A new hybrid force/position control approach for time-varying constrained reconfigurable manipulators

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Model-based control is insufficient for time-varying constrained reconfigurable manipulators due to inevitable system uncertainties.
• Existing methods struggle to adapt to dynamic changes and unknown system behaviors.

Purpose of the Study:

• To propose a novel hybrid force/position control strategy for time-varying constrained reconfigurable manipulators.
• To address uncertainties in manipulator dynamics and improve control performance.

Main Methods:

• Developed a reduced-order dynamic model of the manipulator system.
• Integrated a model-dependent control scheme with a radial basis function neural network (RBFNN) for model-free estimation of unknown dynamics.
• Incorporated an adaptive compensator to mitigate friction effects and RBFNN reconstruction errors.
• Utilized Lyapunov theorem and Barbalat's lemma for stability analysis.

Main Results:

• The hybrid control scheme effectively estimates unknown system dynamics using RBFNN.
• The adaptive compensator successfully overcomes friction and neural network errors.
• Guaranteed bounded tracking errors for joints and force, with asymptotic convergence of joint errors.
• Simulations demonstrated the superiority and applicability of the proposed method on a 2-DOF manipulator.

Conclusions:

• The proposed hybrid force/position control approach offers robust and accurate control for time-varying constrained reconfigurable manipulators.
• The integration of model-based control, RBFNN, and adaptive compensation provides a powerful solution for complex robotic systems.
• The method ensures stability and achieves desired performance levels, outperforming traditional approaches.