Experimental determination of mathematical model of prototype helicopter system and controller design

Mathematical models of physical systems should be known to analyze and design of such systems. In many studies analysis and design are realized according to mathematical model, and then these designs are used in real time applications. However mathematical model of the real time running systems cannot be obtained exactly every time. For this reasons experimental determination of the real time systems' mathematical model are important. In this thesis, mathematical model of vertical and horizontal movement of a prototype helicopter, which is called Twin Rotor MIMO -Multi Input Multi Output- System (TRMS), are obtained via MATLAB Model Identification Toolbox experimentally. Subsequent sections of the thesis include three controller design methods for TRMS using experimentally obtained mathematical model. i. Fractional order controller design method based on Bode's ideal control loop is proposed using the mathematical models of vertical and horizontal movements of TRMS. It is observed that the controller obtained by this method follows the reference signal both in real time application and simulation. Thus, consistency of the mathematical models is shown. ii. In order to design adaptive controller for the systems neural network (NN) structure is used. A MATLAB code is written to generate a trainning data for this study and the NN is trained by using these data. It is observed that the proposed NN has satisfactory response. iii. A fractional order controller is designed by using parameter vector optimization method for vertical and horizontal movements of the TRMS referring to Bode's ideal control loop. During the optimization the coefficient of fractional order controller ? ?, , , , d i p k k k and Bode's ideal control loop coefficients ? ?,c , a totally seven parameters, are optimized. In this thesis, a physical interpretation of mathematical expressions of a fractional order system is proposed to contribute the understanding of fractional order controllers and systems proposed for TRMS.