No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Performance Monitoring and Retuning for Cascaded Control Loops
Abstract
In this work, the problem of monitoring the performance of control loops and retuning is addressed. Poor performance of control loops reduces product quality, generates losses and can affect safety in industrial processes. There is a great interest in the industry in the search for methods that are simple to use and interpret. Cascaded control loops are widely used because they mitigate the effect of disturbances and process variations on the controlled variables. The fast dynamics of the inner loop, whose reference is generated by the controller of the outer loop, represents a challenge when one wants to investigate its performance during operation. The methods proposed here are completely data-based, requiring the user only to provide operating data and to approve the performance of the control loops to be monitored, which must always be done in some way. Both monitoring and retuning are validated by statistical tests, reducing the effect of uncer- tainties introduced by noise that is always present in industrial processes. The data used in all steps are generated by applying small disturbances added to the external loop control signal, reducing the deviations of the controlled variables from their references. The methodology is applied to a pilot plant with industrial equipment, and controlled by a distributed control system. A flow loop acts as an inner loop for a level control loop. The obtained results show that the methodology allows successfully attacking the proposed problem, being additionally replicable in industrial environments in general.
ResearchGate has not been able to resolve any citations for this publication.
In this paper, the current research status of controller performance assessment is reviewed in brief. Solving the problem of proportional–integral–derivative performance assessment usually requires step response data, and several methods are combined and extended. Using the integral of signals, implicit model information contained in process response data becomes explicit, and then the least squares approach is adopted to construct a detailed low-order process model based on process response data in more general types. A one-dimensional search algorithm is used to attain better estimation of process time delay, and integral equation approach is extended to be useful for more general process response. Based on the obtained model, a performance benchmark is established by simulating model output. Appropriate retuning methods are selected when the index of absolute integral error (IAE) indicates bad performance. Simulations and experiments verify the effectiveness of the proposed method. Issues about estimation of process time delay, data preprocessing, and parameter selection are studied and discussed.
Prediction error identification requires that data be informative with respect to the chosen model structure. Whereas sufficient conditions for informative experiments have been available for a long time, there were surprisingly no results of necessary and sufficient nature. With the recent surge of interest in optimal experiment design, it is of interest to know the minimal richness required of the externally applied signal to make the experiment informative. We provide necessary and sufficient conditions on the degree of richness of the applied signal to generate an informative experiment, both in open loop and in closed loop. In a closed-loop setup, where identification can be achieved with no external excitation if the controller is of sufficient degree, our results provide a precisely quantifiable trade-off between controller degree and required degree of external excitation.
The aim of this paper is to present analytic rules for PID controller tuning that are simple and still result in good closed-loop behavior. The starting point has been the IMC-PID tuning rules that have achieved widespread industrial acceptance. The rule for the integral term has been modified to improve disturbance rejection for integrating processes. Furthermore, rather than deriving separate rules for each transfer function model, there is a just a single tuning rule for a first-order or second-order time delay model. Simple analytic rules for model reduction are presented to obtain a model in this form, including the “half rule” for obtaining the effective time delay.
This paper addresses control performance monitoring (CPM) and degradation recovering in automatic control systems. It begins with a re-visit of CPM techniques and a summary of the major limitations of the existing CPM methods. They are (i) deficit in assessing control performance degradation caused by different types of disturbances and environment uncertainties, (ii) incapability for predicting performance degradation, and (iii) deficiency of efficient performance degradation recovering methods. In order to meet increasing demands of next generation automatic control systems for higher system performance, novel CPM methods have been developed in recent years, including performance assessment of control systems with deterministic disturbances and uncertainties, prediction of control performance degradation, and recovery of control performance degradation. Some of these methods and algorithms are introduced in the second part of this paper. The basis of these methods is a so-called residual centred model of feedback control systems, which allows a unified handling of control, monitoring and diagnosis in feedback control systems corrupted by disturbances and uncertainties. The focuses of these methods are on (i) introduction of the loop performance degradation index for the assessment and prediction of performance degradation in automatic control systems, (ii) predictive detection and estimation of loop performance degradation, and (iii) a data-driven performance degradation recovering scheme. The paper is concluded by a short summary of three future perspective topics, (i) prediction of economic system performance monitoring and estimation, (ii) reinforcement learning aided system performance recovery, and (iii) CPM digital twin.
This study deals with an updating algorithm for a database-driven proportional-integral-derivative (DD-PID) controller that uses a database for tuning control parameters.PID controllers are still used in many process systems including chemical processes.However, if systems exhibit nonlinearity, PID controllers with fixed PID parameters cannot achieve the desired control performance when the systems' equilibrium points are changed by set point changes.The DD-PID controller has been proposed to solve this problem.This controller can realize good control performance for nonlinear systems because it updates the PID parameters in its database so that the control performance around each equilibrium point has a desired characteristic.However, many experiments have to be performed for this controller to obtain a suitable database.This paper proposes a new offline database updating method called data-driven extended fictitious reference iterative tuning (DD-E-FRIT) method.In this method, the E-FRIT method, which is a direct control parameter tuning method for the linear system, is applied to the updating rule of the DD-PID controller.The DD-E-FRIT method can update a database offline and obtain the database for realizing the desired tracking property of a closed-loop system by storing one-shot operating data into the database.
A novel data-driven method for simultaneous performance assessment and retuning of PID controllers is presented in this study. The PID tuning problem is cast as a convex approximation problem of a reference model, which makes the resulting closed-loop step response resemble that of a predefined reference model. The proposed data-driven method only requires the current PID controller parameters and a set of closed-loop data with setpoint changes, while the process model is neither necessary to be known a priori nor to be identified directly. The controller effort limitation, which is closely related to variations of controller actions, is an important concern for controller design. In order to evaluate PID controller performance subject to control effort limitations, a trade-off curve displaying the integrated squared error (ISE) against the total squared variation (TSV) of controller actions is obtained, which can be calculated in a cost efficient way and serves as a useful tool for performance assessment and retuning of PID controllers. Illustrative case studies are presented to demonstrate the applicability of the proposed method.
When process control systems do not perform as they should, plant personnel will not obtain positive results and may even give up using some control loops entirely. To increase productivity and efficiency it must be ensured that the control system is used effectively. The best way to do this is to turn the automatic control on and to tune it correctly. This paper describes the expertise in control loop performance monitoring gathered by ABB’s control experts. With more automated production processes and fewer experts, control loop performance monitoring (LPM) needs to be able to evaluate one hundred loops at a time instead of loop-by-loop analysis. This cannot be done with training or manual tools. In this paper, ABB’s perspective on loop performance monitoring as well as several novel features is described. Finally, further research directions are highlighted.
This paper studies direct (non-iterative) data-based method for Model Reference (MR) control design. It shows that the optimal controller can be obtained as the solution of a Prediction Error (PE) identification problem that directly estimates the controller parameters through a reparametrization of the input-output model. The standard tools of Prediction Error Identification can thus be used to analyze the statistical properties (bias and variance) of the estimated controller. It also shows that, for MR control design, direct and indirect data-based methods are essentially equivalent.
In this article, we propose an algorithm for the set-point-following performance assessment of a PID controller and for the retuning of the parameters in case the obtained response is not satisfactory. The technique is based on the SIMC tuning rules and can be applied simply by evaluating a set-point step response. Simulation and experimental results obtained with laboratory-scale equipment show the effectiveness of the methodology.
This paper aims at assessing the setpoint tracking performance of PID control loops, under certain constraints such as the processes under control are linear-time invariant and the instrumentations of control loops work properly. The lower bounds of integrated absolute errors (IAEs) are established, based on the widely used internal model control (IMC) principle, from closed-loop responses subject to setpoint changes in the form of step, ramp or other general types. Taking the lower bound as a benchmark, an IMC-IAE-based index is proposed to assess the setpoint tracking performance of PID control loops. Numerical and experimental examples, as well as an industrial case study, are provided to verify the lower bound as the performance benchmark and to illustrate the effectiveness of the proposed performance index.Highlights► Assessment of the setpoint tracking performance for PID control loops. ► The lower bound of integrated absolute errors. ► The IMC-IAE-based index. ► Setpoint changes in the form of step, ramp or other general types.
The single loop control system with controller of general structure or with PI/PID structure is studied. The control task of the system is to track step set-point changes or reject the intermittent step output disturbances. Performance of a simple feedback system is assessed with its IAE value (J) and its rise time (t r) observed from the response of the system to a step set-point change. Assume that the dynamics of open loop processes can be represented by models of first order or second order plus dead time (i.e., FOPDT or SOPDT). On the basis of these models, the optimal IAEs and the associated rise times are computed. The performance of a system is assessed by comparing its current IAE to the optimal IAEs. An index is thus defined for this quantitative assessment. To be free of a process model for assessment, envelopes of optimal IAE and the rise time are prepared. A method to estimate the step response for set-point tracking by making use of the response of the system to dither inputs is presented. By introduction of dither inputs intermittently to the system, the performance of the single loop can be monitored on-line.
For processes described by linear transfer functions with additive disturbances, the best possible control in the mean square sense is realized when a minimum variance controller is implemented. It is shown that an estimate of the best possible control can be obtained by fitting a univariate time series to process data collected under routine control. No ‘identifiabüity’ constraints need be imposed. The use of this technique is demonstrated with pilot plant and production data.
Pour les procédés décrits par des fonctions de transfert linéaires avec des perturbations additives, le meilleur contrôle possible au sens des moindres carrés s'obtient avec un régulateur à variance minimum. On montre qu'on peut obtenir une estimation du meilleur contrôle possible en adaptant une série chronologique univariée pour traiter des données recueillies lors d'un rétrocontrôle de routine. Il n'est pas nécessaire d'imposer des contraintes d'«identificabilité». L'utilisation de cette technique est décrite par des données d'une usine pilote et des données de production.
The sections in this article are1The Problem2Background and Literature3Outline4Displaying the Basic Ideas: Arx Models and the Linear Least Squares Method5Model Structures I: Linear Models6Model Structures Ii: Nonlinear Black-Box Models7General Parameter Estimation Techniques8Special Estimation Techniques for Linear Black-Box Models9Data Quality10Model Validation and Model Selection11Back to Data: The Practical Side of Identification
Incluye bibliografía e índice
A new methodology is developed to assess the performance of PI
controllers from closed-loop response data for a setpoint step change.
It is based on two new dimensionless performance indices, the
dimensionless settling time and the dimensionless integral of the
absolute value of the error. The methodology is used to quantify how far
a control loop is from the best achievable performance of PI control. It
also identifies poorly performing control loops, such as those that are
excessively sluggish. The performance and robustness properties of the
internal model control (IMC) tuning method are analyzed and are used as
industrially relevant benchmarks. The proposed methodology is also
applicable to PID controllers