Rapid Control Prototyping Simulation of Particle Swarm Optimization-Tuned Backstepping Tracking Control for A Rotary...

Area of Science:

• Control Systems Engineering
• Computational Intelligence
• Robotics

Background:

• Nonlinear control systems require precise parameter tuning for optimal performance.
• Particle Swarm Optimization (PSO) is a metaheuristic optimization algorithm with potential for controller gain tuning.
• Evaluating controller performance under realistic constraints necessitates robust simulation frameworks.

Purpose of the Study:

• To establish a reproducible fixed-step simulation framework for evaluating PSO-based gain tuning in nonlinear control.
• To compare the performance of a manually tuned controller against a PSO-optimized controller for a Furuta pendulum model.
• To assess the impact of PSO optimization on trajectory tracking and disturbance rejection capabilities.

Main Methods:

• Formulation of a Furuta-type pendulum model within a 2 ms fixed-step simulation environment.
• Integration of a backstepping controller and implementation of a four-stage PSO tuning process.
• Characterization of non-ideal constraints, definition of a multi-objective PSO search space, and automated offline tuning.
• Evaluation using trajectory-tracking (step, sinusoidal) and disturbance-rejection scenarios with statistical analysis over 10 trials.

Main Results:

• PSO-optimized controller reduced step-tracking Root Mean Square Error (RMSE) from 0.065 to 0.050 rad.
• Peak pendulum excursions were attenuated by 33.1% with PSO optimization.
• Root Mean Square (RMS) control effort decreased by 22.1%, indicating enhanced energy efficiency.

Conclusions:

• The developed simulation framework provides a reliable method for evaluating nonlinear control strategies.
• PSO-based gain tuning significantly enhances controller performance in terms of accuracy and efficiency.
• This approach facilitates the validation of control strategies prior to physical hardware implementation.