本文分析了时间序列建模方法用于处理动态数据,建立系统模型与进行预测控制的特点及问题;研究了时间序列方法建模与系统辨识方法建模以及滤波理论的内在联系;应用系统辨识方法与滤波理论的成果改进了文献[5]采用的参数估计方法,并提出了进一步提高建模速度与预测控制速度的措施。 In this paper, the characteristics and problems of time series modeling method for processing dynamic data, establishing system model and predictive control are analyzed. The internal relations between time series method modeling and system identification method modeling and filtering theory are studied. By using system identification method Compared with the results of filtering theory, the method of parameter estimation adopted in [5] is improved and some measures to further improve the modeling speed and the predictive control speed are proposed.