Résumé: Holes, tunnels and cavities of two-dimensional (2D) and 3D objects are concise topological features used for object representation and recognition. In this study, the authors are representing any cubical tessellation (regular or not) of 2D and 3D objects and dealing with the extraction and the localisation of these features by using homology-based approach. The cubical tessellation (regular or not) of objects is translated into algebraic language suitable for building a reduced cell complex structure. The extraction of the homology information is equivalent to the estimation of the rank of the homology groups of the reduced complex. The localisation means the reconstruction of the object cycles from the generators of the homology groups. The reduction operation of the cell complex leads to an efficient algorithm. Note that, several objects can be analysed simultaneously by the algorithm conceived in our approach. This algorithm is validated by using 2D and 3D binary images.

Résumé: Artificial Neural Networks (ANNs) is a powerful computational tool to mimic the learning process of the mammalian brain. This book gives a comprehensive overview of ANNs including an introduction to the topic, classifications of single neurons and neural networks, model predictive control and a review of ANNs used in food processing. Also, examples of ANNs in food processing applications such as pasteurization control are illustrated.

Résumé: This book addresses the issue of Model Predictive Control (MPC), supported by a real life application of temperature control in plate heat exchangers, as it is noticed that classical control strategies, for instance PI and PID, often give a relatively large variance in temperature. This results in the producer having to set up a higher control temperature set point than the actual needed one, which introduces energy loss as well as affecting sometimes the treated product itself. One way of reducing temperature variance in order to be able to lower the temperature set-point without re-designing the plant, is to introduce better control. Model Predictive Control (MPC) is proposed as a control alternative. Such a control philosophy requires an internal model of the process, which is used to predict future process behavior. A survey of different MPC techniques is given, highlighting the two most popular linear MPC algorithms, for instance Predictive Functional Control (PFC) and Generalised Predictive Control (GPC) being evaluated for the case study.

Résumé: Cet ouvrage apporte des réponses au problème de l’alignement et de l'hétérogénéité des sources de données. Nous présentons un prototype d’alignement nomé XMap++, capable d'élaborer un mapping sémantique en tenant compte du contexte des sources à aligner. Ainsi, nous proposons des stratégies de matching qui peuvent être combinées de façon dynamique en prenant en considération les spécificités sémantiques des concepts. En particulier, le module aligneur supporte quatre modes de combinaison dynamique (e.g. moyenne, pondération dynamique, pondération avec les RNA, fonction sigmoïde) qui confère un choix parmi les mesures de similarité des entités (e.g., terminologique, linguistique et structurelle), et le degré de confiance approprié a chaque matcher pour aligner une entité donnée, considérant des caractéristiques sémantiques spécifiques (e.g. type de données, nature de contenu, etc.). Les algorithmes implémenté sont testés, sur des ontologies de grande taille. La phase de validation permet de comparer les résultats obtenus par les méthodes proposées avec les résultats d'autres méthodes pionnières de la littérature.

Résumé: The focus of this paper is put on developing Neural Networks approach to predict annual natural gas consumption in Algeria for the three pressure sectors (low pressure, medium pressure and high-pressure sector). Four main distribution areas constitutes the Algerian distribution company (SONALGAZ). Beside each distribution area consists of several distribution divisions (DD). Thus in this paper instead of creating a single neural network model with one dataset to estimate a sector consumption, each DD is considered on its own by selecting the most influential inputs, then developing its specific Multi Layer Perceptron (MLP) model trained with Levenberg-Marquardt learning algorithm, and finally summing their results to get the total consumption for the sectors.

Résumé: The rapid development of semantic Web and exponential growth in the use of the ontology in the field of smart cities, along with World Wide Web, make new and different multi-dimensional character of the smart cities a possibility, in which data is collected from various distributed systems. Consequently, in this paper, we exploit the concept of semantic Web for designing a new smart city ontology that is considered as a system of systems. Such ontology is beneficial for both the citizens and the administrators as it allows interoperability among different systems and frameworks

Résumé: A good communication and interaction between citizens and the administration is important and crucial, also can greatly help in improving the quality of urban life of citizen. In this paper, we propose a semantic data model for managing and resolving the problems that exist in cities such as water leak, street faults, broken street lights, and potholes. The main idea is to focus on the best practices of linked open data to describe all issues, and then integrate them in the dataset provided by DBpedia. Hence, our approach is based on the standards of the World Wide Web Consortium.