ABSTRACT Selection of relevant parameters from a high dimensional process operation setting space... more ABSTRACT Selection of relevant parameters from a high dimensional process operation setting space is a problem frequently encountered in industrial process modelling. This paper presents a method for selecting the most relevant fabric mechanical parameters for each sensory quality feature. The proposed relevancy criterion has been developed using two approaches. The first utilizes a fuzzy sensitivity criterion by exploiting from experimental data the relationship between mechanical parameters and all the sensory quality features for each evaluator. Next an OWA aggregation procedure is applied to aggregate the ranking lists provided by different evaluators. In the second approach, a panel of experts provides their ranking lists of mechanical features according to their professional knowledge. Also by applying OWA, the data sensitivity-based ranking list and the knowledge-based ranking list are combined to determine the final ranking list and the final relevant mechanical parameters for a given sensory quality feature.
2008 4th European Conference on Circuits and Systems for Communications, 2008
Several methods for the identification of FIR systems using cumulants have been proposed in the l... more Several methods for the identification of FIR systems using cumulants have been proposed in the literature. These methods can be classified into three categories of solutions: Linear Algebra, Closed form and Optimization. Only linear algebra solutions are considered in this paper. For the sake of simplicity, these methods use the least squares approach to solve a system of equations characterized by a redundant vector of unknown parameters and assumed to be linear, but it not. Mathematically, this approach is not suitable, since the obtained system is nonlinear and must be treated as an optimization problem. To overcome this problem, we define three optimization problems and based on that the best algorithm to solve it will be selected. Simulations are performed to demonstrate the performance of the proposed methods.
In this study, new approaches for the identification of finite impulse response (FIR) systems usi... more In this study, new approaches for the identification of finite impulse response (FIR) systems using higher-order statistics are proposed. The unknown model parameters are obtained using optimisation algorithms. In fact, the proposed method consists first in defining an optimisation problem and second in using an appropriate algorithm to resolve it. Moreover, a new method is developed for estimating the order of FIR models using only the output cumulants. The results presented in this study illustrate the performance of the proposed methods and compare them with a range of existing approaches.
Cet article présente des méthodes d'estimation paramétrique de systèmes linéaires à Réponse Impul... more Cet article présente des méthodes d'estimation paramétrique de systèmes linéaires à Réponse Impulsionnelle Finie (RIF) à l'aide de Statistiques d'Ordre Elevé (SOE). L'approche envisagée permet de traiter le problème d'identification de processus non gaussiens, à déphasage non minimal.
In this paper, new approaches for the identification of FIR systems using HOS are proposed. The u... more In this paper, new approaches for the identification of FIR systems using HOS are proposed. The unknown model parameters are obtained using optimization algorithms. In fact, the proposed method consists first in defining an optimization problem and second in using an appropriate algorithm to resolve it. Moreover, we develop a new method for estimating the order of FIR Models using only the output cumulants. The results presented in this paper illustrate the performance of our methods and compare them with a range of existing approaches.
2008 4th European Conference on Circuits and Systems for Communications, 2008
Several methods for the identification of FIR systems using cumulants have been proposed in the l... more Several methods for the identification of FIR systems using cumulants have been proposed in the literature. These methods can be classified into three categories of solutions: Linear Algebra, Closed form and Optimization. Only linear algebra solutions are considered in this paper. For the sake of simplicity, these methods use the least squares approach to solve a system of equations characterized by a redundant vector of unknown parameters and assumed to be linear, but it not. Mathematically, this approach is not suitable, since the obtained system is nonlinear and must be treated as an optimization problem. To overcome this problem, we define three optimization problems and based on that the best algorithm to solve it will be selected. Simulations are performed to demonstrate the performance of the proposed methods.
In this paper, new approaches for the identification of FIR systems using HOS are proposed. The u... more In this paper, new approaches for the identification of FIR systems using HOS are proposed. The unknown model parameters are obtained using optimization algorithms. In fact, the proposed method consists first in defining an optimization problem and second in using an appropriate algorithm to resolve it. Moreover, we develop a new method for estimating the order of FIR Models using only the output cumulants. The results presented in this paper illustrate the performance of our methods and compare them with a range of existing approaches.
Eighth International Multi-Conference on Systems, Signals & Devices, 2011
In this paper, we apply an adaptive control algorithm to a nonlinear multivariable process. Such ... more In this paper, we apply an adaptive control algorithm to a nonlinear multivariable process. Such controller is based on the multiple models approach. As the design of the control law requires the knowledge of the dynamical model of the system, we deal firstly with the identification of the system parameters using the recursive least squares and the retro propagation of the gradient algorithms. Then, we focus on the application of the multiple model approach. So, we decomposed the nonlinear model of the system in sub-systems and we adopted a proper criterion of commutation between the various models. The global control consists in the interpolation between the elementary control extracted from each model. The resulting controller is applied to a multivariable process to solve a tracking problem of the water levels into a twin tank process. The control strategy ensures the stability of the closed loop system and guarantees a good behavior when tracking a reference trajectory.
2010 7th International Multi- Conference on Systems, Signals and Devices, 2010
This work deals with the synthesis of an output feedback predictive controller for an induction m... more This work deals with the synthesis of an output feedback predictive controller for an induction motor. The theoretical study showed the possibility of solving the control law design problem by taking into account input constraints. Furthermore, the unavailable state variables of the system are recovered by the incorporation of an observer. Two important titles in the design feature are worth to be emphasized. The first one consists in identification of electrical parameters by the immersion technique. The second consists in using the high gain observer (HGO) to perform on-line an accurate estimation of the mechanical speed, load torque and the rotor fluxes. The global scheme involves a design of a sensorless output feedback predictive controller. Simulation results for an induction motor are addressed to show the effectiveness of the control design method.
2007 International Symposium on Computational Intelligence and Intelligent Informatics, 2007
In this paper, Nonlinear Generalized Predictive Control environment it often happens that measure... more In this paper, Nonlinear Generalized Predictive Control environment it often happens that measurement of the states (NGPC), proposed by Chen [4] in continuous time, is variables is very costly or even impossible. To overcome this reformulated into a quadratic optimization problem in order to problem, observer design is then necessary. And it's shown as take into account constraints, and this approach is applied to a an essential component in control applications. In the linear particular class of nonlinear systems. The states which are case, the well known separation principle allows the problem assumed available will be estimated in this work by a high gain to be split into two sub-problems which can be solved observer (HGO). The closed loop dynamics, under combination * . . of predictive controller and observer synthesis, is shown the design of state feedback ontre and transparent to the designers, and merely manipulated in the .t igosat e obserer ie the solution of the LQG sense that output feedback controller explicitly depend on choice optmiationproblem cnis the omation o the sa of only two parameters (prediction time and high gain feedback controller which iS the optimal solution to the LQ parameter). For a nonlinear example of considered class, we problem and the separately designed optimal Kalman filter. show satisfactory performances and robustness in presence of Although, ingeneral certainty equivalence principle cannot be disturbance applied for nonlinear systems. Recently, significant results in the nonlinear separation principle which originates from [2] I. INTRODUCTION are achieved and then some restricted classes of nonlinear systems admit a global separation principle between controller In the two last decades, model predictive control (MPG) has and state estimator design. For bilinear systems, a separation received a great deal of attention, and is shown by many to be principle has been stated in [I1], and further generalized in one of the most concrete methods in control engineering [9] [12]. In [3], systems which are uniformly observable are Therefore, it has widely accepted by the process industry. A studied, and the main results that, under a lipschitz condition good overview of industrial linear MPi can be found in [18]. of nonlinearity, a high gain observer is suggested which is
This paper proposes the optimization of parameters of neuro-fuzzy system using the particle swarm... more This paper proposes the optimization of parameters of neuro-fuzzy system using the particle swarm optimization. Neuro-fuzzy techniques have emerged from the fusion of neural networks and fuzzy inference systems. They could serve as a powerful tool for system modeling and control. These fuzzy systems are optimized by adapting the antecedent and consequent parameters. Among them, the ANFIS use the least square to optimize the consequent parameters and retropropagation to train the antecedent parameters. Several learning algorithms of fuzzy models have been proposed, e.g. evolutionary algorithms, such as particle swarm optimization. These different methods have been developed to learn the parameters of neuro-fuzzy system and to test them in the on-line control of nonlinear system. I.
21st Mediterranean Conference on Control and Automation, 2013
This paper presents the design of adaptive observers for a class of MIMO nonlinear switched syste... more This paper presents the design of adaptive observers for a class of MIMO nonlinear switched system which is composed on cascade subsystems. It consists on simultaneous estimation of the whole state as well as the unknown parameters.
High gain output feedback control of a quadruple tank process
MELECON 2008 - The 14th IEEE Mediterranean Electrotechnical Conference, 2008
Abstract In this paper, we apply a robust observer-based control structure for a MIMO nonlinear m... more Abstract In this paper, we apply a robust observer-based control structure for a MIMO nonlinear model of a quadruple tank process. Our main goal here is to track a reference trajectory of the water level in the two lower tanks using the information available from the ...
2009 6th International Multi-Conference on Systems, Signals and Devices, 2009
The incorporation of an observer into a state feedback controller results into the so-called outp... more The incorporation of an observer into a state feedback controller results into the so-called output feedback controller. Due to its easiness of design and implementation, many control strategies were reformulated under output feedback. This paper deals with the control problem of a MIMO process represented by a quadruple tank process through the use of a recently proposed output feedback controller. Such one is based on the use of a high gain observer which we propose to substitute with a sliding mode observer. Numerical simulations are developed to show the effectiveness of the proposed observer when applied to estimate the water levels into two bottom tanks. Into the control law is incorporated a filtered integral action. The effectiveness of such component is showed when we evaluate the robustness of the whole process against step like disturbances and stochastic noises.
Uploads
Papers by Faouzi M'SAHLI