The filtering based parameter estimation for two-input single-output bilinear-in-parameter systems

This paper discusses the parameter estimation for a class of two-input single-output bilinear-in-parameter systems with colored noise. Using the over-parameterization model and decomposition technique, the gradient iterative algorithm based on data filtering is derived for the two-input single-output bilinear-in-parameter system. The basic idea is to decompose a two-input single-output bilinear-in-parameter system into several subsystems and estimate the parameters of each subsystem, respectively. The algorithm avoids estimating the correlated noise to enhance the accuracy of parameter estimation, and provides a reference for improving the identification accuracy of multi-input single-output bilinear-in-parameter system with colored noise. The simulation results confirm the effectiveness of the proposed algorithm.