一种基于参数辨识的迭代学习控制算法及应用

摘要

针对间歇过程中模型参数变化的问题,提出了一种基于遗忘因子最小二乘法辨识的迭代学习控制算法.迭代学习律的参数随模型参数变化而更新,利用遗忘因子大大减小参数变化时“错误”数据对算法的影响,使算法具有更强的自适应性.把这一算法应用于黄酒发酵过程,提高了发酵过程的优化控制效果.仿真结果表明当模型参数随着批次变化时,系统的跟踪性能得到了改进.%A new iteration learning control algorithm with parameter identification based on recursive least squares algorithm with a forgetting factor is proposed for model parameters changing in batch process. Parameters of the learning gain are updated as model parameters change. The forgetting factor greatly decreases the impact of wrong data, so the algorithm is more adaptive. Rice wine fermentation is optimized with the new algorithm. The simulation result indicates that the algorithm is much effective and can approach anticipant contrail with less iterative when the parameters update.