计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (29): 201-205.

• 工程与应用 • 上一篇    下一篇

支持向量机抗辐射效应的相关研究

李俊飞,陈晓敏   

  1. 中国科学院 空间科学与应用研究中心,北京 100190
  • 出版日期:2012-10-11 发布日期:2012-10-22

Related study of support vector machines for radiation resistant

LI Junfei, CHEN Xiaomin   

  1. Center for Space Science and Applied Research, Chinese Academy of Sciences, Beijing 100190, China
  • Online:2012-10-11 Published:2012-10-22

摘要: 针对FPGA中实现的支持向量机在航天辐射环境下抗单粒子效应的算法进行了探讨,主要对支持向量机中常用的多项式核和高斯核推导了校验算法,以期实现在满足一定判决结果正确率条件下的检错和容错。仿真结果证明该算法具有一定可行性,将会对支持向量机在航天中的应用起到一定积极作用。

关键词: 算法, 支持向量机, 辐射

Abstract: Algorithms of resisting single particle effect for support vector machine which being implemented in FPGA and applied in space are discussed. The paper mainly designs checkout algorithms for polynomial kernel and Gauss kernel which are mostly used in support vector machine in order to achieve error detection and fault tolerance under the condition of certain judgment accuracy. The results of simulation show that the algorithms are feasible which will play an active role for the application of support vector machine in space.

Key words: arithmetic, support vector machines, radiation