Robust control design on Scilab: Hybrid remotely operated vehicle

Abstract

The goal of this work is to demonstrate the advanced features of Scilab/Xcos software in the process of designing robust controllers and to validate them numerically by a multivariable system with high coupled dynamics; it is a hybrid remotely operated vehicle (HROV). The methodology presents five steps which are the nonlinear model representation, the model linearization, the building of augmented plants, the synthesis of robust controllers through the optimization algorithms, and the plotting of the results. Linear models can be obtained straightforwardly at many operation points. The robust controllers are synthesized using the mixed sensitivity approach and tested in Scilab script and in the Xcos block diagram interface. The surge and yaw rate vehicle dynamics are controlled by a centralized controller; obtaining a good settling time, of less than 14 s, without overshooting and mitigating the coupling dynamics. Moreover, the good shaping in sensitivity and complementary sensitivity indicates robustness in performance and stability for the HROV. The proposed methodology includes time and frequency domain analysis, among other advanced features, to overcome the need for commercial software.