摘要
针对一个线性回归模型的系统矩阵存在的随机扰动情况,提出一种基于均匀设计的稳健参数估计算法。仿真结果表明,采用本文算法估计回归模型参数,能较好地抑制随机扰动对模型预测精度的影响,提高模型的稳定性和抗干扰能力,同时为改进仪表测量误差对预测精度带来的影响提供了一条可行途径。
A robust parameters estimation algorithm is proposed in this paper, which is based on uniform design for a linear regression model in the case of its coefficient matrix with random disturbance. Numerical simulations show that using this algorithm to estimate model parameters can restrain the random disturber＇s influence on precision prediction, and improve model stability and anti-disturbance ability. It also becomes a feasible approach to improve the prediction precision when instrmnents show measuring deviation.
出处
《信息与电子工程》
2006年第2期111-115,共5页
information and electronic engineering
基金
NSAF联合基金资助项目(10276005)
国防行业重点预先研究资助项目(421010401-3)
关键词
计算机仿真
稳健估计
均匀设计
回归模型
随机扰动
computer simulation
robust estimation
uniform design
regression model
random disturbance
