Fuzzy-Immune-Regulated Adaptive Degree-of-Stability LQR for a Self-Balancing Robotic Mechanism: Design and HIL Realization

This letter formulates a fuzzy-immune adaptive system for the online adjustment of the Degree-of-Stability (DoS) of Linear-Quadratic-Regulator (LQR) procedure to strengthen the disturbance attenuation capacity of a self-balancing mechatronic system. The fuzzy-immune adaptive system uses pre-configured control input-based rules to alter the DoS parameter of LQR for dynamically relocating the closed-loop system's eigenvalues in the complex plane's left half. The corresponding changes in the eigenvalues are conveyed to the Riccati equation, which eventually yields the self-adjusting LQR gains. This arrangement allows for the flexible manipulation of the applied control effort and the response speed as the error conditions change. The efficacies of the self-tuning LQR scheme are verified by performing custom-designed hardware-in-the-loop experiments on the Quanser rotary inverted pendulum system. As compared to the DoS-LQR, the proposed controller improves the pendulum's transient recovery time, overshoots, input demands, and offsets by 32.3%, 50.5%, 33.9%, and 33.3%, respectively, under disturbances. These experimental outcomes verify that the proposed self-tuning LQR law considerably improves the system's disturbance attenuation capability.