Optimal design of model predictive controller based on transient search optimization applied to robotic manipulators.

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Robotic manipulator control is challenged by nonlinearity and uncertainty, impacting motion and trajectory tracking.
• Effective controller design is crucial for achieving desired robotic system performance.

Purpose of the Study:

• To design and optimize a Linear Parameter Varying - Model Predictive Controller (LPV-MPC) for a two-link robot manipulator.
• To determine optimal controller parameters using a novel meta-heuristic algorithm, Transient Search Optimization (TSO).

Main Methods:

• Establishment of an LPV-MPC for a two-link robot manipulator.
• Optimization of LPV-MPC parameters using the Transient Search Optimization (TSO) algorithm.
• Verification through set-point and nonlinear trajectory tracking simulations.

Main Results:

• The TSO-optimized LPV-MPC outperformed Computed Torque Controller (CTC) schemes in set-point tracking.
• The proposed LPV-MPC method showed good performance in nonlinear trajectory tracking tasks.
• Robustness to structural uncertainty was confirmed through significant system parameter deviations.

Conclusions:

• The TSO-optimized LPV-MPC offers improved performance for robotic manipulator control.
• This approach enhances motion and trajectory tracking capabilities, addressing nonlinearity and uncertainty.
• The method provides a robust solution for optimizing robotic controller parameters.