PARAMETRIC MULTIBODY MODELING OF ANTHROPOMORPHIC ROBOT TO PREDICT JOINT COMPLIANCE INFLUENCE ON END EFFECTOR POSITIONING

摘要

Nowadays, in the field of robotic, one of the most important objectives is to reduce robot error positioning and improve its dynamic behaviour. One of the main source of error in end effector positioning is due to the joint compliance: robot joint components under operating conditions can be deformed as a function of their stiffness/damping properties. Generally, for industrial robots, harmonic drive gearings are used, their principal characteristics are high transmission ratio and law weight, on the other hand, to realize high transmission ratio, harmonic drive gearings work on inner gear elastic deformation, conferring to the robot joints an excessive compliance that, in some robot applications, cannot be neglected. In this research activity multibody modelling and simulation approach has been used to analyse joint compliance influence on robot position accuracy. The principal aim of this work was the formulation of a modelling procedure that starting from classical robots modelling approach (i.e. Denavit Hartenberg) defines an universal database and a parametric modelling procedure that allows the designer to use any multibody commercial codes to analyse anthropomorphic robots considering or not the compliance effect. All the procedure was developed and managed into a numerical code environment (Matlab/Simulink). An example of commercial anthropomorphic robot was considered by assuming its principal kinematic and dynamic characteristics. Parametric models of the robot have been developed in two different multibody modelling environments (Simmechanics, Adams/View). Moreover the models structure has been built in order to control the robot movements both in motion (open loop) or in force (closed loop). In this case they are interfaced with Simulink code in a so called co-simulation approach that allows to developed a generic control system and test it by using one or more models, less or more refined.