Robust-adaptive control and observer design for mechanical systems with friction

• Autores: Claudiu Iurian
• Directores de la Tesis: José Julián Rodellar Benedé (dir. tes.) , Fayçal Ikhouane (dir. tes.)
• Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2008
• Idioma: español
• Tribunal Calificador de la Tesis: Enric Fossas Colet (presid.) , Carles Batlle i Arnau (secret.) , Ningsu Luo (voc.) , Carlos Canudas de Wit (voc.) , Josep Maria Guerrero Zapata (voc.)
• Texto completo no disponible (Saber más ...)
• Resumen

  • This thesis presents work in the area of friction compensation and velocity observer design for mechanical systems with friction, After bringing in a comprehensive background discussion on friction modelling for control aims, the work develops a robust friction-compensation algorithm of high-precision position-tracking mechanical systems, and formulates two asymptotic velocity-observation algorithms for mechanical drives with friction without the need of integrating velocity-sensing devices.

    First, the thesis focuses on friction modeling for control purposes. Friction science is seen as fundamentally experimental, in which relevant predictions of basic theoretical ideas are tested thoroughly against observational data. Friction models have been introduced in the literature to reproduce static and dynamic friction phenomena that have been experimentally observed. While somewhat of a makeshift nature, these models can largely be divided into those which explain friction and those which try to reproduce in an accurate manner its dynamical properties, and almost all of them are still used nowadays. In mechanical positioning systems, where high precision is required, one of the most important uncertainties is represented by the friction phenomena.

    Second, the dissertation develops a robust control law for friction compensation in position regulation and tracking of mechanical systems. In the area concerned with control of mechanisms, friction has been recognized as the source of tracking errors, limit cycles, and stick-slip motions, among other phenomena, degrading the performance of the controlled mechanism. We present a novel friction-compensation design for control of mechanisms that is robust against the friction model and adaptive with respect to parameter uncertainties in the mechanical plant. The performance and practical robustness of the obtained control law is evaluated numerically by employing the LuGre model as the ""true'' friction in the system. Very acc