Internal model control design using Just-In-Time learning technique

Abstract

Internal model control (IMC) is a convenient and powerful controller design strategy for the open-loop stable dynamic systems. IMC design is expected to perform satisfactorily as long as the plant is operated in the vicinity of the point where the process model is obtained. However, many chemical processes exhibit a certain degree of nonlinearity. When the process dynamics of such processes is forced away from its nominal design condition due to the increasingly stringent requirements on product quality and energy utilization, as well as on safety and environmental responsibility, the performance of linear IMC controller will degrade or even become unstable. The extension of the linear IMC strategy to nonlinear systems has been a popular model-based control approach. In literature, several nonlinear IMC schemes that incorporate concepts from linear IMC have been developed recently. However, these control schemes relied on computationally demanding analytical or numerical methods and neural networks to learn the inverse process dynamics for the necessary construction of nonlinear operator inverses. To overcome aforementioned difficulties, a nonlinear IMC (NLIMC) design strategy based on partitioned model inverse is proposed for a class of nonlinear systems that can be described by Just-in-Time Learning (JITL) technique. The proposed controller consists of a conventional linear IMC controller augmented by an auxiliary loop to account for nonlinearities in the system. In addition, a memory-based IMC design strategy is proposed for nonlinear systems. Simulation results show that the proposed control strategies give better performance than their conventional counterparts.