Shorter Communication

Plug-flow type reactor versus tank reactor

This work develops algorithms for the estimation of sparse interpretable data-driven models. Linear combination of linear and nonlinear basis functions is considered for candidate models, where the optimal basis functions are selected by using the branch and bound algorithm. Model parameters are optimally estimated by employing Bayesian inferencing to seek the maximum a posteriori estimates conditioned on the available data. This is implemented in a two-step expectation-maximization algorithm enabling simultaneous estimation of noise characteristics along with the model parameters. A priori analysis is undertaken based on the ranking of the basis functions in the library. In presence of correlated noise, special properties of the Jacobian matrix are exploited for efficient pruning. Model selection is done by using modified Akaike information criterion that rewards model fitness while penalizing model size and complexity. The algorithm, called Bayesian identification of Dynamic Sparse Algebraic Models (BIDSAM), is tested for three case studies.

This paper presents a method to estimate process dynamic gradients along the transient that combined with the idea of Modifier Adaptation (MA) improves the economic cost fuction of the examples presented. The gradient estimation method, called TMA, aims to reduce the large convergence time required for traditional MA in processes of slow dynamics. TMA is used with an economic predictive control with MA (eMPC+TMA) and was applied in two case studies: a simulation of the Williams-Otto reactor and a hybrid laboratory plant based on the Van de Vusse reactor. The results show that eMPC+TMA could reach the plant real steady-state optimum despite process-model mismatch, due to the inclusion of the effect of process dynamics in the TMA algorithm. Despite the estimation errors, the proposed methodology improved the profit of the experimental case study, with respect to the use of an eMPC with no modifiers, by about 20% for the unconstrained case, and by 130% in the constrained case.

The present paper brings together openly available datasets and simulators for testing of process monitoring and fault diagnosis techniques. Some general characteristics of these benchmark models and datasets are then assessed, including the richness of the information, availability of real data, ease of use, challenges that must be considered and potential for scientific advances. A critical comparison is made based on Exploratory Data Analysis (EDA) of readily available datasets. Previous works that used each benchmark are also reviewed and discussed. This reference can be used by researchers and practitioners who wish to choose a reference benchmark that best suits their particular interests and needs.

Environmental and economic pressure leads to an increasing desire to operate processes over various operational conditions, adapting to changing conditions such as feed quality, available energy, customer demand, or product prices. This requires frequent changes in the process setpoints, involving transitions between those. We focus on the explicit utilization of varying storage levels between process elements to increase the flexibility and thus the operational space of the overall process. To do so, we propose a so-called closed-loop static real-time optimization (RTO). It extends classical static RTO by integrating the time evolution of the storage levels and the underlying storage level controls. This enables the optimum and safe operation of the production plant, taking into account storage size limitations and filling levels of the elements. We approximate the dynamic effect of the controlled storage elements on the process dynamics by a suitable mapping, which allows us to capture the transition behavior of the storage element. The mapping can be, for instance, given by inversion-based control ideas or learned, e.g., by Gaussian processes. At the same time, the approach allows minimizing the dynamic effects of the storages on the overall system. The proposed formulation requires the efficient solution of a bilevel optimization problem. Thus, we will discuss a decomposition strategy to solve the overall closed-loop static RTO problem efficiently. A process example illustrates the effectiveness of the approach.

Due to the ability to handle constraints systematically and predict system evolution with models, model predictive control (MPC) methods have been widely studied and implemented in many industries. At the same time, low-cost MPC has received widespread attention due to its simple principle and easy implementation. This paper proposes a new low-cost MPC method and uses this method to put forward new insights on the performance improvement of complex dynamic system control. Firstly, using the concept of preprocessing, a novel MPC prediction structure is proposed under the independent model mode. Then through rigorous proofs, the properties of the proposed MPC algorithm are analyzed. Finally, through the study of three industrial cases, the proposal’s deployment procedure and efficacy are illustrated in detail. Compared with other low-cost predictive control algorithms, the effectiveness of the method has been presented.

Regression analysis constitutes an important tool for investigating the effect of explanatory variables on response variables. When outliers and bias errors are present, the weighted least squares estimator can perform poorly. For this reason, alternative robust techniques have been studied in several areas of science. However, often these different scientific communities are disconnected from each other, culminating in the scarcity of knowledge exchange among these areas. Thus, this paper presents a review on robust M-estimators in various knowledge areas. 50 (48 robust) M-estimators are illustrated, including the Weighted Least Squares estimator (non-robust), the Contaminated Normal estimator (quasi-robust), the Huber estimator (monotone), the Correntropy estimator (soft-redescending), the Smith estimator (hard-redescending), and the adaptive Barron and Generalized T-distribution. The mathematical functions that describe the estimators and their respective graphical forms are presented. The tuning constants of all these estimators, for 90%, 95%, 98%, and 99% relative efficiency levels in respect to the Normal distribution are also presented.