Design of Control Systems for Biped-Wheeled Exoskeletons

This thesis describes a control logic for a new exoskeleton model conceived by my supervisor Muscolo Giovanni Gerardo and on which a group of students, to which I belong, is studying. The idea was born from the need to help people with disabilities, such as ALS patients. ALS is a neurodegenerative disease that leads to a degeneration of motor neurons and causes total paralysis over time. At the moment there is no cure and the purpose of the exoskeleton is to give the patient the opportunity to move and to delay complete paralysis of the lower limbs. The proposed exoskeleton is a wheeled-robot and is designed to load the patient's weight and provide the necessary comforts. Just to carry out these studies it was necessary to know the general needs of an ALS patient and we as team attended a meeting like the doctors of the neuroscience department "Rita Levi Montalcini" (University of Turin) organized by our supervisor. From this meeting it has prevailed that the primary need is to walk, therefore the purpose of this thesis is precisely to simulate a human walk. The first purpose of this thesis was to create and simulate a control logic in a small-scale prototype and then perform the simulation in Arduino in order to control the servomotors, necessary for the movement of the robot joints. Unfortunately, because of COVID-19 it was not possible to carry out this thesis, therefore it was then decided to focus attention on a larger prototype and on tests in the Simulink environment, analysing the human-machine interaction. In the first chapter of this thesis, an introduction on the exoskeleton is presented, paying particular attention to the state of the art and to the types of control most implemented in this area. What will be referred to is the "trajectory-based" type, on the basis of which the angular position of the motors is controlled. A description of the human anatomy, in particular of the joints of the lower body, is presented in chapter 2. The cycle of a human walk is described and stiffness parameters of human joint are indicated because they will be useful for testing the human-machine interaction. In chapter 3 there are the presentations of the reduced-scale model and of the full-size model implemented in Simscape, which allows to create models of physical systems within the Simulink environment. Chapter 4 is dedicated to control architecture. The type of control implemented and the methods of implementation are described. The main requirement in exoskeleton applications is to obtain a tracking error for each joint close to 0. The simulation results are shown in chapter 5. Two different tasks are considered for the full-size exoskeleton: in the first the human on board has no stiffness in lower limb joints (ankle, knee, hip), in the second each human joint presents a parameter of stiffness taken from literature. A comparison with a different control logic implemented in past studies will be done. It is highlighted how the stiffness of human lower limb joints influences the various motors in terms of torque and power. In chapter 6 there are the conclusions and possible future works to arrive at the complete realization of the real prototype.