摘要
利用变点统计学和黄金分割法讨论具有跳变参数的多模型离散线性动态系统的参数估计。提出基于黄金分割法搜索最佳变点估计（参数变化时间）和同时得到参数估计的最小二乘算法。理论分析和计算机仿真结果表明，该算法具有较好的收敛性并能得到良好的变点及参数的估计值。
The change-point statistics and the golden section method is used to discuss the parameter estimation of multiple-model discrete linear systems with jumping parameters in this paper, and then the least squares algorithm which can obtain the optimal estimation of the change-points and the parameters based on the golden section method is proposed. The theoretical analysis and the simulation results show that the algorithms in this paper have good convergence and the good estimation of the change-points and parameters.
出处
《控制与决策》
EI
CSCD
北大核心
1995年第4期352-356,共5页
Control and Decision
关键词
多模型系统
黄金分割法
参数估计
离散系统
change-point statistics, parameter estimation, the golden section method, the least squares method
