Jesus Flores-Cerrillo

Data-based latent variable methods for process analysis, monitoring and control

This paper gives an overview of methods for utilizing large process data matrices. These data matrices are almost always of less than full statistical rank, and therefore, latent variable methods are shown to be well suited to obtain useful subspace models from them for treating a variety of important industrial problems. An overview of the important concepts behind latent variable models is presented and the methods are illustrated with industrial examples in the following areas: (i) the analysis of historical databases and trouble-shooting process problems; (ii) process monitoring and FDI; (iii) extraction of information from novel multivariate sensors; (iv) process control in reduced dimensional subspaces. In each of these problems, latent variable models provide the framework on which solutions are based.

Integration of control theory and scheduling methods for supply chain management

Abstract In this paper, we propose to use distributed model predictive control for supply chain optimization. In particular, we focus on inventory management in supply chains. We use cooperative model predictive control, in which each agent makes their local decisions by optimizing the overall supply chain objective. Motivated by recent results in Stewart, Wright, and Rawlings (2011) , we develop a new cooperative MPC algorithm that is applicable to any stabilizable system, and in particular to supply chain models. We illustrate cooperative MPC for a two node supply chain example and compare its performance and properties with other classical distributed operating policies.

A non-Gaussian pattern matching based dynamic process monitoring approach and its application to cryogenic air separation process

Abstract Principal component analysis (PCA) based pattern matching methods have been applied to process monitoring and fault detection. However, the conventional pattern matching approaches do not specifically take into account the non-Gaussian dynamic features in chemical processes. Furthermore, those techniques are more focused on fault detection instead of fault diagnosis. In this study, a non-Gaussian pattern matching based fault detection and diagnosis method is developed and applied to monitor cryogenic air separation process. First, independent component analysis (ICA) models are built on the normal benchmark and monitored data sets along sliding windows. The IC subspaces from the benchmark and monitored data are then extracted to evaluate the non-Gaussian patterns and detect process faults through a mutual information based dissimilarity index. Further, a difference subspace between the two IC subspaces is computed to characterize the divergence of the dynamic and non-Gaussian patterns between the benchmark and monitored data. Subsequently, the mutual information between the IC difference subspace and each process variable direction is defined as a new non-Gaussian contribution index for fault identification and diagnosis. The presented approach is applied to a simulated cryogenic air separation plant and the monitoring results are compared against those of PCA based pattern matching techniques and ICA based monitoring method. The application study demonstrates that the developed non-Gaussian pattern matching approach can effectively monitor the complex air separation process with superior fault detection and diagnosis capability.

Subspace-Based Model Identification of a Hydrogen Plant Startup Dynamics

Abstract This work addresses the problem of determining a data-driven model for the startup of a hydrogen production unit, and demonstrates the approach both on a detailed first principles simulation model and by application to real data. To this end, first a detailed first principles model of the hydrogen plant is developed in Honeywells UniSim design by adapting the plant standard operating procedure (SOP). Illustrative simulations are next presented to establish the meaningfulness of approximating process nonlinearity with a (higher order) linear time invariant (LTI) model. Then an LTI data-driven model of the hydrogen unit startup process using subspace identification based methods is identified. The framework is then implemented and successfully validated data on simulated data and on data from an industrial hydrogen unit.

Practical optimization for cost reduction of a liquefier in an industrial air separation plant

Abstract Commercial and in-house simulation software used by industrial practitioners are often of a “black box” type from which derivatives cannot be directly obtained. This paper demonstrates a linkage between available industrial tools and cost reduction opportunity creation through the application of a derivative-free optimization technique. An operational liquefier in an air separation unit is used in our study due to the increasing importance of liquid production in the plants overall operation strategy, and limited evaluation on the operation of such systems under disturbances. Particle swarm optimization is implemented, and optimization results show that when the plant is forced to operate away from its nominal operating/design conditions, it is possible to reduce the unit power consumption by adjusting different operation set-points. A reference map is generated to guide the operation under selected realizations of cooling water temperature, production load and feed conditions.

An independent component analysis and mutual information based non-Gaussian pattern matching method for fault detection and diagnosis of complex cryogenic air separation process

The conventional principal component analysis (PCA) based pattern matching methods have been applied to dynamic process monitoring. However, they do not take into account the non-Gaussian features in industrial processes and are also more focused on fault detection instead of fault diagnosis. In this paper, an independent component analysis and mutual information based non-Gaussian pattern matching approach is developed for fault detection and diagnosis of complex chemical processes. The presented approach is applied to a simulated cryogenic air separation process and the application study demonstrates that the developed non-Gaussian pattern matching method can effectively monitor the complex air separation process with strong capability of fault detection and diagnosis.

Semi-Batch Trajectory Control in Reduced Dimensional Spaces

Abstract A novel inferential strategy for controlling end-product quality properties using complete trajectories of manipulated variables is presented. Control through complete trajectory manipulation using empirical models only is possible by controlling the process in the reduce space (scores) of a latent variable model rather than in the real space of the manipulated variables. Model inversion and trajectory reconstruction is achieved by exploiting the correlation structure in the manipulated variable trajectories captured by a Partial Least Squares (PLS) model. The approach is illustrated with a condensation polymerisation example for the production of nylon. The data requirements for building the model are shown to be modest.

Development of a high fidelity and subspace identification model of a hydrogen plant startup dynamics

In this work, the problem of determining a data-driven model of a hydrogen production unit is addressed. The framework is applied to a high fidelity simulation model developed in this work. To this end, first a high fidelity model of the entire plant is developed in Honeywells UniSim Design, capable of simulating the startup and shutdown phase, with appropriate adaptation of the plant standard operating procedure (SOP). Several startups are simulated to generate training data for identification of a data-driven model. Then an LTI data-driven model of the process using subspace identification based methods is determined and validated against new simulated startup. Simulation results demonstrate the prediction capabilities of the identified model.

Model Predictive Control for Batch Processes Using Latent Variable Methods

Abstract A novel multivariate model predictive control strategy for trajectory tracking and disturbance rejection for batch processes, based on multi way PCA models, is presented. It directly computes the manipulated variable trajectory adjustments over a future horizon using the structure of the PCA model. The advantages and the modest data requirements are illustrated using an emulsion polymerization process for temperature tracking.