论文部分内容阅读
基于解耦辨识和多步计算思想,本文提出了一种状态空间模型参数辨识的多步算法.计算法包括二步加权最小二乘法、一步输出信息序列的修正和一步自适应 Kalman滤波过程,具有全局收敛且对待辨识参数初始估值设置不敏感等特点.仿真结果表明,在受控系统承受测量噪声和过程噪声的情况下,该算法对线性状态空间模型的参数辨识是十分有效的.
Based on the idea of decoupling identification and multi-step calculation, a multi-step algorithm for parameter identification of state space model is proposed in this paper. The calculation method includes two-step weighting least square method, one-step output information sequence modification and one-step adaptive Kalman filter. Global convergence and insensitive to the initial estimation of the parameter to be identified.The simulation results show that the proposed algorithm is very effective in the parameter identification of the linear state-space model when the controlled system is subjected to measurement noise and process noise.