Main Content

估计Polynomial Models in the App

Prerequisites

  1. 在系统标识应用程序中,选择估计>Polynomial Modelsto open the Polynomial Models dialog box.

    有关对话框中选项的更多信息,请单击Help

  2. In theStructure列表,从以下选项中选择要估算的多项式模型结构:

    • ARX:[NA NB NK]

    • ARMAX:[na nb nc nk]

    • OE:[NB NF NK]

    • BJ:[NB NC ND NF NK]

    This action updates the options in the Polynomial Models dialog box to correspond with this model structure. For information about each model structure, see什么是多项式模型?

    笔记

    对于时间序列数据,仅可用AR和ARMA模型。有关估计时间序列模型的更多信息,请参见Time Series Analysis

  3. In the订单字段,指定模型订单和延迟,如下:

    • For single-output polynomial models, enter the model orders and delays according to the sequence displayed in theStructurefield. For multiple-input models, specifynbandnk作为具有输入元素的许多元素的行向量。如果您正在估计BJ和OE模型,则还必须指定nf作为矢量。

      For example, for a three-input system,nb[1 2 4], where each element corresponds to an input.

    • 对于多出输出模型,请按照所述输入模型订单多项式多项式模型的多项式大小和顺序

    小费

    To enter model orders and delays using the Order Editor dialog box, clickOrder Editor

  4. (ARX models only) Select the estimation方法作为arxorIV(instrumental variable method). For information about the algorithms, seePolynomial Model Estimation Algorithms

  5. (仅ARX,ARMAX和BJ模型)检查Add noise integration复选框将集成器添加到噪声源,e

  6. Specify the delay using the输入延迟编辑框。该值必须是长度等于数据中输入通道数量的矢量。对于离散时间估计(使用具有非零样品时间的数据的任何估计),必须在滞后数中表达延迟。这些延迟与由“模型内”延迟分开nkorder in the订单编辑框。

  7. In the姓名field, edit the name of the model or keep the default.

  8. In the重点列表,选择how to weigh the relative importance of the fit at different frequencies. For more information about each option, seeAssigning Estimation Weightings

  9. In theInitial statelist, specify how you want the algorithm to treat initial conditions. For more information about the available options, seeSpecifying Initial Conditions for Iterative Estimation Algorithms

    小费

    If you get an inaccurate fit, try setting a specific method for handling initial states rather than choosing it automatically.

  10. In the协方差列表,选择估计如果您希望算法计算参数不确定性。这种不确定性的影响显示在图作为模型置信区域上。

    To omit estimating uncertainty, selectNone。跳过大型多输出模型的不确定性计算可能会缩短计算时间。

  11. 点击正则化获得模型参数的正则估计。在“正则化选项”对话框中指定正规化常数。要了解更多,请参阅模型参数的正则估计值

  12. (ARMAX, OE, and BJ models only) To view the estimation progress in the MATLAB Command Window, select theDisplay progress复选框。这启动了一个进度查看器窗口,其中报告了估计进度。

  13. 点击估计to add this model to the Model Board in the System Identification app.

  14. (仅预测错误方法)停止搜索并在当前迭代完成后保存结果,请单击停止迭代。To continue iterations from the current model, click the继续迭代按钮将当前参数值分配为下一个搜索的初始猜测。

Assigning Estimation Weightings

You can specify how the estimation algorithm weighs the fit at various frequencies. In the app, set重点到以下选项之一:

  • 预言- 使用噪声模型的倒数H权衡如何在各种频率范围内拟合数据的相对重要性。对应于最大程度地减少一步预测的预测,这通常在短时间间隔内有利于拟合度。针对输出预测应用进行了优化。

  • Simulation- 使用输入频谱在特定频率范围内权衡拟合的相对重要性。不使用噪声模型来权衡如何在各种频率范围内拟合数据的相对重要性。针对输出模拟应用进行了优化。

  • Stability- 估计最佳稳定模型。有关模型稳定性的更多信息,请参见不稳定的模型

  • 筛选— Specify a custom filter to open the Estimation Focus dialog box, where you can enter a filter, as described inSimple Passband FilterorDefining a Custom Filter。这个前置过滤只适用于估算thedynamics from input to output. The disturbance model is determined from the unfiltered estimation data.

Next Steps

相关示例

更多关于