Systems and Control

For model-based production control and optimisation it is crucial to properly identify those input variables that have the strongest influence on production performance. This way, production operators can focus only on the relevant variables, and production control problems can be reduced. In order to identify previously unknown relationships among the production variables, hidden knowledge in historical production data needs to be explored. In the article, two decisive steps are considered. First, an input variable selection methodology, typically applied for selecting model regressors, is applied. Next, the appropriateness of the selected inputs and their manipulative strength is validated by an operating-space-based controllability analysis. To use the most appropriate input variable selection approach, different input selection methodologies are compared with synthetic datasets. Moreover, a case study of Tennessee Eastman process is applied to demonstrate a complete input variable selection procedure for model-based production control and optimisation. Keywords model-based production control • holistic production control • input variable selection • controllability

Correct planning is crucial for efficient production and best quality of products. The planning processes are commonly supported with computer solutions; however manual interactions are commonly needed, as sometimes the problems do not fit the general-purpose planning systems. The manual planning approach is time consuming and prone to errors. Solutions to automatize structured problems are needed. In this paper, we deal with material requirements planning for a specific problem, where a group of work orders for one product must be produced from the same batch of material. The presented problem is motivated by the steel-processing industry, where raw materials defined in a purchase order must be cut in order to satisfy the needs of the planned work order while also minimizing waste (leftover) and tardiness, if applicable. The specific requirements of the problem (i.e., restrictions of which work orders can be produced from a particular group of raw materials) does not fit the regula...

Abstract:-In the paper a velocity-based linearisation approach for a design of a gain-scheduling controller for ship steering is presented. While ship steering and gain-scheduling are well known problems with welldeveloped solutions, the described method differs from previous approaches to gain-scheduled controller design in that the closed-loop system with nonlinear controller retains properties of the linearised system with linear controller designed in one of the design stages. Stochastic robustness analysis with Monte Carlo ...

Exact or approximate solutions to constrained linear model predictive control problems can be pre-computed off-line in an explicit form as a piecewise linear state feedback defined on a polyhedral partition of the state space. This leads to efficient real-time computations and admits implementation at high sampling frequencies in real-time systems with high reliability and low software complexity. In this paper, an explicit model predictive controller for a gas-liquid separation plant is designed and experimentally tested.

This paper presents an innovative self-tuning nonlinear controller ASPECT (advanced control algorithms for programmable logic controllers). It is intended for the control of highly nonlinear processes whose properties change radically over its range of operation, and includes three advanced control algorithms. It is designed using the concepts of agent-based systems, applied with the aim of automating some of the configuration tasks. The process is represented by a set of low-order local linear models whose parameters are identified using an online learning procedure. This procedure combines model identification with pre-and postidentification steps to provide reliable operation. The controller monitors and evaluates the control performance of the closed-loop system. The controller was implemented on a programmable logic controller (PLC). The performance is illustrated on a field test application for control of pressure on a hydraulic valve. r

The Gaussian process model is an example of a flexible, probabilistic, nonparametric model with uncertainty predictions. It offers a range of advantages for modelling from data and has been therefore used also for dynamic systems identification. One of the noticeable drawbacks of the system identification with Gaussian process models is computation time necessary for modelling. The modelling procedure involves the inverse of covariance matrix which is as large as the length of input samples vector. The computation time for this inverse regardless of the use of efficient algorithm is rising with the third power of input data number. Intensive research is going on for finding algorithms that would accelerate the training of Gaussian process models. The purpose of this paper is to show approach from the used hardware point of view. The assessment of computational efficiency of two different hardware platforms for GP model identification are given in the paper. These are: single core personal computer and personal computer with graphic card used for computations. The assessment has been done with comparison of computational load on a toy case study of nonlinear dynamic system identification. The assessment reveals that the parallel computation solutions are efficient for larger amount of data when the initial and communication overhead of parallel computation becomes sufficiently small part of the whole process.

In this paper, an approach to multivariable combustion control design within the Individual Channel Design (ICD) framework for analysis and control design is presented. ICD is a framework which involves an interplay between customer specification, uncertain plant characteristics, and the multivariable feedback design itself. Established multivariable methods and process engineering knowledge can be incorporated or evaluated within the ICD framework. The combustion control has been designed and evaluated with a computer simulation of both a linearized model and a nonlinear model of the closed-loop system. The ICD multivariable framework shows in a highly transparent manner, by way of simple graphical frequency response indicators, what the main possibilities and difficulties posed by a combustion process for multivariable control are, and how much trade-off between control specifications is possible. Solutions are also presented for problems such as: integrity of closed-loop control, balance of input-output channels, simple and transparent controller structure, and robustness.

In this contribution a possible approach for designing multiple blended controller for wastewater treatment reactor is presented. Velocity-based linearisation framework for analysis of non-linear systems is used to assist control design. Due to this analysis tool the design of relatively simple controller is possible. The design procedure enables use of knowledge from linear systems control design and achieving desired performance not just in equilibrium points where system can be linearised but in whole operating region. This approach has confirmed its advantages on the wastewater treatment reactor control problem in tracking and regulatory control of dissolved oxygen level.

Gaussian process Pareto front a b s t r a c t This paper presents a multi-criteria evaluation methodology for determining the operating strategies for bio-chemical, wastewater treatment plants based on a model analysis under an uncertainty that can present multiple steady states. The method is based on Monte Carlo (MC) simulations and the expected utility theory in order to deal with the analysis of choices among risky operating strategies with multi-dimensional outcomes. The motivation is given by a case study using an anaerobic digestion model (ADM) adapted for multiple co-substrates. It is shown how the multi-criteria analyses' computational complexity can be reduced within an approximation based on Gaussian-process regression and how a reliability map can be built for a bio-process model under uncertainty and multiplicity. In our uncertainty-analyses case study, the reliability map shows the probability of a biogas-production collapse for a given set of substrates mixture input loads. ª

A Control Performance Monitor (CPM) as a software agent is developed for assessment the SISO control loops behaviour. A dedicated system for data pre-processing is applied in order to guaranty robust properties of the extracted process features. Using a fuzzy and rule based evaluation procedure an overall performance index is also calculated from the estimated indicators. The CPM provides early warning information when an inadmissible performance is detected. Some results of experimental verification of the CPM are reported.

Gaussian process models provide a probabilistic non-parametric modelling approach for black-box identification of nonlinear dynamic systems. The Gaussian processes can highlight areas of the input space where prediction quality is poor, due to the lack of data or its complexity, by indicating the higher variance around the predicted mean. Gaussian process models contain noticeably less coefficients to be optimized. This chapter illustrates possible application of Gaussian process models within model predictive control. The extra information provided by the Gaussian process model is used in predictive control, where optimization of the control signal takes the variance information into account. The predictive control principle is demonstrated via the control of a pH process benchmark.

This paper describes model-based predictive control based on Gaussian processes. Gaussian process models provide a probabilistic nonparametric modelling approach for black-box identification of non-linear dynamic systems. It offers more insight in variance of obtained model response, as well as fewer parameters to determine than other models. The Gaussian processes can highlight areas of the input space where prediction quality is poor, due to the lack of data or its complexity, by indicating the higher variance around the predicted mean. This property is used in predictive control, where optimisation of control signal takes the variance information into account. The predictive control principle is demonstrated on a simulated example of nonlinear system.

ABSTRACT This paper addresses the task of identification of nonlinear dynamic systems from measured data. The discrete-time variant of this task is commonly reformulated as a regression problem. As tree ensembles have proven to be a successful predictive modeling approach, we investigate the use of tree ensembles for solving the regression problem. While different variants of tree ensembles have been proposed and used, they are mostly limited to using regression trees as base models. We introduce ensembles of fuzzified model trees with split attribute randomization and evaluate them for nonlinear dynamic system identification. Models of dynamic systems which are built for control purposes are usually evaluated by a more stringent evaluation procedure using the output, i.e., simulation error. Taking this into account, we perform ensemble pruning to optimize the output error of the tree ensemble models. The proposed Model-Tree Ensemble method is empirically evaluated by using input–output data disturbed by noise. It is compared to representative state-of-the-art approaches, on one synthetic dataset with artificially introduced noise and one real-world noisy data set. The evaluation shows that the method is suitable for modeling dynamic systems and produces models with comparable output error performance to the other approaches. Also, the method is resilient to noise, as its performance does not deteriorate even when up to 20% of noise is added.