Model predictive control method for a system with constraints on the polarity of control inputs and application to a pneumatic isolation table

Abstract

This paper addresses a plant that includes different time delays dependent on the polarity of control inputs. An example of such a system is a pneumatic isolation table, which consists of two valves of different lengths. For this plant, a switching control method depending on the control input polarity has been employed. Each polarity is considered as a separate single input system. Alternatively, this plant can also be considered as a two-input system by using both control inputs simultaneously. Thus, the conventional method might have restricted the control performance. Moreover, the switching control method determines an input value by considering the next step input beforehand. To improve control performance, a method comprising several steps in advance is proposed. It is a model predictive control (MPC). However, this method is difficult to apply to the switched system as it results in a problem of hybrid control. In the present paper, we consider this plant as a two-input system, and thus, improvement of control performance is expected. Moreover, the proposed view of the plant is not considered as a switched system. It enables MPC to be applied to the plant by solving a linearly constrained quadratic optimization problem. Finally, the control performance achieved by the proposed method is verified through numerical simulations.