The system identification for multivariable system in frequency domain

In this paper, a new technique for multivariable process identification in the frequency domain based on a simple test is presented. Step changes in the controller set point are performed on the process and the process frequency response matrix is estimated with signal decomposition and analysis method in the frequency domain. Consequently, an accurate transform function matrix can be also identified by matching the model in the frequency range of interest. The significance of the proposed method is that it only requires response data of process input and output, and no prior knowledge of the process dynamics is needed. Simulation examples are given to show both effectiveness and practicality of the identification method.