T-S fuzzy model identification of MIMO nonlinear systems based on data-driven

For complicated relationship among variables in muti-input muti-output(MIMO) nonlinear system, a fuzzy C-means clustering data-driven algorithm is proposed, then a T-S fuzzy model identification method of MIMO nonlinear systems based on data-driven is presented in this paper. The approach realizes data self-adapting identification of systems fuzzy cluster centers and avoids high randomness and poor convergence of the fuzzy clustering algorithm that caused by giving matrixes of initial membership degrees. The effectiveness of proposed method is validated by the experiment studies of cement rotary kiln control process.