Feedback Control—Based Inverse Kinematics Solvers for a Nuclear Decommissioning Robot

摘要

Abstract: The article develops two novel feedback control-based Inverse Kinematics (IK) solvers. They are evaluated for a dual-manipulator mobile robotic system with application to nuclear decommissioning. The first algorithm has similarities to other feedback control based solvers, and borrows ideas from the Cyclic Coordinate Decent and the Jacobian Transpose methods. This yields a particularly straightforward algorithm with tunable Proportional-Integral-Derivative (PID) gains to determine performance. The second approach utilises a discrete-time state space modelling framework to solve the IK problem. Although the second solver is more complex to implement, preliminary simulation results for the case study example, show that it can converge quicker, and has improved immunity to the kinematic singularities that can occur in Jacobian based methods.