Tracking control of industrial manipulator based on adaptive RBF neural network with local model approximation

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Industrial robots require precise control for complex tasks.
• Existing control methods often depend on accurate mathematical models, which are difficult to obtain for real-world systems.
• System nonlinearities and uncertainties pose significant challenges to achieving high-precision trajectory tracking.

Purpose of the Study:

• To develop a Lyapunov-stability-guaranteed local model adaptive control algorithm for industrial robots.
• To enable real-time learning and compensation of system nonlinearities and uncertainties without an exact plant model.
• To achieve high-precision trajectory tracking control for robot manipulators.

Main Methods:

• Proposed a local model adaptive Radial Basis Function (RBF) neural network control algorithm.
• Implemented an adaptive control law for online adjustment of neural network parameters.
• Utilized Particle Swarm Optimization (PSO) to optimize RBF basis width parameters.
• Validated the algorithm using the ABB IRB1600 industrial robot in MATLAB Simscape and ADAMS co-simulation.

Main Results:

• Demonstrated effective real-time trajectory tracking control.
• Significantly reduced tracking errors compared to traditional methods.
• Showcased enhanced robustness and adaptability of the control system.
• Maintained guaranteed stability within the Lyapunov framework.

Conclusions:

• The proposed adaptive RBF neural network control algorithm is effective for industrial robot trajectory tracking.
• The algorithm successfully compensates for system uncertainties and nonlinearities in real-time.
• The method offers a robust, stable, and adaptable solution for high-precision robot control.