SDRE controller for motion design of cable-suspended robot with uncertainties and moving obstacles

2017 International Conference on Structural, Mechanical and Materials Engineering

材料科学;力学

Behboodi, Ahad^1 ; Salehi, Seyedmohammad^2

Biomechanics and Movement Sciences University of Delaware, United States^1

Computer and Information Sciences University of Delaware, United States^2

关键词: Approximation methods;    Artificial potential fields;    Cable suspended robot;    Efficient designs;    Moving obstacles;    Optimal controls;    Power series expansions;    Systematic designs;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/248/1/012031/pdf
DOI  :  10.1088/1757-899X/248/1/012031

来源: IOP

【 摘 要 】

In this paper an optimal control approach for nonlinear dynamical systems was proposed based on State Dependent Riccati Equation (SDRE) and its robustness against uncertainties is shown by simulation results. The proposed method was applied on a spatial six-cable suspended robot, which was designed to carry loads or perform different tasks in huge workspaces. Motion planning for cable-suspended robots in such a big workspace is subjected to uncertainties and obstacles. First, we emphasized the ability of SDRE to construct a systematic basis and efficient design of controller for wide variety of nonlinear dynamical systems. Then we showed how this systematic design improved the robustness of the system and facilitated the integration of motion planning techniques with the controller. In particular, obstacle avoidance technique based on artificial potential field (APF) can be easily combined with SDRE controller with efficient performance. Due to difficulties of exact solution for SDRE, an approximation method was used based on power series expansion. The efficiency and robustness of the SDRE controller was illustrated on a six-cable suspended robot with proper simulations.