A new procedure to design parameter estimators with enhanced performance is proposed in the technical note. For classical linear regression forms, it yields a new parameter estimator whose convergence is established without the usual requirement of regressor persistency of excitation. The technique is also applied to nonlinear regressions with “partially” monotonic parameter dependence—giving rise again to estimators with enhanced performance. Simulation results illustrate the advantages of the proposed procedure in both scenarios.
19 June 2020 in IEEE Transactions on Automatic Control
We present some new results on the dynamic regressor extension and mixing parameter estimators for linear regression models recently proposed in the literature. This technique has proven instrumental in the solution of several open problems in system identification and adaptive control. The new results include: (i) a unified treatment of the continuous and the discrete-time cases; (ii) the proposal of two new extended regressor matrices, one which guarantees a quantifiable transient performance improvement, and the other exponential convergence under conditions that are strictly weaker than regressor persistence of excitation; and (iii) an alternative estimator ensuring convergence in finite-time whose adaptation gain, in contrast with the existing one, does not converge to zero. Simulations that illustrate our results are also presented.
A sensorless algorithm is developed on the basis of a rotor flux observer in the stationary frame. In particular, it involves a parameter adaptive algorithm for an initial rotor flux like the observer, which was recently proposed by Bobtsov et al. (Automatica, vol. 61, Nov. 2015). In the proposed method, the flux observer is linked to the parameter estimator via a compensating term that results from parameter error. This method has a robust property against dc bias errors, i.e., it cures the inherent weakness of the pure integrator (flux observer) to dc offsets that frequently occur in current measurements and voltage estimates. The robust performance is demonstrated through simulations and experimental results.
We present a new approach for rejection of a sinusoidal disturbance of unknown frequency, bias, amplitude, and phase for a linear unstable plant with a delay in the control channel. To solve the problem, we combine the well-known predictor feedback approach with the adaptive scheme that identifies the frequency of the disturbance. Compared to the existing results, the dynamic order of our adaptive scheme is low (equal to three) and the approach applies to plants that are unstable and have an arbitrary relative degree. The results are illustrated by the numerical example.