Exploiting the natural dynamics of parallel robots for energy-efficient pick-and-place tasks

During the execution of pick-and-place operations, the links of industrial robots are continuously accelerating and braking. Consequently, their kinetic energy constantly oscillates and a huge potential for energy savings appears. The energy that is introduced to accelerate the links can be recuperated during the braking phases and used for acceleration instead of dissipating it as heat. One possibility for taking advantage of this is to transform the mechanism into a vibratory system by introducing elastic elements, such that the robot naturally oscillates between the required positions at the desired frequency.Parallel robots are excellent candidates for energy-efficient pick-and-place operations. A typical example is the Delta robot, which can achieve very high speeds in packaging processes. In this type of robots, the motors remain stationary, allowing the use of lighter links. This results in higher speeds of operation, reduced energy consumption and increased precision. However, the presence of passive joints complicates the kinematic and dynamic analysis of such structures.In this thesis, the feasibility of increasing the energy efficiency of parallel robots performing pick-and-place tasks by introducing linear springs and exploiting the natural dynamics of the ensuing system is investigated. Two methods are presented to find the required parameters of the springs, such that the free vibration of the modified robot matches a nominal pick-and-place task. Moreover, as the task requirements may vary continuously, a method is proposed to minimize the energy consumption including these disturbances. These methods are implemented in the simulation of a Delta robot and a five-bar linkage.To estimate the energy consumption, a set of tasks is defined within the workspace of the robots and typical controllers are simulated. It was found that it is possible to obtain a vibratory system, the free vibration of which matches the nominal task requirements by including linear springs with constant stiffness, despite the nonlinearities of the parallel robots. Moreover, it was found that a significant reduction in the energy consumption is possible without affecting the versatility of the robot considerably. Throughout the work, the results are compared with those of implementing an alternative energy-saving strategy in the robot without springs