Prescribed Performance Control for Flexible Joint Robots With All State Quantization: Design and Experiments

Abstract

A novel low-complexity prescribed performance tracking control scheme is proposed in this study for flexible-joint robots under full-state quantization. All system states are quantized using a uniform quantizer, which introduces the discontinuities. Unlike existing methods that rely on command filters, a backstepping-like design is employed to handle these discontinuities while maintaining a simple controller. By integrating a prescribed performance function with quantized states, the proposed method ensures both transient and steady-state tracking performance while significantly reducing communication bandwidth. Furthermore, neural networks are employed to approximate unknown nonlinearities, thereby eliminating the need for global Lipschitz conditions commonly required in quantized control. Rigorous stability analysis confirms that all tracking errors remain within prescribed performance bounds. Finally, experiment results on the Quanser flexible-joint robot platform are provided to demonstrate the availability of the proposed method.