Modelling and Control of Cable-Driven Hyper-Redundant Robots

Nowadays, inspection and maintenance of industrial sites are often carried out by specialised operators who need to enter narrow and potentially dangerous environments. In this context, cable-driven hyper-redundant robots support the operators to carry out inspection and maintenance tasks proving to be an effective solution in terms of safety. The cable transmission keeps the actuators safe from the possible extreme environmental conditions of the inspection site, and the kinematic redundancy allows the robot to move into narrow environments avoiding obstacles. This dissertation aims to study cable-driven hyper-redundant robots providing a unified modelling framework that accurately describes their dynamics and could be employed to design control algorithms. The proposed modelling technique defines a simulation environment by interconnecting all the components of a mechatronic system. This approach allows speeding up the preliminary analysis necessary to evaluate the effectiveness of different solutions highlighting possible critical aspects. Furthermore, describing each subsystem by different complexity models allows employing the optimal model for the various scenarios the designer would analyse, such as secondary dynamics studies or control algorithm design. The obtained model provides the framework to design a control algorithm able to damp the residual vibration induced by cable transmission. Since fast actuators need to be employed to increase the vibrations damping, a novel design methodology to passively compensate for gravity forces is presented to improve the actuators dynamic response. Finally, the proposed modelling framework is applied to a novel cable-driven hyper-redundant robot to prove the versatility of this method.