Robust Design for Dynamic System Under Model Uncertainty

Abstract

In real-world applications, a nominal model is often used to approximate the design of an industrial system. This approximation could make the traditional design method less effective due to the existence of model uncertainty. In this paper, a novel stability-based approach is proposed to design the system ensuring robust stability under model uncertainty. First, the design variables and their variation bounds are configured to make the system stable. Then, a robust design is developed to incorporate system eigenvalues that are less sensitive to model uncertainty. Finally, the tolerance of the design space will be maximized under given performance constraints. A simulation example is conducted to demonstrate the effectiveness of the proposed robust design method.