【摘要】 介绍多变量系统辨识在电加热炉系统中的应用。在确定合适的采样周期和输入信号后,获得了可供辨识的输入输出数据,并利用Guidorzi法离线辨识出模型的结构。为了克服有色噪声的影响,在辨识过程中用IV(InstrumentalVariable)法估计出模型的参数。通过对残差的白性检验,验证了所建立模型的可靠性。最后,对IV法和RLS(RecursiveLeastSquare)法的辨识精度进行了比较。更多还原

【Abstract】 To introduce application of MIMO system identification in electric heating furnace.Input and output data used for identification can be acquired after right sampling period and input signals have been determined,Then,structure of model can be identified by use of Guidorzis method offline.In order to avoid influence of colour noise,in identification process,parameters of the model will be estimated by use of IV method.Through checking the white characteristic of residual error,reliability of the model can be proved.Finally,identification precision of IV and RLS (Recursive Least Square) methods are compared.更多还原