Adaptive identification of bilinear systems

The paper considers the adaptive identification of bilinear systems using the equation-error approach. An improved least squares (ILS) objective function is presented to reduce the bias of coefficient estimation in the case of large measurement noise when the standard least squares (LS) technique is used. An adaptive algorithm based on the ILS criterion is proposed for the identification of the bilinear system. Numerical simulations are given to demonstrate the effectiveness of the adaptive ILS algorithm. Compared with the least mean square (LMS) technique, the proposed algorithm has superior identification performance