Introduction Part I: Preliminaries Linear Algebra and Preliminaries Disctrete-time Linear Systems Stochastic Processes Kalman Filter Part II: Realization Theory Realization of Deterministic Problems Stochastic Realization Theory I Stochastic Realization Theory II Part III: Subspace Identification Subspace Identification I: ORT Subspace Identification II: CCA Identification of Closed-loop System Appendices Least-squares Method Input Signals for System Identification Overlapping Parametrization Matlab(R) Programs Solutions to Problems

Subspace methods for system identification

This paper presents an efficient method for identification of nonlinear Multi-Input Multi-Output (MIMO) systems in the presence of colored noises. The method studies the multivariable nonlinear Hammerstein and Wiener models, in which, the nonlinear memory-less block is approximated based on arbitrary vector-based basis functions. The linear time-invariant (LTI) block is modeled by an autoregressive moving average with exogenous (ARMAX) model which can effectively describe the moving average noises as well as the autoregressive and the exogenous dynamics. According to the multivariable nature of the system, a pseudo-linear-in-the-parameter model is obtained which includes two different kinds of unknown parameters, a vector and a matrix. Therefore, the standard least squares algorithm cannot be applied directly. To overcome this problem, a Hierarchical Least Squares Iterative (HLSI) algorithm is used to simultaneously estimate the vector and the matrix of unknown parameters as well as the noises. The efficiency of the proposed identification approaches are investigated through three nonlinear MIMO case studies.

Identification of multivariable nonlinear systems in the presence of colored noises using iterative hierarchical least squares algorithm.

Modeling, simulation and advanced control of methanol production from variable synthesis gas feed

This paper develops a model for optimal and online tuning of PID controllers and evaluates its performance. The proposed model is based on computational intelligence approaches and can be used in industrial plant operating processes. The proposed tuning model is called PID-neuro-fuzzy model, a formulation based on a structured artificial neural network and fuzzy rules. The neural network is used to perform an optimal adjustment of PID controller gains to ensure the operating point of the system required to reduce the time of accommodation and the steady-state error. A fuzzy system is incorporated into a real-time gain-scheduling scheme to compensate for the possible variations in plant parameters. The performance of the proposed model is evaluated based on its ability to deal with uncertainties and disturbances in the process. The efficiency of the model is investigated through computational simulations in five plants of two industrial processes in the mining sector: 1) solid bulk unloading processes by car dumper and 2) solid bulk resumption process by bucket wheel reclaimers. In conclusion, the main advantage of the proposed model is its adaptability in relation to variations in plant parameters during the operational process.

A Neuro-Fuzzy Model for Online Optimal Tuning of PID Controllers in Industrial System Applications to the Mining Sector

This paper discusses the empirical modeling using system identification technique with a focus on an interacting series process. The study is carried out experimentally using a gaseous pilot plant as the process, in which the dynamic of such a plant exhibits the typical dynamic of an interacting series process. Three practical approaches are investigated and their performances are evaluated. The models developed are also examined in real-time implementation of a linear model predictive control. The selected model is able to reproduce the main dynamic characteristics of the plant in open-loop and produces zero steady-state errors in closed-loop control system. Several issues concerning the identification process and the construction of a MIMO state space model for a series interacting process are deliberated.

MIMO model of an interacting series process for Robust MPC via System Identification.

This paper presents the empirical modeling of the gaseous pilot plant which is a kind of interacting series process with presence of nonlinearities. In this study, the discrete-time identification approach based on subspace method with N4SID algorithm is applied to construct the state space model around a given operating point, by probing the system in open-loop with variation of input signals. Three practical approaches are used and their performances are compared to obtain the most suitable approach for modeling of such a system. The models are also tested in the real-time implementation of a linear model predictive control. The selected model is able to well reproduce the main dynamic characteristics of gaseous pilot plant in open loop and produces zero steady-state errors in closed loop control system. Several issues concerning the identification process and the construction of MIMO state space model are discussed.

/pdf/system-identification-of-an-interacting-series-process-for-4ammclhxf2.pdf

System identification of an interacting series process for real-time model predictive control

This paper demonstrates the use of the external feedback of manipulated variables in model predictive controller to produce a bumpless transfer of actuator when the controller is switched from one mode to another. The work is carried out experimentally using a lab-scale pilot plant, be named the gaseous pilot plant. The controller is built in MATLAB/Simulink environment which able to communicate and control the plant in real-time condition. It is shown that the use of the external feedback of manipulated variables provides a better estimation and gives effective outcomes to make bumpless transfer transitions of the actuators.

Tri Chandra S. Wibowo

Papers

System identification of an interacting series process for real-time model predictive control

MIMO model of an interacting series process for Robust MPC via System Identification.

Steady-state bumpless transfer MPC using external feedback of Manipulated Variables