论文部分内容阅读
首次提出了时变迭合ＡＲ模型，该模型在实际应用中具有广泛的应用价值．应用两步最小二乘法和限定记忆递推最小二乘法，给出了模型中时变参数的递推估计算法，该算法仅依靠量测数据即能自适应进行．仿真计算及应用结果表明：算法能够自适应地跟踪量测数据模型参数的变化，效果是令人满意的．
For the first time, the time-varying and overlapping AR model is proposed. The model has wide application value in practical application. Applying the two-step least square method and the limited memory recursive least square method, the recursive estimation algorithm of time-varying parameters in the model is given. The algorithm can be adaptively implemented only by measuring the data. The simulation results and the application results show that the algorithm can adaptively track the changes of the parameters of the measured data model, and the results are satisfactory.