Adel Abdennour

An intelligent supervisory system for drum type boilers during severe disturbances

Severe disturbances cause a number of problems in the operation of power plants. The plant subsystem that often suffers the most is the boiler. Alleviating the severity of the problem requires improving the boilers capability in dealing with disturbances. One of the more severe upsets is partial or full load rejections. Withstanding load rejections depends almost exclusively on the boiler control. This paper presents a fuzzy supervisory scheme to improve the performance of the boiler response during such disturbances. This scheme consists of a robust local controller designed using the Linear Quadratic Gaussian with Loop Transfer Recovery (LQG/LTR). The set-points to the controller are modified whenever necessary by a supervisor partially based on fuzzy logic.

Evaluation of neural network architectures for MPEG-4 video traffic prediction

Multimedia applications and particularly MPEG-coded video streams are becoming major traffic components in high speed networks. Traffic prediction is important in enhancing the reliable operation over these networks. However, MPEG video traffic exhibits periodic correlation structure and a complex bit rate distribution, making prediction difficult. Neural networks can effectively be used to overcome such problem. In the literature, the problem has been mostly evaluated using standard feed-forward neural networks. However, a significant improvement can be expected using different types of neural networks. In this paper, six separate neural network predictors (including feed-forward) that can predict the basic frame types of MPEG-4: I, P, and B are developed and evaluated using long entertainment and broadcast video sequences. The performance is also compared to the widely used linear predictor. Comparison with results published in a recent work is also presented.

Particle Swarm Algorithm variants for the Quadratic Assignment Problems - A probabilistic learning approach

A new probability-based approach is proposed for the learning in discrete PSO.A generic framework is proposed to discretize PSO and its variants.Five well-known PSO variants are discretized based on the proposed framework.Comparative evaluation and search landscapes analysis is presented. The Quadratic Assignment Problem (QAP) has attracted considerable research efforts due to its importance for a number of real life problems, in addition to its acknowledged difficulty. Almost all of the well-known nature-mimicking algorithms have been applied to solve the QAP. However, the Particle Swarm Optimization (PSO), which has proven to be very effective in various applications, has received little attention at this front. The reason can be ascribed to the Euclidian-distance based learning concept (at the core of the algorithm) which makes PSO, in its present form, unsuitable for combinatorial optimization problems. In this article, a new probability-based approach is proposed for the learning in PSO. Based on this learning concept, a generic framework is developed to discretize PSO and its variants, to make them suitable for combinatorial optimization. Five well-known PSO variants are discretized based on this proposed framework. A comparative study of all discretized PSO variants is also included. Moreover, the proposed framework is compared to other attempts to discretize PSO, in addition to three other meta-heuristic approaches. The comparison revealed that the proposed technique is more effective.

A Real-Time Intelligent Wireless Mobile Station Location Estimator with Application to TETRA Network

Mobile location estimation has received considerable interest over the past few years due to its great potential in different applications such as logistics, patrol, and fleet management. Many mobile location estimation techniques had been proposed to improve the accuracy of location estimation. Location estimation based on artificial intelligence techniques is a recent alternative approach. In this paper, adaptive neuro-fuzzy inference system (ANFIS) is used as a robust location estimator to locate the mobile station (MS) using the MS geo-fencing area data within 9 km from a serving base station. Extensive evaluations and comparisons have been performed, and a set of statistical parameters has been obtained. From the comparison of the proposed ANFIS estimator with the neural-network-based estimators, it is found that ANFIS estimator is faster and more robust. Its average computation time (ACT) is 0.076 sec. While the ACT for multilayer perceptron (MLP) and radial-based function (RBF) neural networks is 0.88 and 1.7, respectively. Whereas on comparing ANFIS with other techniques, it is found that in ANFIS estimator, 67 percent of the estimated location errors do not exceed 149 m, while these for the statistical, multiple linear regression, and geometric are 170, 280, and 2,346 m, respectively. Thus, the results clearly reveal that the proposed ANFIS estimator outperforms all other techniques.

A team-oriented approach to particle swarms

The Particle Swarm Optimization (PSO) is a simple, yet very effective, population-based search algorithm. However, degradation of the population diversity in the late stages of the search, or stagnation, is the PSOs major drawback. Most of the related recent research efforts are concentrated on alleviating this drawback. The direct solution to this problem is to introduce modifications which increase exploration; however it is difficult to maintain the balance of exploration and exploitation of the search process with this approach. In this paper we propose the decoupling of exploration and exploitation using a team-oriented search. In the proposed algorithm, the swarm is divided into two independent teams or sub swarms; each dedicated to a particular aspect of search. A simple but effective local search method is proposed for exploitation and an improvised PSO structure is used for exploration. The validation is conducted using a wide variety of benchmark functions which include shifted and rotated versions of popular test functions along with recently proposed composite functions and up to 1000 dimensions. The results show that the proposed algorithm provides higher quality solution with faster convergence and increased robustness compared to most of the recently modified or hybrid algorithms based on PSO. In terms of algorithm complexity, TOSO is slightly more complex than PSO but much less complex than CLPSO. For very high dimensions (D>400), however, TOSO is the least complex compared to both PSO and CLPSO.

Decentralized media access vs. credit-based centralized bandwidth allocation for LR-PONs

Several centralized dynamic bandwidth allocation algorithms have been proposed for Ethernet Passive Optical Networks (EPONs), making the optical line terminal (OLT), located in the central office, the intelligent device that arbitrates time-division access to the shared upstream channel. When the network span is extended to 100 km and beyond, as suggested by next generation long-reach PONs (LR-PONs), the increased propagation delays severely degrade the performance of these algorithms since they are based on bandwidth negotiation messages frequently exchanged between the optical network units (ONUs) and the OLT. In this paper, we propose a decentralized media access scheme for the emerging LR-PON that would enable sooner transmission of upstream packets. We accompany this scheme with centralized control over the network. Through simulation, we compare the performance of the proposed scheme with centralized credit-based schemes that are thought to be able to mitigate the degradation of centralized schemes in LR-PONs. Results show that the proposed scheme is less affected by the network extension and can reduce average packet delays beyond those of credit-based schemes.

VBR video traffic modeling and synthetic data generation using GA-optimized Volterra filters

Variable bit rate (VBR) video traffic models, which accurately represent the traffic characteristics and statistical properties of real videos, can provide significant information about expected traffic behavior. This knowledge can be used in the development of effective control schemes and improved network quality of service. An interesting class of models based on the idea of generating a number of chi-square sequences, by passing a Gaussian autoregressive (AR) process through a simple nonlinearity, was recently introduced. The gamma process in this class of models is obtained by a linear combination of the chi-square sequences. This model is simple and allows for arbitrary selection of both the AR model and the shape parameter of the gamma probability density function. However, the AR filter order is chosen mainly on a trial-and-error basis. In addition, while the approach uses a linear combination of K chi-square sequences, it fixes (K - 2) coefficients and solves for only the remaining two because it has more equations than unknowns. Occasionally, the resulting solution is not feasible and additional trials for different solutions are required. It is therefore the objective of this paper to use genetic algorithms to provide a more systematic approach to find the various model parameters. The paper also presents a thorough statistical analysis of the generated synthetic data in order to assess its suitability for representing MPEG video traffic. A comparison with published results is carried out in terms of how close are the means, standard deviations, and the autocorrelation functions to those of the real data. A comparison of over 10000 replications and a number of different video traces reveals that a significant improvement can be achieved in almost all measures and for almost all the movies tested.

Energy Resolution of CsI (TL) Crystal Coupled with PIN Photodiode@@@قدرة تحليل الطاقة للبلورة CsI TL المقرونة بالشرائح الضوئية من نوع PIN

The emission spectrum of the CsI (Tl) scintillation crystal does not match well with the spectrum sensitivity of the standard pho- to-multiplier. This is the main handicap limiting the use of this crystal for radiation detection. However, the production of the new photodi- odes, with better characteristics, is likely to be a good candidate as a readout device for this crystal. One way to investigate this possibility is to study the energy resolution of the CsI (Tl) crystal coupled with a new large area PIN photodiode. The purpose of this paper is to deter- mine the different contributions affecting this resolution including that of the readout device.