摘要
对极限学习机的特点及适用条件进行了探讨,在此基础上提出和实现了一种基于极限学习机的空间配准方法,并与基于广义最小二乘和神经网络的配准方法在多种场景下进行了仿真比较,结果验证了基于极限学习机的空间配准方法的性能优势。
The characteristic and applicability of extreme learning machine have been studied inthis paper, a method of space registration based on extreme learning machine is proposed, it iscompared with the method of sensor registration based on neural network and the method of GLS inmulti-kind parameters. The result illustrate that the method of space registration based on extremelearning machine is effective.
出处
《火力与指挥控制》
CSCD
北大核心
2017年第10期10-13,共4页
Fire Control & Command Control
基金
国家自然科学基金资助项目(61371064)
关键词
空间配准
系统误差
非参数估计
极限学习机
space registration, systematic error, nonparametric estimation, extreme learning machine
