The Permanent Magnet Synchronous Motors (PMSM) have wide application in the robotics field due to its efficiency and reliability. As a servo system, it demands high precision in different control applications. Torque ripple is a critical issue resulting in mechanical vibrations and shortening the life of PMSMs, especially at low speeds. Because the magnitude of speed harmonics is proportional to the magnitude of the torque harmonics of the same order, methods to reduce speed harmonics can be utilized for torque ripple minimization. This thesis work proposes three methods for torque ripple reduction. One method is based on harmonic speed control (HSC) and harmonic current control (HCC). Another method uses the fuzzy to adjust PI parameters based on HSC-HCC. The third method utilizes torque ripple estimation (TRE) and HCC. In the proposed methods, torque ripples are estimated using a torque ripple model (TRM). At low speeds, speed harmonics and current harmonics are obtained based on an adaptive linear neural-based filter. The errors between the optimal harmonic current reference from HSC or TRE and the harmonic current from extraction are used to generate harmonic voltage in HCC. This harmonic voltage is fed back to compensate and reduce torque ripple. Furthermore, a feedforward compensation method is proposed to minimize torque ripple across a range of speeds based on the feedback compensation results. Finally, simulations and experiments are carried out to demonstrate the validity and performance of the proposed torque ripple reduction methods.

Dans cette thèse, nous proposons des solutions pour la conception de systèmes de contrôle actif de vibration robustes (AVC). Le manuscrit de thèse comporte deux grandes parties.Dans la première, les problèmes d'incertitude paramétrique dans les systèmes de contrôle actif de vibration sont étudiés. En plus des incertitudes sur la fréquence des perturbations, nous avons trouvé que la présence de zéros complexes peu amortis soulevait des problèmes de conception difficiles, même pour des systèmes et des modèles parfaitement connus. Dans ce contexte, nous avons proposé des solutions pour le problème linéaire. Une procédure améliorée d'identification en boucle fermée a été développée pour réduire les incertitudes dans l'identification de ces zéros. Pour traiter les incertitudes sur la perturbation, l'adaptation de la fréquence est de toute façon incontournable.La seconde partie est consacrée au développement et/ou à l'amélioration de deux algorithmes, désormais classiques, de compensation par feedback adaptatif direct, fondés sur la paramétrisation de Youla-Kučera. Le premier résulte de l'amélioration d'un précédent travail (Landau et al., 2005) ; les contributions concernent la synthèse du contrôleur central robuste et l'utilisation optionnelle de la surparamétrisation du filtre Q-FIR (réponse à temps fini) avec pour effet de minimiser l'effet « waterbed » sur la fonction de sensibilité de sortie. Le second algorithme présente une structure hybride directe/indirecte qui utilise un filtre Q-IIR (à temps de réponse infini). Les améliorations sont dues principalement au dénominateur du filtre, obtenu à partir d'une estimation de la perturbation. Cette solution permet également de simplifier la conception du contrôleur central.Les algorithmes ont été testés, comparés et validés sur un procédé réel du laboratoire Gipsa-lab, dans le cadre d'un benchmark international. / In this thesis, solutions for the design of robust Active Vibration Control (AVC) systems are presented. The thesis report is composed of two main parts.In the first part of the thesis uncertainties issues in Active Vibration Control systems are examined. In addition of the uncertainties on the frequency of the disturbances it has been found that the presence of low damped complex zeros raise difficult design problems even if plant and models are perfectly known. Solutions for the linear control in this context have been proposed. In order to reduce the uncertainties in the identification of low complex zeros and improved closed loop identification procedure has been developed. To handle the uncertainties on the disturbance frequency adaptation has any way to be used.The second part is concerned with the further development and/or the improvement of the now classical direct adaptive feedback compensation algorithms using Youla Kucera controller parametrization. Two new solutions have been proposed in this context. The first one results from the improvement of a previous work (Landau et al., 2005). The contributions are a new robust central controller design to the optional use of over parameterization of the Q-FIR filter which aims to ensure a small waterbed effect for the output sensitivity function and therefore reducing the unwanted amplification. The second algorithm presents a mixed direct/indirect structure which uses a Q-IIR filter. The improvements are mainly the effect of the Q filter denominator, which is obtained from a disturbance identification. This solution in addition drastically simplifies the design of the central controller.The algorithms have been tested, compared and validated on an international benchmark setup available at the Control System Department of GIPSA-Lab, Grenoble, France.