Online Estimation of Moment of Inertia Incorporating Disturbance Estimation

This article presents an estimation strategy that combines an improved recursive least-squares (RLS) algorithm and the equivalent-input-disturbance (EID) approach to estimate the time-varying moment of inertia (MOI) in a permanent-magnet synchronous motor in a real-time fashion. The main feature of the method is that an MOI estimate is incorporated into the disturbance estimation and an EID estimate is also used in the MOI estimation so as to improve the estimation accuracy of MOI. The improved RLS algorithm balances the estimation accuracy and the convergent speed of the MOI estimation by adjusting a forgetting factor with respect to an asymptotic memory length. Furthermore, the error of disturbance estimation using the EID approach is analyzed theoretically, and the stability of disturbance estimation is guaranteed in terms of the concept of globally uniformly ultimately bounded. Finally, a self-tuning PI speed-control system, in which the PI controller is tuned based on the MOI estimate, is used to verify the effectiveness of the method. Experimental results demonstrate the validity of the method.