Dynamic interaction and control of an indoor blimp inside the CERN FCC-hh magnetic environment

The purpose of this thesis is to study the dynamic interaction and control of an indoor airship robot, also called blimp, inside the new generation of CERN FCC-hh detector cavern. The problem falls in the field of robotics in harsh environment due to the presence of high magnetic fields and radiation. This work is part of an R&D program of the Experimental Physics department at CERN which deals with the development of new robotic technologies for automated and robotized services and interfaces for the maintenance and disposal of the detector. After studying and deriving the 6 DOF nonlinear dynamic model of an indoor blimp, the work focused on the estimation of magnetic disturbances acting on the brushless electric motors with which the blimp is equipped. Through electromagnetic simulations with FEM modeling software such as FEMM and CST it was possible to characterize the magnetic behavior of an electric motor by associating it with a magnetic dipole moment and calculating the interaction in terms of forces and torques of this within the simulated nonuniform magnetic field of the detector. Then, the control problem was analyzed using linear control techniques such as PID and LQR for the trajectory tracking control of the linearized and decoupled longitudinal and lateral dynamics simulated in MATLAB. The last part concerns the practical tests that were carried out during the internship period at CERN. The activities involved the assembly of a real COTS blimp like Blimpduino from jjrobots and its identification dynamic parameters with the creation of a CAD model on CATIA. The tests has been done inside a cleanroom of a CERN laboratory without the presence of disturbances in which the PhaseSpace motion capture system was installed. Before doing the test was necessary to setup the motion caputure system which required cameras calibration, reference frame alignment and tracking rigid body triad definition. After implementing and updating the PID controller in the Arduino board, the station keeping test was carried out with two blimp of different sizes for altitude and yaw control. The blimp is followed with the telemetry of the on board sensors and with the tracking data of the motion capture system and the results were finally compared with a MATLAB simulation in which the same scenario was reproduced and the state was generated with the numerical integration of the dynamics model.