Search

Fault detection and diagnosis (FDD) plays an important role in risk and safety management system. According to the ways faults influence the actual system, they can be divided into additive (mainly actuator and sensor faults) and multiplicative (mainly process faults). The two types of faults should be diagnosed with specific methods then handled with different maintenance strategies. This paper presents a combined passive-active fault diagnosis method, which allows a simultaneous consideration of additive and multiplicative faults and to distinguish between them. It also enables detailed diagnosis that assists to enhance the subsequent risk assessment and management. System identification is used as the modeling tool and forms the basis of the method. The passive-active feature of the method reflects in that: it uses online generated residual as a fault indicator for real-time monitoring; it also uses test signals to magnify the fault characteristics and helps to reveal the fault location. Specifically, a method is proposed to distinguish between additive and multiplicative fault according to the different residual behavior after adding test signals. Once fault type is determined, by investigating the identified models with error bounds, a method is further developed to determine the location of multiplicative fault in the multi-variable system. The statistical properties of the identified models are utilized to generate perturbed realizations of the model and derive probabilistic bounds of model errors, enabling both methods to deal with model errors. The proposed method does not require to break the control loops when adding test signals and does not require fault data/model to start with. The effectiveness of the proposed method is validated through a numerical example and Tennessee Eastman process (TEP). [ABSTRACT FROM AUTHOR]

Optimal detection filters can greatly enhance fault detection performance, but designing these filters requires fault data which is difficult to obtain in practice. This paper proposes a scheme that automatically determines the optimal detection filter from a filter bank online without using fault data. The method can improve fault detection rate and accelerate detection speed. In order to reduce the false alarm rate, a method of threshold setting is introduced based on kernel density estimation. Implementation issues concerning filter bank design and online decision rule are also discussed. The method is validated in a numerical example and Tennessee Eastman process, and its performance is compared to those of other state-of-the-art methods. • Enabling automatic determination of optimal fault detection filter. • Using kernel density estimation to refine the thresholds. • Enhancing fault detection performance without using fault data. • Performance validated in Tennessee Eastman benchmark. [ABSTRACT FROM AUTHOR]

In this work, an identification based fault detection method is proposed. The idea is to identify a dynamic process model from test data and to generate residuals using the identified model for fault detection. The method intends to improve fault detection performance while taking disturbance and model error into account. To this end, a fault detection performance index is introduced in a statistical framework. Then it is shown that the output error residual is more suitable for fault detection than the prediction error residual. Further an optimal detection filter maximizing the performance index is developed. Practical issues for implementing the detection filter are also addressed. Finally, the proposed method is illustrated through a numerical example and Tennessee Eastman process. • Proposing an identification based fault detection method. • Demonstrating output error residual is more suitable than the prediction error residual. • Developing optimal detection filters to enhance the detection performance. • Comprehensive case studies including method illustrations and comparison with other methods.. [ABSTRACT FROM AUTHOR]

The dynamic performance (or flexibility) of the combined heating and power (CHP) unit operating in a heat-led mode is restricted by the strong interdependence between electricity generation and thermal generation. This paper proposes a method that balances the flexibility and the heating quality using a coupled two-timescale model predictive control (MPC) scheme. The indoor temperatures of thermal users are regulated in a comfort region by a slow timescale MPC using zone control; a fast MPC is used as the coordinated control system (CCS). The two controllers are coupled in that both controllers use the extraction steams as their controlled variable. In this way, the dynamic flexibility of the CHP unit is improved by utilizing energy buffers from district heating systems and buildings. The proposed scheme is verified using a simulation study on a real heating system with 2 CHP units, 28 thermal nodes, and 58 pipelines. The result shows that the dynamic performance of the system is improved considerably while the temperatures of thermal users remain in their comfort regions. [Display omitted] • An MPC scheme is proposed to optimize the dynamic flexibility of CHP systems. • Two coordinated MPC controllers are used to solve the problem. • Zone control is used in order to give more room for electrical power control. • The method is verified using simulations of a CHP system. [ABSTRACT FROM AUTHOR]

Summary: Identification‐based proportional, integral, and derivative (PID) controller tuning is studied in this work. A method of closed‐loop PID tuning without using test signals is developed. The basic idea is to shift the PID control parameters properly to achieve persistent excitation condition in closed‐loop test. In this way, not only the closed‐loop system can be identified accurately without using external exciting signals but also the control performance can be improved during the test. A stochastic model error bound is introduced for model validation. An internal model control design method is applied to tune the PID parameters. The effectiveness of the proposed approach is illustrated by simulations. [ABSTRACT FROM AUTHOR]

Identification of systems with slowly sampled output is studied. A linear parameter varying (LPV) model with multi-model structure is used to solve the problem. The output error (OE) method is used to estimate model parameters. Firstly, the local models and weighting functions are estimated separately using optimization methods. Then, a relaxation iteration method is developed to refine the parameters of the total model. For LPV model structure determination, an engineering approach is proposed that combines process knowledge with the so-called final output error criteria (FOE). The method is verified using both simulation data and industrial data. In the industrial case study, the LPV models give more accurate prediction of product qualities than that of a linear dynamic model and that of a static nonlinear model; the result also indicates the necessity of using test signals in soft-sensor development. [ABSTRACT FROM AUTHOR]

In this work, two fundamental issues in system identification, sampling frequency and anti-aliasing filtering are revisited, questions like, “Can higher sampling frequency help increase model accuracy?”, and, “Can we do better than anti-aliasing filtering?” will be answered. First it is shown that, the traditional anti-aliasing filtering procedure does not give a consistent estimate of the desired model. Then an identification method is proposed based on the over-sampling scheme where a high frequency model is first identified and then converted to the (lower) control sampling frequency. It is shown that when the output noise contains energy beyond the bandwidth of the plant, the proposed method performs anti-aliasing in both open-loop and closed-loop identification; and in closed-loop identification, extra excitation is achieved from the high frequency noise. Simulation studies are used to illustrate the findings. The result has important implications in control applications. [ABSTRACT FROM AUTHOR]

This work studies the identification of errors-in-variables (EIV) systems. An asymptotic method (ASYM) is developed for the EIV system. Firstly, an auto regressive with exogeneous (ARX) model estimation method is proposed, which is consistent for EIV systems. Then the asymptotic variance expression of the estimated high-order ARX model is derived, which forms the basis of the ASYM method. In parameter estimation, the ASYM starts with a high-order ARX model estimation followed by a frequency domain weighted model reduction. The obtained model is consistent, and its efficiency needs to be investigated. Besides parameter estimation, a criterion for model order selection is proposed, which is based on frequency domain considerations, and the frequency domain error bound is established that can be used for model validation. Simulations and comparisons with other methods are used to illustrate the performance of the method. [ABSTRACT FROM AUTHOR]

This paper studies sensor and actuator fault diagnosis for a class of linear time-invariant (LTI) industrial control systems. Based on the dynamic model obtained using system identification, a method is proposed that can diagnose all sensor and actuator faults in the control system when they do not happen simultaneously. First, a structured residual design method is developed for coupled systems. Then model errors that occur in system identification are considered and the robust isolability is analyzed. Finally, online decision rules are given which are easy to implement. The method does not require fault data and can deal with small faults that cannot be detected by inspecting input–output signals. Verification in the high temperature gas-cooled reactor pebble-bed module (HTR-PM) simulation platform shows good potential of the method for complex industrial control systems. [ABSTRACT FROM AUTHOR]

In this work, firstly, an asymptotic variance expression is derived for the transfer function estimates in the over-sampling based identification scheme. Then the result is used to analyze the over-sampling scheme. The asymptotic variance expression says that the joint covariance matrix of the transfer function estimates is proportional to the “generalized” noise-to-signal ratio. It is an application of the result in Ljung (1985); and it covers both open-loop and closed-loop tests. Using the result, one can show that informativity of the closed-loop tests without external excitation can be attained by using the over-sampling scheme, but only when the output disturbance has high frequency content beyond the bandwidth of the plant. Moreover, when the output disturbance has high frequency content and when test signal is used, the result points out that the over-sampling scheme can increase model accuracy, and, it will outperform conventional method with anti-aliasing filtering. Numerical examples are used to illustrate the asymptotic variance expression and the analysis about over-sampling scheme of identification. [ABSTRACT FROM AUTHOR]

This work is concerned with the minimum variance (MV) benchmark estimation without a prior knowledge of time-delay. The idea is to use a small high-frequency test signal in order to obtain the accurate time-delay estimation (TDE) result, then the MV benchmark can be calculated from the closed-loop data. Based on the analysis, a well designed test signal for control performance monitoring (CPM) is not only possible but also necessary, as it increases the identifiability of the closed-loop data without having a large effect on routine operation. A new TDE algorithm by using high-order auto regressive exogenous (ARX) model is proposed as well as the multivariable issue is investigated. Simulation examples of performance monitoring with low signal-to-noise ratio (SNR) are employed to demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

An algorithm for identification of single-input single-output Box–Jenkins models is presented. It consists of four steps: firstly a high order ARX model is estimated; secondly, the input–output data is filtered with the inverse of the estimated disturbance model; thirdly, the filtered data is used in the Steiglitz–McBride method to recover the system dynamics; in the final step, the noise model is recovered by estimating an ARMA model from the residuals of the third step. The relationship to other identification methods, in particular the refined instrumental-variable method, are elaborated upon. A Monte Carlo simulation study with an oscillatory system is presented and these results are complemented with an industrial case study. The algorithm can easily be generalized to multi-input single-output models with common denominator. [ABSTRACT FROM AUTHOR]

Most real-time optimization (RTO) methods in process industries are based on rigorous (mechanistic) models or on some special performance tests. Both methods are difficult to apply due to low model accuracy and high costs in modeling, in computation or in performance tests. In this work, a real-time optimization method based on system identification is developed. No rigorous models or special performance tests are needed in this approach, making it cost-effective and applicable in various process industries and especially suitable for energy and utility systems. First, the selection of objective function and of decision variables is discussed. Then the identification-based optimization method is proposed where a closed-loop multivariable identification approach is incorporated. The method is verified using a simulated power plant and is also partially applied to a real power plant. • A real-time optimization method based on system identification is developed. • The steady-state gradient is obtained from dynamic data using system identification. • The gradient uncertainty issue is solved using system identification theory. • The proposed method is verified using a simulated 1000 MW power plant and partially applied to a 600 MW power plant. [ABSTRACT FROM AUTHOR]

• Proposing a method of fault isolation filter design directly based on transfer-function models. • Formulating the isolation problem as a dual decoupling control problem and using a model predictive control (MPC) controller as the decoupler. • Efficient to handle large-scale processes with time delays and right-half-plane (RHP) zeros. • Model error is treated and optimal filters that suppress the disturbances are also designed. • Comprehensive case studies including a coal power generator and the Tennessee Eastman process (TEP). This work studies model-based fault isolation. A method of fault isolation filter design is developed. The fault to output transfer-function model is obtained using system identification. The basic idea is to formulate the isolation problem as a dual decoupling control problem then use a MPC controller as the decoupler. In this way an efficient algorithm can be developed to obtain the fault isolation filter; time delays and right-half-plane (RHP) zeros can be handled. Model errors that occur during system identification are considered in designing the isolation filter. Optimal fault isolation filters that suppress disturbances are also developed. The method is demonstrated using a simulated 600MW supercritical power generation unit and the Tennessee Eastman process (TEP). [ABSTRACT FROM AUTHOR]

Highlights: [•] k-Step-ahead prediction error model identification is studied. [•] It will be shown that a model estimated using a k-step-ahead prediction error criterion may not be optimal for k-step-ahead prediction. [•] A normal one-step-ahead prediction error criterion can be optimal. [•] For MPC identification it is suggested that one-step-ahead prediction error criterion is used. [•] Simulations and industrial data will be used to illustrate the relevance of the result. [ABSTRACT FROM AUTHOR]

Abstract: The high cost of model predictive control (MPC) technology has hampered its wide application in process industries beyond the refining/petrochemical industry. This work strives to increase the efficiency of MPC deployment. First, a semi-automatic MPC system is introduced. It consists of three modules: an MPC module, an online identification module and a control monitor module. The goal of the MPC technology is twofold: (1) to considerably reduce the cost of MPC commissioning and maintenance and (2) to increase control performance. System identification plays important roles in all the three parts of the MPC system. In the identification module, the so-called ASYM method of identification is used. It is demonstrated with an industrial application. In the control module, adaptive disturbance model identification is developed for improving control performance; in the monitor module, a method of model error detection method is developed. Industrial applications and simulations are used to demonstrate the ideas. Finally, we comment on some industrial needs on MPC research and development. [Copyright &y& Elsevier]

Abstract: This paper proposes an MPC method that uses an adaptive disturbance model to improve the accuracy of prediction. In unmeasured disturbance model identification, a novel multi-iteration pseudo-linear regression (MIPLR) method is used which is more accurate and has faster convergence than traditional recursive identification methods. The adaptive disturbance model is used in an MPC scheme for improved performance in disturbance rejection. The method is demonstrated by the simulation of a distillation column and also tested on the real process. The test results show that the proposed MPC scheme can not only increase control performance, but also increase robustness. [Copyright &y& Elsevier]

Abstract: Identification of systems with slowly and irregularly sampled output data is considered. The identified models can be used in inferential control or as soft sensor. Dual-rate system with slow output samples is a special case of the problem studied here. In this work, it will be show that an output error method is a natural choice for the problem that can identify the fast rate model directly from the fast input and slow and irregular output data. When the system is in the model set, consistence of the output error model is established and minimum variance property is proved. When the model order is lower than the true order, bias distribution in the frequency domain is given. Simulation studies and industrial data will be used to illustrate the method. [Copyright &y& Elsevier]

Multi-energy systems have been leaping forward for its various benefits, e.g., energy conservation and emission reduction. Multi-energy systems exhibit multi-time scaled characteristic and broad bandwidth, thereby causing difficulties in dynamic modeling. In this work, two-time scaled system identification is studied. A method is developed to solve the problem, which uses proper test signal design, signal pre-filtering and subtraction. The high and low frequency parts of the two-time scale system are identified separately and then combined to form the total system in a parallel structure. The consistency of the method is proved and case studies are used to verify the effectiveness of the method. • The difficulties of the multi-time scaled system identification are discussed. • A two-time scaled system identification method is developed. • The consistency of the proposed method is established. • Case study on a simulated multi-energy system is provided. [ABSTRACT FROM AUTHOR]

Abstract: The identification of a class of ill-conditioned processes is considered. The goal of identification is model predictive control (MPC). For this class of processes, it is essential to have good estimation of the very small difference between certain transfer functions, or, low gain direction. Two simple and practical test methods will be proposed that can enhance the model quality of low gain direction. The main idea is to use the high amplitude and high correlation test signals that excite the low gain direction more and that will not disturb process operation. The test signals can be used in both open-loop and closed-loop tests. A high purity distillation column model is used to show the effectiveness of the methods. [Copyright &y& Elsevier]

Estimation of a single-input single-output block-oriented model is studied. The model consists of a linear block embedded between two static nonlinear gains. Hence, it is called N–L–N Hammerstein–Wiener model. First, the model structure is motivated and the disturbance model is discussed. The paper then concentrates on parameter estimation. A relaxation iteration scheme is proposed by making use of a model structure in which the error is bilinear-in-parameters. This leads to a simple algorithm which minimizes the original loss function. The convergence and consistency of the algorithm are studied. In order to reduce the variance error, the obtained linear model is further reduced using frequency weighted model reduction. Simulation study will be used to illustrate the method. [Copyright &y& Elsevier]

In this work we study multivariable and closed-loop identification of industrial processes for use in model predictive control. The advantages of closed-loop identification are discussed and related problems of identification are outlined. Subsequently, two case studies are used to demonstrate the advantages of closed-loop identification. The first process is a simulated high purity distillation column; the second process is a deethanizer of an industrial scale ethylene unit. [Copyright &y& Elsevier]

Identification of single-input single-output Hammerstein models is studied in this work. The basic idea here is to extend the recently developed asymptotic method (ASYM) of linear model identification to include input non-linearity in the model set. First identification test design will be discussed. In parameter estimation, prediction error criterion is used in order to maintain consistence when the process is operating in closed-loop. A relaxation iteration scheme is proposed by making use of a model structure in which the error is bilinear in the parameters. The order of the linear part and nonlinear part are determined by looking at an output error related criterion which is control-relevant. The frequency domain upper error bound of the linear part will be derived and used for model validation. Simulation study will be used to illustrate the method and comparisons with other methods are also given. [ABSTRACT FROM AUTHOR]

Most MPC (Model Predictive Control) algorithms used in industries and studied in the control academia use a two-term QP (quadratic programming), where the first term is the weighted norm of the output errors, and the second term is that of the input increments. In this work, a DMC (Dynamic Matrix Control) algorithm that uses three-term QP is developed and studied, where the third term is the weighted norm of the output increments. A relationship between the three-term DMC and the two-term DMC is established; and based on that, ideal closed-loop response curves are derived. These results are useful for the tuning of the three-term DMC. Then, a method for comparing two control methods is proposed and it is shown that the three-term DMC outperforms the two-term DMC. The findings are verified using simulation studies. Finally, the three-term DMC is successfully applied to a 1030 MW ultra-supercritical coal-fired power plant. According to simulation study using the identified model, in comparison to traditional two-term DMC, there is a roughly 10% reduction in the standard deviation of load tracking error for similar control efforts. In real-life tests on the power plant, the performance of the three-term DMC is compared to that of the existing PID controller. The load rate of change increased from 0.8%/min to 1.5%/min while the standard deviation of steam pressure decreased 38%; the standard deviations of the main steam temperature and reheat steam temperature decreased more than 40% with their mean values increased by more than 2 ∘ C and 4 ∘ C respectively. • A DMC (Dynamic Matrix Control) algorithm that uses three-term QP is developed and studied. • A relationship between the three-term DMC and the two-term DMC is established. • A method for comparing two control methods is proposed and it is shown that the three-term DMC outperforms the two-term DMC. • The three-term DMC is applied to a 1030 MW coal fired power plant. [ABSTRACT FROM AUTHOR]

This paper presents a method of closed-loop identification for multivariable systems without external excitation. The method is specially designed for model predictive control (MPC) systems. Without using external excitation (test signals), the method ensures the informativity of the closed-loop data and, at the same time, improve the control performance during the test period. The purpose of the study is to reduce the cost of identification test. The basic idea is to switch the input weighting matrix in the MPC controller which leads to the informativity of the data-set. A preliminary test is carried out in order to find a new input weighting matrix which improve the control performance; then a switching scheme is developed based on the two weighting matrices. Traditional simulation based model validation no longer works in closed-loop identification without excitation, and model error bounds on the frequency responses can be used instead. The effectiveness of the proposed method is demonstrated by a simulation study. • A method of an MPC closed-loop identification without excitation is developed. • The method ensures the informativity and improves the control performance during the test. • The method uses input weighting switching to ensures the informativity of the data-set. • The simulation based model validation no longer works and model error bounds can be used instead. [ABSTRACT FROM AUTHOR]

In this paper, a model predictive control (MPC) case study is presented on the selective catalytic reduction (SCR) in a coal-fired power plant. The MPC is used for optimizing the volume of ammonia injection in order to meet the emission standard of nitrogen oxides (NO x ) as well as to prevent ammonia slip. A gain scheduling model is applied to deal with the nonlinearity of the SCR unit. The model gain is developed by kernel density estimation and the dynamic model is estimated by the asymptotic method of identification. The MPC control result shows that the standard deviation of outlet NO x concentration can be reduced by two-thirds, which leads to about 25% reduction of reagent consumption and the prevention of ammonia slip. [ABSTRACT FROM AUTHOR]

Purpose: The aim of this study was to explore the improved image quality of the portal vein using the contrast-enhancement boost (CE-boost) technique for the improved visibility of abdominal-enhanced computed tomography (CT) scans in clinical practice. Methods: This retrospective study included 50 patients in Group A who underwent routine abdominal-enhanced CT and 50 patients in Group B who underwent abdominal computed tomography angiography (CTA) with matched body mass index, age, and sex. Images in Group A were postprocessed with the CE-boost technique for further enhanced visibility of the portal vein. Both subjective and objective assessments of different branches of the portal vein in three types of images (i.e., Group A with CE-boost and without CE-boost, Group B) were statistically analyzed. Results: The subjective scores of two experienced radiologists showed good consistency (kappa value > 0.624, p < 0.001), and the score of Group A with CE-boost (mean, 4.64) was significantly higher than that of the others (p < 0.001). The liver parenchyma and most target veins in Group A with CE-boost showed the highest CT, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) values and the lowest standard deviation (SD), while the CNR of most portal veins in Group A without CE-boost had the lowest CNR (p < 0.001). There were no differences in the SNR of the portal vein in Group A without CE-Boost and Group B (p > 0.05). Conclusion: CE-boost can significantly improve image quality in portal vein imaging without any additional scanning settings or changes in the clinical workflow. [ABSTRACT FROM AUTHOR]

Abstract: In order to model complex industrial processes, this work studies the identification of linear parameter varying (LPV) models with two scheduling variables. The LPV model is parameterized as blended linear models, which is also called multi-model structure. Several weighting functions, linear, polynomial and Gaussian functions, are used and compared. The usefulness of the method is tested using a high purity distillation column model in a case study. The case study shows that a good fit of identification data is not enough to verify model quality and can even be misleading in nonlinear process identification; other measures related to process knowledge should be used in model validation. The case study also shows that commonly used LPV model based on parameter interpolation can fail for the high purity distillation column. Finally, several pitfalls in nonlinear process identification are pointed out. [Copyright &y& Elsevier]

Abstract: This work introduces a method of multivariable model error detection in model prediction control (MPC). The idea is to use non-disturbing small sinusoidal test signals to obtain accurate estimates of process frequency responses at several frequency points. Then, the differences between estimated frequency responses and the frequency responses of current MPC model are used to form the model error index matrix which is used to access the model error of the MPC controller. An upper error bound is developed for quantifying the error of frequency response estimation. The method works in closed-loop operation with the MPC controller online. Simulation studies are used to demonstrate the use of the method. [Copyright &y& Elsevier]

Industrial model predictive control (MPC) typically includes a steady-state target calculation (SSTC) layer that provides targets for the dynamic MPC controller to track. This paper develops the first closed-loop robust SSTC method, which explicitly addresses the closed-loop uncertainty, for stable linear systems with time-invariant parametric uncertainty. The problem is initially formulated as a stochastic bi-level optimization problem, where the lower-level optimization problem describes the nominal MPC law. The problem is then transformed into a deterministic mixed-integer second-order cone program for ellipsoidal uncertainty. Case study results show that the proposed method is more robust than the classical nominal SSTC in that it helps reduce cycling of the steady-state targets. In addition, the proposed method is less conservative than a typical open-loop robust SSTC method, in that it drives the system to the steady state more quickly and can push the steady state closer to the operating boundary when desired. • The first closed-loop robust steady-state target calculation method for MPC. • A bilevel stochastic program where the lower-level models steady-state control. • Reformulation of the bilevel stochastic program to a deterministic MISOCP. • The proposed method outperforms the nominal method and the open-loop robust method. [ABSTRACT FROM AUTHOR]

• A two-dimensional (2D) kernel-based Gaussian process regression identification method for batch process is proposed. • The 2D kernel is designed by means of addition / multiplication operation based on coordinate decomposition, and the properties of the proposed 2D kernel are analyzed. • The developed GPR-based identification method considers the time-varying noise, which is modeled as another zero-mean Gaussian process. In this work, a two-dimensional (2D) kernel-based Gaussian process regression (GPR) method for the identification of batch process is proposed. Under the GPR framework, the estimate of the time-varying impulse response is a realization from a zero-mean Gaussian process (GP), wherein the kernel function encodes the possible structural dependencies. However, the existing kernels designed for system identification are one-dimensional (1D) kernels and underutilize the 2D data information of batch process. Utilizing the 2D correlation property of batch process impulse response, we propose the amplitude modulated 2D locally stationary kernel by means of addition / multiplication operation based on coordinate decomposition. Then, a nonparametric identification method using 2D kernel-based GPR for batch process is developed. Furthermore, the properties of the proposed 2D kernel are analyzed. Finally, we demonstrate the effectiveness of the proposed identification method in two examples. [ABSTRACT FROM AUTHOR]

Abstract: This paper introduces the design and presents the research findings of the identification and control application for an industrial Circulation Fluidized Bed (CFB) boiler. Linear Parameter Varying (LPV) model is used in the model identification where steam flow is selected as the operation-point (scheduling) variable. Three kinds of weighting functions, namely linear, cubic splines and Gaussian functions are compared. LPV model based Model Predictive Control (MPC) is also simulated. Test results show that LPV model is more accurate than linear model, and LPV MPC yields a better control effect than linear MPC. [Copyright &y& Elsevier]

• Two-dimensional incremental recursive least squares identification methods for batch processes are proposed. • Latest estimation results along both time direction and batch direction is used to obtain the new estimated value. • The convergence property is established by the associated ordinary differential equation method. In this paper, two-dimensional incremental recursive least squares identification methods for batch processes are proposed. First, parameter increment along the time direction is introduced to represent the batch process, which changes slowly along the batch direction in comparison with the parameters. Then, we apply RLS method along the batch direction to derive two kinds of two-dimensional incremental recursive identification algorithm, which can use the latest estimation results along both the time direction and batch direction to obtain the new estimated value. By extending the model assumption from parameter batch-invariant to parameter increment batch-invariant, this method can effectively accelerate the convergence rate. Furthermore, the convergence property of the proposed method is established by the associated ordinary differential equation method. Finally, we demonstrate the superiority of the proposed identification method in two case studies. [ABSTRACT FROM AUTHOR]

In this paper, a cooperative distributed model predictive control (DMPC) algorithm based on topological hierarchy decomposition is proposed. Utilizing the connection topology information of the distributed system, we decompose subsystems into a hierarchy structure model through interpretative structural modeling. Subsystems with strong coupling are grouped into the same layer, and weak coupling exists between subsystems in different layers. Then, we propose an improved cooperative DMPC algorithm, in which the optimal input trajectories of subsystems in each layer are evaluated and propagated in a hierarchical order. Instead of all-to-all communication, only intra-layer communication is required at each iteration of DMPC optimization, which can significantly lessen the communication burden without losing much performance. Furthermore, the feasibility and stability of the proposed algorithm are proven in detail. Finally, the effectiveness and merits of the proposed method are demonstrated by applying it to a reheating furnace system and a six-area power system. • We present a cooperative DMPC based on topological hierarchy decomposition. • Subsystems are decomposed into a hierarchy structure model. • Less communication burden is needed to achieve a desired performance. • Ensure the feasibility and stability properties of the proposed algorithm. [ABSTRACT FROM AUTHOR]