LEANDRO, G. V.; http://lattes.cnpq.br/0365437321038162; LEANDRO, Gideon Villar.
Résumé:
This work aims to study via digital simulation the performance of an additional adaptive stabilizer signal (self-adjusting type) for an infinite machine-bar system. This compensator aims to introduce additional damping into the system when disturbances occur. The compensator is designed as a self-adjusting regulator that utilizes a recursive Least Square estimator and controller adaptation using pole allocation. The basic Recursive Least Squares algorithm is modified with the introduction of robustness techniques and techniques that increase the convergence rate of parameters such as; filtering of the signals used by the estimator, periodic reset of the covariance matrix P (t) with the "weighted" main diagonal elements, normalization of the regression vector. Adaptive controller performance comparisons are made with fixed parameter controllers designed by the optimal control and classical control techniques.