Bias compensation based partially coupled recursive least squares identification algorithm with forgetting factors for MIMO systems: Application to PMSMs

Aiming at a class of systems with multiple-input multiple-output (MIMO) output-error, this paper proposes a novel bias compensation based partially coupled recursive least squares (RLS) algorithm with forgetting factors, based on the coupled identification concept and the bias compensation technique. The proposed algorithm can not only give the unbiased estimates of the system parameters in the presence of colored noises, but also improve the tracking capability of the time-varying parameters. Additionally, complex matrix inversion is avoided in the proposed algorithm, which is required in the multivariable RLS algorithm to identify MIMO systems. The analysis indicates that the proposed algorithm requires less computational burden than the traditional multivariable RLS algorithm. Finally, the proposed algorithm is tested on a surface permanent-magnet synchronous motor example, the efficiency of which is demonstrated by the experimental results.