Ultra-Local Model-Based Intelligent Robust Control of PKMs: Theory and Simulations

• Published in: IFAC-PapersOnLine Vol. 55; no. 12; pp. 634 - 640
• Main Authors: Hassan, G., Chemori, A., Rafei, M. El, Gouttefarde, M., Francis, C., Hervé, P.E., Sallé, D.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 2022
• Subjects: Intelligent Robust Control; Model-Free Control; numerical simulations; Parallel Kinematic Manipulators; Pick; RISE feedback control; Throw; numerical simulations; Pick; Parallel Kinematic Manipulators; Throw; Model-Free Control; RISE feedback control; Intelligent Robust Control
• ISSN: 2405-8963; 2405-8963
• DOI: 10.1016/j.ifacol.2022.07.383

In this paper, we propose a novel Intelligent Robust Control (IRC) suitable for controlling highly nonlinear Multiple-Input-Multiple-Output (MIMO) systems. The proposed IRC scheme takes advantage of the Robust Integral of the Sign of the Error (RISE) control law and Model-Free Control (MFC). The MFC scheme is mainly composed of: (i) a nonlinear function estimated from an ultra-local model representing the input-output behavior of the system, (ii) the νth derivative of the reference trajectory as a feedforward term, and (iii) a feedback control term. MFC is characterized by its simple concept and its ability to compensate for the modeled and unmodeled system dynamics through its nonlinear compensation term. The proposed IRC approach consists of redesigning the feedback term of MFC scheme based on RISE feedback law to further improve its robustness against external disturbances and to guarantee a semi-global asymptotic tracking despite the presence of disturbances and uncertainties. Numerical simulations under different operating conditions have been conducted on T3KR parallel manipulator, in a pick-and-throw task, to validate the relevance of the proposed IRC strategy. The comparison with a model-based feedforward RISE and a feedforward super-twisting sliding mode control, by exploiting different performance indices, confirms the superiority of the proposed IRC approach.