摘要： 将支持向量机应用于飞机重着陆的诊断研究。通过分析飞机着陆阶段的运动方程，确定造成飞机重着陆的主要影响因素，将传统的单一指标诊断扩展到多指标诊断。提出一种多核支持向量机，并在此基础上建立飞机重着陆诊断模型。与传统支持向量机和神经网络模型比较表明，该模型精度高，具有更强的泛化能力。

关键词: 重着陆, 诊断模型, 多核支持向量机, 核函数

Abstract: Support Vector Machine(SVM) is used to diagnose airplane hard landing. According to airplane’s motion equation of landing phase, major influencing factors are determined and extended airplane hard landing diagnosis from one index to several. Multi-kernel Support Vector Machine(MKSVM) is proposed, on this basis, airplane hard landing diagnosis model is established. Compared with traditional SVM and BP neural networks, the MKSVM diagnosis model is feasible and accurate, has high generalization ability.

Key words: hard landing, diagnosis model, Multi-kernel Support Vector Machine(MKSVM), kernel function

中图分类号: 

• V328