Abstract: A feedback control algorithm is presented, based on adaptive inverse control theory. This approach introduces a e-filter structure, using the desired single mode vector and white disturbance as input, trying to attain an ideal controller by means of the special error filter and error elimination techniques. The simulation calculation result: (1) The Signal-to-Error Ratios (SERs) curves can reach the perfect SNR bound quickly; (2) The controller c can get convergence without needing the accurate replica Q. The results show that the proposed approach is simple, effective, and can eliminate plant disturbance and noise efficiently to excite the desired single mode by using the Multi-Input Multi-Output (MIMO) system consisting of a source array and a feedback array.