Anti-windup design for systems with input quantization

This paper presents a full-order anti-windup compensation strategy for LTI systems with input quantization. The anti-windup compensator functions by comparing the quantized and un-quantized control signal and attempting to ensure that the system's behavior with quantization is close as possible to the ideal system behavior without quantization. The paper proposes an anti-windup compensator which can provide ultimate boundness stability conditions and also guarantee large-signal L-2 gain bounds. The problem is tackled by using the sector and boundness characteristics of the quantizer. The conditions for ultimate boundedness and large signal L-2 gain are formulated as a convex optimization problem which allows one to trade-off the size of the ball of ultimate boundedness with the L-2 performance objectives. The effectiveness of the proposed technique is shown through a simulation example.