摘要
针对多维传感器工作过程中的信息耦合问题,提出了基于多尺度特征的解耦数值计算方法,该方法通过对系统数学模型进行基于多尺度的特征提取,运用Hermitte插值以及分段线性插值方法,根据不同的尺度,逐步缩小系统自由度以完成解耦数值计算。仿真结果表明使用本方法的解耦误差平均为0.08%,对比以往的解耦方法运算精度提高20%,速度提高50%。
According to the information coupling in the function of multi-variant sensor,a computational decoupling method based on multi-scale characteristic is proposed.After the multi-sacle characteristic of the sensing information is evaluated based on the mathematical model of the system,the numerical analysis is carried out according to the scale,in order to reduce the variation of the system step by step,and complete the decoupling computation.The result indicates that this method could get(0.08%) error.Comparing to the traditional method,the precision is raised about 20%,and the speed is improve 50% at average.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2005年第z1期164-167,共4页
Optics and Precision Engineering
基金
广东省科技计划项目(No.2004B10201011)
广州市科技计划项目(No.053Z208008)
关键词
多尺度特征
传感器
解耦
数值计算
multi-scale characteristic
sensor
decoupling
computation method
