Modelling, simulation and motion control of climbing parallel robots

This dissertation has been dedicated to the study, modelling, simulation and motion control of 6 DOF parallel robots used as climbing robots. In order to fulfil the idea of using the Stewart platform as a mobile robot, computational tools have been developed based on the multibody dynamics. These tools conform the basic frame for the kinematics and dynamics analysis of this new robotics application. And also, they have allowed research and development of algorithms for the parallel robot's kinematics control, to enable it to climb, based on sensorial information, through long cylindrical structures describing unknown spatial trajectories. The complex problems of the multiple solutions of the forward kinematics have been solved, using numeric methods, to achieve a unique solution that fulfils the conditions for the path planning and motion control in real time. The dynamic modelling also includes the pneumatic power actuators. This tool has been essential for the simulation, research and adjustment of the control systems of the pneumatic actuators of the robot. As a practical case of simulation, a conventional PID control law has been implemented and the robot's tracking has been evaluated for a trajectory given by the algorithm. The results of this thesis have contributed to the development of a climbing parallel robot prototype and its application for the maintenance of palms trees (pruning, fumigation, germination). The contributions have been reflected in several phases in the development of the prototype. In the design phase in particular, the contributions of the kinematics and dynamics analysis have produced excellent results. At the same time, the computational modelling of the control system has allowed the verification of the dynamics and theoretical control hypotheses with the experimental results. This has led to the deduction and to the interpretation of the tendencies of the robot control system in relation to the regulators tuning of the six axes of the control card. Finally, the application of the results of the kinematics and dynamics modelling has been fundamental for the motion control of the parallel robot in real time. These results are closely related to the unique solution of the forward kinematics problem.