Comparison between ultra-local model and auto-tuning PID controls strategies applied to the redundant manipulator robot for several inputs

This paper deals with the development of an auto-tuning PID and an ultra-local model controls strategies applied to a redundant manipulator robot with a load on its endpoint. This system is dynamically modeled using finite element and Euler-Lagrange approaches. The proposed approach of ultra-local model control consists of estimating the ultra-local model parameter via an algebraic derivation approach. In order to validate its efficiency towards the flexibility's structure, the ultra-local model control is compared to auto-tuning PID controller using several inputs which are characterized by its optimal time and its smoothness. Numerical simulations show the simplicity and the robustness of the ultra-local model control compared to the auto-tuning PID control with respect to the flexibility's structure for several motions profiles.