Optimal Dynamic Quantizers for Feedback Control With Discrete-Level Actuators: Unified Solution and Experimental Evaluation

Abstract

This paper proposes to use optimal dynamic quantizers for feedback control in mechatronics systems when the actuator signals are constrained to discrete-valued signals. Here, the dynamic quantizer is a device that transforms the continuous-valued signals into the discrete-valued ones depending on the past signal data, as well as the current data. First, a closed form optimal quantizer is presented in a general linear fraction transformation representation setting. The optimal quantizer minimizes the deviation of the output produced by the quantized signals from the corresponding output yielded by the continuous-valued signals before quantization. Then, its experimental evaluation is performed by using a crane positioning system with a discrete-valued input to demonstrate the effectiveness of the proposed quantizers.