摘要
本文提出了有截距项指数模型的线性最小二乘辨识方法。证明了所提算法的收敛性和它与非线性规划算法的一致性。数值计算结果表明该方法具有简单易行,基本上无需猜测参数初值,收敛速度较快且与参数个数无关等一系列特点。
In this paper a new linear least-squares identification algorithm for the exponential model with an intersection parameter is proposed. Its convergence and the consistency between it and nonlinear programming approaches are proved. Numerical calculation results demonstrate that the algorithm is quite simple, easy to be used, unnecessary to guess the initial values of the parameter, and that its Convergence rate is very fast and independent of the number of the parameters identified, etc.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
1989年第Z01期67-76,共10页
Control Theory & Applications
