Search

To address the escalating demand and supply constraints for semiconductor devices, manufacturing techniques such as thermal atomic layer etching (ALE) have been utilized, which require robust temperature control systems. Radiative heating with lamps is a promising method for achieving fast and precise temperature control. However, there is limited research in analyzing the dynamic control of radiative lamp heating systems. In this study, a transient, two-dimensional (2-D) radiative heating model in a cross-flow thermal ALE reactor is constructed through Ansys Fluent. A sparse identification (SINDy) modeling approach is applied to build a reduced-order dynamic model for the prediction of system states in the control system. A model predictive controller (MPC) is then developed to bring the wafer surface temperature to the target temperature while preserving the uniformity of the temperature on the wafer surface. Control is implemented by adjusting the powers of three groups of heating lamps independently. Notably, a novel feedback-based time-varying steady-state penalty approach is applied with the MPC in this study, which enables the system to reach the target temperature range within 10 s while maintaining temperature uniformity for 1000 s. • Computational fluid dynamics modeling of wafer temperature profile. • Sparse model identification using spatio-temporal temperature data. • Model predictive control using nonlinear sparse model. • Implementation of MPC on CFD model of the process. [ABSTRACT FROM AUTHOR]

This paper proposes a tube model predictive control method considering uncertainties of renewable energy generation in the low-voltage distribution network. For relatively low dependence on communication networks, a decentralized voltage control framework is urgently needed to be developed for voltage regulation in the low-voltage distribution network. Opposed to the traditional voltage control techniques, a novel decentralized tube model predictive voltage control (TMPC) strategy is proposed. First, this strategy entails the formulation of a discrete model encompassing distributed photovoltaic units and energy storage systems intertwined with the distribution grid model based on voltage sensitivity. Then, considering the worst case of voltage deviations caused by uncertain and stochastic power fluctuation, we specifically discuss a robust positively invariant to determine the control section of controllers. Further, the strategy employs tube model predictive controllers encompassing nominal and auxiliary controllers in distribution areas. The nominal controller generates a central path for different voltage operating points of each distribution area, and the auxiliary controller makes all trajectories lie in a bounded neighborhood called a tube. The effectiveness of the proposed control strategy is investigated on a test system containing renewable energy resources. • A novel decentralized control structure is proposed to serve voltage control for distribution networks, which can cooperate for accurate voltage control with random and fluctuant renewable energy. • In the nominal controller, operation points of DGs are generated considering the generation and load scheduling. In the auxiliary controller, detailed control objectives and constraints are proposed for suppressing disturbances of DGs. • To balance the conservation and performance, a worst-case-based robust positively invariant is developed to quantify the range of tubes to cope with the randomness. [ABSTRACT FROM AUTHOR]

Distributed energy resources (DER) are integrated into a microgrid through dc-dc power electronic converters. The bidirectional dc-dc converter regulates charging and discharging operations of ESS. Model predictive control (MPC), is a high-performance control technique for these converters, but it is limited in robustness to parameter mismatch, model uncertainties and sensor measurement noise. Therefore, in this paper, an improved hybrid cascade-parallel extended state observer (CP-ESO) is proposed, for model-free predictive control, to guarantee both robustness to parameter/model uncertainties and measurement noise suppression. A novel structurally-adaptive ESO scheme is also proposed to improve the disturbance rejection of CP-ESO during transient response. These results are supported by analysis and design guidelines for the selection of optimal sub-frequencies. Experimental results, for a bidirectional dc-dc boost converter, validate the effectiveness of the proposed methods against model uncertainties, external disturbances from variable input voltage and load, as well as measurement noise. • Analysis and design guidelines for multi-frequency extended state observers (ESO). • Improved cascade-parallel extended state observer with better noise suppression. • Novel structurally-adaptive ESO that gives better transient disturbance rejection. • Experimental validation with a bidirectional dc-dc boost power converter. [ABSTRACT FROM AUTHOR]

This paper is about robust Model Predictive Control (MPC) for linear systems with additive and multiplicative uncertainty. A novel class of configuration-constrained polytopic robust forward invariant tubes is introduced, which admit a joint parameterization of their facets and vertices. They are the foundation for the development of novel Configuration-Constrained Tube MPC (CCTMPC) controllers that freely optimize the shape of their polytopic tube, subject to conic vertex configuration constraints, as well as associated vertex control laws by solving convex optimization problems online. It is shown that CCTMPC is—under appropriate assumptions—systematically less conservative than Rigid- and Homothetic- Tube MPC. Additionally, it is proven that there exist control systems for which CCTMPC is less conservative than Elastic Tube MPC, Disturbance Affine Feedback MPC, and Fully Parameterized Tube MPC. [ABSTRACT FROM AUTHOR]

In this paper, the application of a robust nonlinear model predictive control (NMPC) framework to mammalian cell cultures is proposed, dealing with possible large kinetic parameter uncertainties. Industrial constraints formulated in view of good manufacturing practice and quality-by-design approach are also considered, namely the assurance that all state trajectories are contained within a corridor defined by lower and upper safety bounds. The latter are assimilated to the well-known tube-based paradigm which is used to formulate the corresponding robust NMPC problem. Both classical and tube-based NMPC performances are assessed in numerical simulations where specific key-species are regulated while dealing with an uncertain plant model. The capability of the tube-based method to reduce the impact of the parameter variations on the state trajectories and the violation of the constraints is highlighted, suggesting the transfer of the method on a real pharmaceutical process. • NMPC design for a 25-state data-driven mammalian cell culture model. • Qualitative model sensitivity analysis using confidence corridors. • Robust tube-based NMPC framework application to mammalian cell culture model. • Practical guidelines for experimental validation in industrial environment. [ABSTRACT FROM AUTHOR]

We propose a robust model predictive control (MPC) method for discrete-time linear systems with polytopic model uncertainty and additive disturbances. Optimizing over linear time-varying (LTV) state feedback controllers has been successfully used for robust MPC when only additive disturbances are present. However, it is challenging to design LTV state feedback controllers in the face of model uncertainty whose effects are difficult to bound. To address this issue, we propose a novel approach to over-approximate the effects of both model uncertainty and additive disturbances by a filtered additive disturbance signal. Using the System Level Synthesis framework, we jointly search for robust LTV state feedback controllers and the bounds on the effects of uncertainty online, which allows us to reduce the conservatism and minimize an upper bound on the worst-case cost in robust MPC. We provide a comprehensive numerical comparison of our method and representative robust MPC methods from the literature. Numerical examples demonstrate that our proposed method can significantly reduce the conservatism over a wide range of uncertainty parameters with comparable computational effort as the baseline methods. [ABSTRACT FROM AUTHOR]

We present a robust nonlinear model predictive control (NMPC) algorithm for dynamic systems with mixed degrees of freedom, i.e., continuous and discrete manipulated variables. The discrete variables are often excluded from the NMPC layer and determined at a supervisory layer, which typically leads to suboptimal solutions. In contrast, the proposed controller optimizes both classes of degrees of freedom within the NMPC layer, thereby enhancing the closed-loop performance. Our approach relies on a computationally efficient relaxation and integrality restoration strategy. Initially, a nominal controller computes relaxed central reference trajectories for the states and the input variables. These trajectories are then tracked by an ancillary NMPC controller before the integrality of the discrete variables is restored. The role of the ancillary controller is to ensure robustness against the integer restoration error as well as other disturbances based on a nonlinear Tube MPC approach. Furthermore, we provide sufficient conditions to establish recursive feasibility and guarantee robust closed-loop stability. Finally, the effectiveness of the approach is demonstrated through two nonlinear simulation examples, affirming its viability and robustness. [ABSTRACT FROM AUTHOR]

This work proposes an approach to replace the use of a robust control invariant set by a pair of simpler sets that provide an inner and an outer approximation of the former. In the proposed approach, the outer set plays the role of the target region and the inner set is such that the trajectories that start inside it can be kept inside the outer set and be driven back to the inner set within a finite-time horizon. We show that the existence of these two sets implies the existence of a robust control invariant set between both regions. We also provide results that allow finding an inner set from a given target outer set and we show a way of using both sets in Model Predictive Control schemes such that the target region is never abandoned in spite of the fact that nor that region neither the inner set are invariant. We also illustrate the ideas with an example in which the inner and outer sets are very simple notwithstanding that any robust invariant set is not convex. [ABSTRACT FROM AUTHOR]