SLIDING MODE CONTROL (SMC) BASED ON GENETIC ALGORTIHM OPTIMIZATION APPLIED TO MANIPULATOR ROBOT

• Main Author: RIYAD FIRDAUS (NIM 23206025), AHMAD
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/7096
• Institution: Institut Teknologi Bandung

The dynamical model of manipulator robot is represented by equations systems which are nonlinear and strongly coupled. Furthermore, the inertial parameters of manipulator depend on the payload which is often unknown and variable. So, to avoid these problems we studied sliding mode controller which is well suited to manipulator robot application. The sliding mode controller provides an effective and robust means of controlling nonlinear plants. The performance of sliding mode controller depends on parameter selection of gain switching (k) and sliding surface constant (S). A parameter selection algorithm is proposed by genetic algorithm to select the gain switching and sliding surface constants parameter so that the controlled system can achieve a good overall performance in the sliding mode controller design. The searching of these parameter values is conducted by objective function which has been specified, that are: settling time (ts), steady state error (ess), and control input (u). The evaluation of searching result is done by evaluating the fitness function of chromosomes which are defined in this algorithm. The PML parameters optimization is done by using MATLAB 6.5 which applied to PUMA 260 2-DOF model. Simulation shows a better performance of PML if using genetic algorithm optimization. A better overall performance is presented by smaller settling time and steady state error from manipulator to achieve position reference. For example: each joints of manipulator needs 1.5s and 1.86s (if not using optimization) and, 0.97s and 0.98s (if using optimization) to achieve reference position 90o. At the realization with the real plant, there are some degradations of performance. These matters seen with settling time to achieve the reference position 46.88o needs 3.3 secons. These are because of usage MATLAB for PML implementation and unreckoned of mechanic friction. <br />