Developement of the SmartGimbal control system of the SmartBay platform

The final work is focused on the development of the control system of the SmartGimbal holder camera, a 2 axis tracking system, also called IPS (Inertial Platform Stabilization). The study evolves is in collaboration with the Digisky firm, which offers high technology and low cost solutions in the avionic system field. Precisely the SmartGimbal is one of the components compatible with the SmartBay platform and it can holds a camera to achieve the purpose of aerial monitoring, surveillance, fire detection and medical emergencies. Proceeding according to the Model-Based approach, four main parts of the thesis can be identified. The necessity to have a model of the system around which design the control system, means that at first the mechanics and the dynamics of the gimbal system is analyzed. The gimbal dynamic model is derived by means of the Denavit-Hartenberg conventions from robotic, and the dynamic equations of the gimbal are obtained. Then they are linearized in order to design a suitable controller for an LTI system. In the second part, thanks to the well known control strategies of Loop-Shaping and LQR, is created the control system of the gimbal (which is able to rotate around the azimuth and elevation axis due to the presence of two independent DC-motors). From the simulations, taking place in the Matlab and Simulink environments, the maximum torque and angular speed values required by the gimbal system are extracted and based on these a survey concerning the most suitable DC-motor is conducted. In a third phase, the DC-motors jet available in the Digisky’s laboratory are taken as plant of a new designed PID controllers, once the mathematical model and the corresponding state-space representation of the motors are achieved. Actually the controllers are of PI kind. Finally the control algorithms are converted from Simulink’s block models into C code: the purpose is to export the generated codes on the microcontroller board, which is represented by the Arduino Mega 2560. In this manner, the two control algorithms can run on the microcontroller and the control system can be tested on the real existing prototype. Before the validation phase on the real gimbal system, as recommended in the model-based design approach, a Software-in-the-Loop and a Processor-in-the-Loop steps are completed.