摘要
基于最小二乘辨识原理,针对快时变系统,提出了一种将遗忘因子算法与限定记忆算法相结合的辨识算法,该方法在固定记忆长度中加入遗忘因子,并根据估计参数的变化率来在线调节遗忘因子的大小,提高了时变系统的辨识精度,数值仿真结果表明该方法有较好的辨识效果。
Based on the least square estimation theory, the authors present an identification method for fast time varying system, which combines forget factor with restricted memory length. When we added forget factor into an arithmetic of restricted memory length, and adjusted forget factor online according to the varying speed of the estimated parameters, the estimation precision was improved greatly. We made simulation for a one-rank system, the conclusion was that this algorithm may get good identification effect for those fast time varying system.
出处
《电光与控制》
北大核心
2006年第1期48-49,66,共3页
Electronics Optics & Control
关键词
遗忘因子
限定记忆法
时变系统
forget factor
restricted memory length
time varying system