一种可同时估计传递函数阶次和参数的沃尔什函数建模法

A Walsh Function Modeling Method which Estimates the Parameter and Order of Transfer Function Simultaneously

引入沃尔什函数，可以直接由传感器的时域动态校准数据得出其连续传递函数的系数。本文研究了采用沃尔什函数的最小二乘建模方法，通过合理地构造待估计参数向量，可以在模型参数确定之前得到各阶次对应的最小残差平方和以便确定阶次，并同时可求解出相应的模型参数。求解过程中采用了Ｈｏｕｓｅｈｏｌｄｅｒ变换，减小了计算量。

By introducing Walsh function Into dynamic modeling,the transfer function coefficient can be obtained from the dynamic calibrating data directly. This paper presented a least square modeling method based on Walsh function. By constructing the parameter vector reasonably,the square sum of residuals used in determining the order can be obtained before the getting of parameter,and the relative model parameter can also be calculated. The Householder transform is used during the calculation,which can reduce the calculating quantity obviously.