results indicate that ARVM mode has superiority in the sparsity and precision of algorithm, and ARVM can predict the vari- ation trend of thrust vector in orbit and attitude engine test.
作者
陈文丽
马军强
杨思锋
田国华
李志刚
CHEN Wenli MA Junqiang YANG Sifeng TIAN Guohua LI Zhigang(Academy of Aerospace Propulsion Technology, Beijing Institute of Aerospace Testing Technology, Beijing 100074, China)
出处
《测试技术学报》
2017年第2期114-119,共6页
Journal of Test and Measurement Technology
