SESSION A8: Recent Advances in Decision Tools for Optimization

[165] The use of binary outranking relations to infer criteria weights for ELECTRE I method
Hela Frikha and Sawsan Charfi

Most of multi-criteria aggregation methods require fixing parameters in order to model decision maker’s preferences. These parameters’ values must be provided directly by the decision maker. However, this information is delicate and difficult to give. In addition, it is often partial and usually subjective since it is based only on his experience, his intuition and his psychological state. Subsequently, the decision-maker should be helped to express his preferences, to formalize his subjectivity and to elicit objective information. In this paper, we will develop an approach to infer ELECTRE I’s criteria weights using the information furnished by the decision-maker and taking support on mathematical programming.

[34] Dominance Measuring Methods within MAVT/MAUT with Imprecise Information concerning Decision-Makers’ Preferences
Antonio Jiménez-Martín, Alfonso Mateos and Pilar Sabio

Dominance measuring methods are an approach for dealing with complex decision-making problems with imprecise information within multi-attribute value/utility theory. These methods are based on the computation of pairwise dominance values and exploit the information in the dominance matrix in different ways to derive measures of dominance intensity and rank the alternatives under consideration. In this paper we review dominance measuring methods proposed in the literature for dealing with imprecise information (intervals, ordinal information or fuzzy numbers) about decision-makers’ preferences and their performance in comparison with other existing approaches, like SMAA and SMAA-II or Sarabando and Dias’ method.

[121] A Hybrid Approach Combining Fuzzy Consensus-Based Goal Programming and TOPSIS
Idris Igoulalene and Lyes Benyoucef

In hotly competitive international industrial and economic environments, supply chain coordination (SCC) is one of active research topics in production and operation management. In this research work, we present a new consensus-based fuzzy goal programming approach for supply chain coordination problem. It is formulated as a multi-criteria group decision making (MCGDM) problem and solved by combining consensusbased goal programming with TOPSIS method in a fuzzy environment. To demonstrate the applicability of the proposed approach, an example of robot selection problem is presented and the numerical results analyzed.

[139] Trace-Based Decision Making in Interactive Application: Case of Tamagotchi systems
Hoang-Nam Ho, Mourad Rabah, Samuel Nowakowski and Pascal Estraillier

We present our exploratory work for situation preselecting in interactive applications; assuming that the application is an Interactive Adaptive System based on a sequence of contextualized “situations”. Each situation confines activities and interactions related to a common context, resources and system actors. When one situation is completed, the system has to determine which is the best following one. We introduce in this paper a new preselecting method that selects possible situations among all available situations. We propose a method using Naïve Bayes to analyse the traces (the past of users). The result explains among all available situations, which one can be executed for the next step. Combining all obtained results, we get a set of situations, called set of alternatives that can be used in any decision algorithm. A case study of Tamagotchi game is presented to illustrate this new approach to optimize the preselecting in multicriteria environment.

SESSION B8:

[227] Optimization of a Reliable Network on Chip dedicated to partial reconguration
Camel Tanougast and Cedric Killian

We present an optimization of reliable Network on Chip (NoC) structure dedicated to dynamic recongurable systems (DRS) based on FPGA. The originality of our approach is based on a strate- gic placement of router incorporating elements of dependability. The solution is a factorization of these reliable routers encompassing routers without any error detection block. This ensures the global reliability of the network and reduce the cost of area, the latency of the data packets and the power consumption. The proposed approach can be applied to the majority NoC topologies.

[215] Multiple Closed Loop System Control with Digital PID Controller Using FPGA
Sirin Akkaya, Onur Akbati and Haluk Gorgun

PID (Proportional – Integral - Derivative) controllers are the most widely used closed loop controllers due to their simplicity, robustness, effectiveness and applicability for much kind of systems. With the rapid development of technology, implementation of PID controller has gone several steps from using analog components in hardware to using some software-based program to execute PID instructions digitally in some processor-based systems. And also, these developments have brought an alternative solution to implement PID instructions in Programmable Logic Devices (PLD). Field Programmable Logic Array (FPGA) is the most advanced members of PLDs. This paper presents the digital PID algorithm on FPGA. The controller algorithm is developed using VHDL and implemented using Altera DE0 Nano Board. As the controlled system, five axis robot arm is selected, which have five dc motor and four potentiometer to determine the positions of motors. The results show that digital PID controller and also multi-feedback control systems can be implemented successively using FPGA devices.

[228] Reliable Router for Dynamic Network on Chip
Mohamed Frihi, Mostefa Boutalbi, Camel Tanougast, Mikael Heil and Salah Toumi

This paper proposes a new reliable router allowing accurate online error detections in dynamic Network on Chip (NoC). The proposed router has the capability to detect and localize accurately inner or outer data packet errors of the router while distinguish between temporary and permanently errors. The error detection mechanisms of the proposed switches and advantages with regards to the other main already proposed router approaches are detailed while proving the feasibility and efficiency through several simulations online detection cases. Performance evaluation and FPGA implementation results are also given.

[90] Novel Experimental Synchronization Technique for Embedded Chaotic Communications
Said Sadoudi, Mohamed Salah Azza and Camel Tanougast

In this paper, we propose an interesting and original solution to the problem of chaotic synchronization which is focused in its high sensitivity to channel noise. We demonstrate through experimental design, the synchronization of tow embedded hyperchaotic Lorenz generators implemented separately in two FPGA circuits. The basic idea of the proposed solution consists of synchronizing two hyperchaotic systems without perturbing the hyperchaotic dynamic of the slave system. This means that, in an eventual wireless hyperchaotic communication system, whatever the perturbations suffered by the transmitted hyperchaotic carrier (drive signal) through the channel, our solution permits to generate an identical copy of the hyperchaotic carrier at the receiver and then recover the information signal correctly.

[80] Hardware Implementation for a New Design of the VBSME Used in H.264/AVC
Amira Yahi, Kamel Messaoudi, Salah Toumi and El Bey Bourennane

Motion estimation (ME) in video coding standard H.264/AVC adopts variable block size (VBSME) which provides high compression rates but requires much higher computation compared to the previous coding standards. To overcome this complexity, this paper describes a VHDL design and an implementation of VBSME. The design is based on partitioning each 16×16 macroblock into sixteen 4×4 non overlapping subblocks. The motion estimation of these subblocks is performed in parallel in order to use them to form the 41 subblocks of different sizes specified by the standard. As a result, this new design has in consideration low latency and high throughput with a maximum frequency which reaches over than 277 MHz on a Xilinx-Vittex5-LX110T FPGA.

SESSION C8: Advances in Sliding Mode Control Techniques: From Theory to Real Implementation

[116] New Algorithms of Adaptive Switching Gain for Sliding Mode Control: Part I – Ideal Case
Jiang Zhu and Karim Khayati

Based on recent results on adaptive sliding mode control (ASMC) design for nonlinear systems with uncertainties we propose some lemmas and theorems to discuss finite time convergence (FTC) and alternative approaches to smoothen the adaptation tuning algorithm of the ASMC design. These modifications are proposed to enhance accuracy without overestimation of the uncertainty magnitude and to suppress the chattering phenomenon. In fact, the new adaptation laws are designed in ways to assign a minimum admissible value to the switching gain. The robustness is proven using the Lyapunov stability criterion combined with an intuitive analysis of the control behavior. Simulation results are performed to demonstrate the effectiveness of the proposed algorithm. Part I introduces the new designs for ideal ASMC while the real case design will be shown in Part II.

[117] New Algorithms of Adaptive Switching Gain for Sliding Mode Control: Part II – Real Case
Jiang Zhu and Karim Khayati

Based on new results on ideal adaptive sliding mode control (ASMC) design for nonlinear systems with uncertainties discussed in Part I of the present contribution, we extend the new designs to the real case in this paper (Part II) by using boundary layer method and filtered rate of the sliding variable. These modifications are proposed to enhance accuracy without overestimation of the uncertainty magnitude and to suppress the chattering phenomenon, often magnified within real measured signals. In fact, the new adaptation laws are designed in ways to assign a minimum admissible value to the switching gain. The robustness is proven using the Lyapunov stability criterion combined with an intuitive analysis of the control behavior. Simulation results are performed to demonstrate the effectiveness of the proposed algorithm.

[159] Robust IM Control with Mras-Based Speed and Parameters Estimation with ANN using Exponential Reaching Law
Salah Eddine Rezgui, Said Legrioui, Adel Mehdi and Hocine Benalla

As known, the main cause of the degradation in indirect rotor field oriented induction motor (IM) control (IRFOC) is the time-varying machine parameters, especially the rotor-time constant (Tr) and stator resistance (Rs), more pertinently, in cases of proportional-integral control with speed observation. In this work, a new exponential reaching law (ERL) based sliding mode control (SMC) is introduced to improve significantly the performances when compared to the conventional SMC which are well known susceptible to the annoying chattering phenomenon, so, the elimination of the chattering is achieved while simplicity and high performance speed tracking are maintained. In addition, an artificial neural network (ANN) technique is used to achieve an accurate on-line conjoint estimation. This technique is integrated in the adaptation mechanism of the model reference adaptive system (MRAS) in order to obtain adaptive sensorless scheme. The merits of the proposed method are demonstrated experimentally through a test-rig realized via the dSPACE DS1104 card in various operating conditions.

[137] Graphical conditions for R-controllability of generalized linear switching systems
Mohamed Bendaoud, Hicham Hihi and Khalid Faitah

This paper investigates the structural R-controllability problem for generalized linear switching systems. Causal manipulations on the bond graph models are carried out in order to determine graphical conditions for structural R-controllability.

[112] Design of Full Order Observers with Unknown Inputs for Delayed Singular Systems with Constant Time Delay
Malek Khadhraoui, Montassar Ezzine, Hassani Messaoud and Mohamed Darouach

In this paper, new time and frequency domain designs of Unknown Inputs Full Order Observers (UIFOO) for a class of delayed singular systems arre proposed. A constant time delay present in both state and input vectors. The timedomain approach is based on Lyapunov-Krasovsii stability theory where the optimal gain implemented in the design is obtained in terms of linear matrix inequalities (LMIs). The frequency domain approach is obtained by using the factorization from time-domain results. The proposed approach procedure is justified by a numerical example.

SESSION D8: Identification and Control

[2] Statistical Assessment of New Coordinated Design of PSSs and SVC for Multimachine Power System
Sahar Eldeep and Ehab Ali

In this paper, the assessment of new coordinated design of Power System Stabilizers (PSSs) and Static Var Compensator (SVC) in a multimachine power system via statistical method is proposed. The coordinated design problem of PSSs and SVC over a wide range of loading conditions is handled as an optimization problem. The Bacterial Swarming Optimization (BSO), which synergistically couples the Bacterial Foraging (BF) with the Particle Swarm Optimization (PSO), is employed to seek for optimal controllers parameters. By minimizing the proposed objective function, in which the speed deviations between generators are involved; stability performance of the system is enhanced. To compare the capability of PSS and SVC, both are designed independently, and then in a coordinated manner. Simultaneous tuning of the BSO based coordinated controller gives robust damping performance over wide range of operating conditions and large disturbance in compare to optimized PSS controller based on BSO (BSOPSS) and optimized SVC controller based on BSO (BSOSVC). Moreover, a statistical T test is executed to validate the robustness of coordinated controller versus uncoordinated one.

[48] Comparison between asymptotic and non asymptotic identification techniques of a first-order plus time delay model from step response
Abdelbacet Mhamdi, Kaouther Ben Taarit and Moufida Ksouri

A comparison between asymptotic and non-asymptotic identification algorithms for a first order system is discussed. This paper describes three methods to identify a mathematical model for a real process with a time delay, namely: area, parametric and algebraic method. The process is the Process Trainer, PT326 from Feedback Instruments Limited. The main goal of this communication is to test the robustness and convergence of each method. The best results are obtained using the algebraic approach.

[50] H∞ decentralized dynamic-observer-based control for large-scale uncertain nonlinear systems
Nan Gao, Mohamed Darouach, Marouane Alma and Holger Voos

In this paper an H∞ decentralized observer-based control is proposed for large-scale uncertain nonlinear systems. These systems are coupled by N interconnected subsystems where the interconnections satisfy the quadratic constraint. The proposed control is based on a new form of dynamic observer (DO) and it generalizes the existing results on proportional ob- server (PO) and proportional integral observer (PIO). The design approach is derived from the solution of linear matrix inequalities (LMIs) and based on the algebraic constraints obtained from the analysis of the estimation error. A numerical example is provided to show the effectiveness of the proposed control.

[113] Design of Full Order H infinity Filter for Delayed Singular Systems with Unknown Input and Bounded Disturbance
Malek Khadhraoui, Montassar Ezzine, Hassani Messaoud and Mohamed Darouach

In this paper we propose a new method for the design of the full order H1 filter for a class of linear singular delayed systems with unknown inputs and bounded disturbances. The proposed procedure is based on the unbiasedness of the state estimation error and on the Lyapunov-Krasovskii stability theory, where the optimal filter gain implemented in the proposed design is a solution of linear matrix inequalities (LMIs). The proposed approach is tested on a numerical example and the results were successful.

[155] Robust PID-PSS Based Genetic Algorithms Implemented Under GUI - MATLAB
Ghouraf Djamel-Eddine and Naceri Abdellatif

Power System Stabilizer (PSS) is a supplementary control signal of a generator’s excitation system based on Automatic Voltage Regulator (AVR), are now routinely used in the industry to damp out power system oscillations. Optimal tuning gain of AVR - PSS is necessary for satisfactory performance of power system. Genetic algorithms (GA) have been widely used for global optimization problems. This paper presents a systematic approach for designing and optimal tuning an advanced Russian AVR-PSS gains, realized on PID-schemes (called AVR-SA), to improve the effectiveness and investigates its robustness under uncertainly constraints on a SMIB system, using Genetic Algorithms. The proposed approach employs GA search for optimal setting of AVR-PSS parameters. The performance of the proposed GA-PSS under small and large disturbances, loading conditions and system parameters variations are tested. The simulation results have proved that GA are powerful tools for optimizing the AVRPSS parameters, and obtained more robustness of the studied power system. This present work was performed and simulated using our graphical interface ‘GUI’ realized under MATLAB.

SESSION E8: Data Mining and Data/Information Analysis

[74] Linked Data, Data Mining and External Open Data for Better Prediction of at-risk Students
Farhana Sarker, Thanassis Tiropanis and Hugh C Davis

Research in student retention is traditionally survey-based, where researchers use questionnaires to collect student data to analyse and to develop student predictive model. The major issues with survey-based study are the potentially low response rates, time consuming and costly. Nevertheless, a large number of datasets that could inform the questions that students are explicitly asked in surveys is commonly available in the external open datasets. This paper describes a new student predictive model that uses commonly available external open data instead of traditional questionnaires/surveys to spot ‘at-risk’ students. Considering the promising behavior of neural networks led us to develop student predictive models to predict ‘at-risk’ students. The results of empirical study for undergraduate students in their first year of study shows that this model can perform as well as or even out-perform traditional survey-based ones. The prediction performance of this study was also compared with that of logistic regression approach. The results shows that neural network slightly improved the overall model accuracy however, according to the model sensitivity, it is suggested that logistic regression performs better for identifying ‘at-risk’ students in their programme of study.

[202] Forecasting Price Volatility Cluster of Commodity Futures Index by Using Standard Deviation with Dynamic Data Sampling Based on Significant Interval Mined from Historical Data
Kwan Hua Sim, Kwan Yong Sim and Patrick Hang Hui Then

Forecasting price volatility of financial time series has been a major challenge confronting investors, speculators, businesses and also governmental organization in view of its impacts, not only on financial aspect, but also social and possibly political aspects. While businesses have been struggling in making financial decision to hedge their risk against possible future price fluctuation, governmental bodies and policy makers often caught in the midst of severe volatility. This paper presents a standard deviation approach with dynamic data sampling to forecast the price volatility cluster of a commodity futures index in Malaysia derivative market. Data sampling to derive the mean of standard deviation is taken dynamically based on the last price reversal mined from the historical data. Experiment was conducted on historical price data for the period of twenty seven years to assess the competency of this standard deviation approach with mean values through dynamic data sampling in comparison to static mean values through fixed data sampling. The outcome of the experiment reveals a promising performance demonstrating the relevancy of the proposed approach. This study constitutes a novel approach using standard deviation to quantify price equilibrium, and subsequently forecasting possible future price volatility to allow better decision making with a more reliable analysis.

[56] A Study of the Data Pre-Processing Module of the Dendritic Cell Evolutionary Algorithm
Zeineb Chelly and Zied Elouedi

Data reduction as a critical step in the process of data pre-processing presents a central point of interest across a wide variety of fields. Data pre-processing has a significant impact on the performance of any machine learning algorithm. In this context, we focus our research paper on investigating the data preprocessing phase of a recent evolutionary algorithm named the Dendritic Cell Algorithm (DCA). We aim at reviewing the data pre-processing phase of the DCA while making a comparative study of the used data reduction techniques within the DCA. This is needed to clarify the differences, the advantages and the characteristics of the previously proposed techniques with the DCA. The output of the comparison will facilitate the task of the developer to select the most useful technique to be adopted and integrated in the DCA data pre-processing module.

[6] Context in Information Retrieval
Djalila Boughareb and Nadir Farah

The evolution of information retrieval is intimately linked to the evolution of the Web. Although that the involvement of different context dimensions in the improvement of the search task was greatly studied, the abundant development of hardware and software opens ways to explore new contextual dimensions. This paper presents the different contextual dimensions studies in the literature and proposes a new context taxonomy gathering all discussed dimensions.

[247] Distributed Real-time sentiment analysis of big data social stream
Amir Rahnama

Big data trend has enforced the data-centric systems to have continuous fast data streams. In recent years, real-time analytics on stream data has formed into a new research field, which aims to answer queries about “what-is-happening-now” with a negligible delay. The real challenge with real-time stream data processing is that it is impossible to store instances of data, and therefore online analytical algorithms are utilized. To perform real-time analytics, pre-processing of data should be performed in a way that only a short summary of stream is stored in main memory. In addition, due to high speed of arrival, average processing time for each instance of data should be in such a way that incoming instances are not lost without being captured. Lastly, the learner needs to provide high analytical accuracy measures. Sentinel[1] is a distributed system written in Java that aims to solve this challenge by enforcing both the processing and learning process to be done in distributed form. Sentinel is built on top of Apache Storm, a distributed computing platform. Sentinel’s learner, Vertical Hoeffding Tree, is a parallel decision tree-learning algorithm based on the VFDT, with ability of enabling parallel classification in distributed environments. Sentinel also uses SpaceSaving to keep a summary of the data stream and stores its summary in a synopsis data structure. Application of Sentinel on Twitter Public Stream API is shown and the results are discussed.