| description abstract | Many mechanical systems often show nonlinear behavior related to particular operating conditions or to the nonlinear characteristic of the elements (springs, dampers, etc.) making up the system. In these cases, common engineering practice is to linearize the equation of motion around a particular operating point and to design a linear controller. Although this approach is simple, its main disadvantage is that stability properties and validity of the controller are only local. For these reasons, over the last decades, nonlinear control techniques have been investigated more and more in order to improve control performance. In particular, in this paper, sliding mode control (SMC) technique, which is based on the model of the system (model-based), is considered because of its easy implementation, especially on simple mechanical systems, and the considerable robustness of the controller even under significant model uncertainties. This technique is analyzed numerically with respect to the pendulum system to better understand the influence of the control action on the system dynamics. A nonlinear sliding surface is also considered, recalling the terminal sliding mode (TSM) control already analyzed in the scientific literature. This sliding surface is characterized for the numerical system, and then it is applied experimentally in order to control a highly nonlinear system, consisting of a three-link flexible manipulator. For this system, a nonlinear modal model is developed, and a nonlinear observer is designed. Finally, results of experimental tests on the manipulator are reported, in order to compare the performances of the linear embedded control and the sliding mode controllers with the linear and nonlinear sliding surface. | |
