In this paper, we propose a control design method for single-link flexible manipulators. The proposed technique is based on the integral resonant control (IRC) scheme. The controller consists of two nested feedback loops. The inner loop controls the joint angle and makes the system robust to joint friction. The outer loop, which is based on the IRC technique, damps the vibration and makes the system robust to the unmodeled dynamics (spill-over) and resonance frequency variations due to changes in the payload. The objectives of this work are: 1) to demonstrate the advantages of IRC, which is a high-performance controller design methodology for flexible structures with collocated actuator-sensor pairs and 2) to illustrate its capability of achieving precise end-point (tip) positioning with effective vibration suppression when applied to a typical flexible manipulator. The theoretical formulation of the proposed control scheme, a detailed stability analysis and experimental results obtained on a flexible manipulator are presented.

This paper presents the detailed design, analysis, and application of the controller for a shunt active power filter based on a pulsewidth modulation dc-to-ac voltage source converter. The controller is mainly tailored to compensate harmonic currents of nonlinear loads connected to the mains. However, it can also achieve reactive-power compensation and mains-current balancing when required. The controller has a two-layer structure. The outer layer generates the current references for the inner layer. The former uses a plug-in discrete-time repetitive algorithm for current-harmonic compensation, a proportional-integral algorithm to maintain the dc-capacitor voltage in spite of unmodeled losses and a reactive-power-reference generator. The inner layer uses state-feedback with integral action for current control. The repetitive controller is justified to improve the tracking of the periodic current references required by the active filters. The stability of the resulting closed-loop system is studied and some indication of the system robustness is given. The proposed controller has been tested in a prototype with balanced and unbalanced nonlinear loads. A discrete-time model of the filter has been used from the beginning. The microcomputer delay when calculating the controller output and the delay due to the anti-aliasing filters have been included in the inner system state-variable model