New prospects in the adaptive control of robots under unmodeled environmental interactions

A new method for the simultaneous tackling of the effects of incomplete and approximate robot models and that of the dynamic interaction with an unmodeled environment in robot control was recently outlined. The basic idea was to gain adaptivity via tuning the parameters of the symplectic group inherent in the Hamiltonian description of classical mechanical systems. This description results in equations of uniform structure independent of the particular properties of the system to be controlled. Until now, only the effects of tuning of only special parameters were investigated by simulation for a special 3 DOF robot arm. In this paper, the behavior of a similar control with extended parameter space is considered in the case of a vertical 3 DOF SCARA robot. Simulation results are discussed for different nominal trajectories. The results reveal that this approach may open new prospects in the control of mechanical devices