Le travail de recherche effectué dans le cadre de cette thèse concerne le développement d’un système de perception de la saillance en environnement 3D en tirant l’avantage d’une représentation pseudo-3D. Notre contribution et concept issue de celle-ci part de l'hypothèse que la profondeur de l’objet par rapport au robot est un facteur important dans la détection de la saillance. Sur ce principe, un système de vision saillante de l’environnement 3D a été proposé, conçu et validée sur une plateforme comprenant un robot équipé d’un capteur pseudo-3D. La mise en œuvre du concept précité et sa conception ont été d’abord validés sur le système de vision pseudo-3D KINECT. Puis dans une deuxième étape, le concept et les algorithmes mis aux points ont été étendus à la plateforme précitée. Les principales contributions de la présente thèse peuvent être résumées de la manière suivante : A) Un état de l'art sur les différents capteurs d'acquisition de l’information de la profondeur ainsi que les différentes méthodes de la détection de la saillance 2D et pseudo 3D. B) Etude d’un système basé sur la saillance visuelle pseudo 3D réalisée grâce au développement d’un algorithme robuste permettant la détection d'objets saillants dans l’environnement 3D. C) réalisation d’un système d’estimation de la profondeur en centimètres pour le robot Pepper. D) La mise en œuvre des concepts et des méthodes proposés sur la plateforme précitée. Les études et les validations expérimentales réalisées ont notamment confirmé que les approches proposées permettent d’accroitre l’autonomie des robots dans un environnement 3D réel / The research work, carried out within the framework of this thesis, concerns the development of a system of perception and saliency detection in 3D environment taking advantage from a pseudo-3D representation. Our contribution and the issued concept derive from the hypothesis that the depth of the object with respect to the robot is an important factor in the detection of the saliency. On this basis, a salient vision system of the 3D environment has been proposed, designed and validated on a platform including a robot equipped with a pseudo-3D sensor. The implementation of the aforementioned concept and its design were first validated on the pseudo-3D KINECT vision system. Then, in a second step, the concept and the algorithms have been extended to the aforementioned robotic platform. The main contributions of the present thesis can be summarized as follow: A) A state of the art on the various sensors for acquiring depth information as well as different methods of detecting 2D salience and pseudo 3D. B) Study of pseudo-3D visual saliency system based on benefiting from the development of a robust algorithm allowing the detection of salient objects. C) Implementation of a depth estimation system in centimeters for the Pepper robot. D) Implementation of the concepts and methods proposed on the aforementioned platform. The carried out studies and the experimental validations confirmed that the proposed approaches allow to increase the autonomy of the robots in a real 3D environment

Baigiamajame magistro darbe nagrinėjamos neraiškiųjų verslo taisyklių naudojimo informacinėse sistemose problemos, „minkštųjų skaičiavimų“ intelektinėse informacinėse sistemose problematika, neuroninių-neraiškiųjų sistemų principai. Išnagrinėti pagrindiniai neraiškiosios logikos dėsniai, kuriais remiantis naudojamos neraiskiosios verslo taisyklės intelektinėse informacinėse sistemose. Pateiktas būdas, kaip neuroninės-neraiškiosios sistemos gali būti naudojamos verslo taisyklių sistemose naudojant RuleML, taisyklių žymėjimo kalbos, standartą. Baigiamajame darbe aprašomas eksperimentas, atliktas naudojant Matlab aplinką, XMLBeans taikomąją programą ir autoriaus sukurta neraiškaus išvedimo sistemos perkelimo į RuleML formatą taikomąją programą. Išnagrinėjus teorinius ir praktinius neuroninių-neraiškiųjų sistemų naudojimo aspektus, pateikiamos baigiamojo darbo išvados ir siūlymai. Darbą sudaro 5 dalys: įvadas, analitinė-metodinė dalis, eksperimentinė-tiriamoji dalis, išvados ir siūlymai, literatūros sąrašas. Darbo apimtis – 58 p. teksto be priedų, 30 iliustr., 30 bibliografiniai šaltiniai. Atskirai pridedami darbo priedai. / This work investigates the problems of use of fuzzy business rules in information systems, „soft computing“ in intelligent information systems issues, neuro-fuzzy systems principles. Main fuzzy logic laws are considered, which are used as the basis of fuzzy business rules in intelligent information systems. Suggested an approach, based on RuleML standard, how neuro-fuzzy systems could be used together with business rules engines. This paper describes the experiment carried out using the Matlab environment, XMLBeans application and the author created application for fuzzy inference system migration to RuleML standard format. Structure: introduction, analysis , project, conclusions and suggestions, references. Thesis consist of: 58 p. text without appendixes, 30 pictures, 30 bibliographical entries. Appendixes included.

In 2004, students at the University of Guelph designed and constructed an all-wheel-drive Formula SAE vehicle for competition. It utilized an electronically-controlled, hydraulic-actuated limited slip center coupling from Haldex Traction Ltd, to transfer torque to the front wheels. The initial control system design was not comprehensively conceived, so there was a need for a thoroughly developed control system for the all-wheel-drive actuator augmented with commonly available sensors and a low cost controller. This thesis presents a novel all-wheel-drive active torque transfer controller using a neural network estimated slip angle velocity. This controller specifically targets a racing vehicle by allowing rapid direction changes for maneuverability but damping slip angle changes for increased controllability. The slip angle velocity estimate was able to track the actual simulated value it was trained against with excellent phase matching but with some offsets and phantom spikes. Using the estimated slip angle velocity for control realized smooth control output, excellent stability, and a fast turn-in yaw response on par with rear-wheel-drive configurations. A full vehicle simulation with software-in-the-loop testing for control software was also developed to aid the system design process and avoid vehicle run time for tuning. This design flow should significantly decrease development time for controls algorithm work and help increase innovation within the team.

The hybridization of soft computing methods of Radial Basis Function (RBF) neural networks, Multi Layer Perceptron (MLP) neural networks with back-propagation learning, fuzzy sets and rough sets are studied in the scope of this thesis. Conventional MLP, conventional RBF, fuzzy MLP, fuzzy RBF, rough fuzzy MLP, and rough fuzzy RBF networks are compared. In the fuzzy neural networks implemented in this thesis, the input data and the desired outputs are given fuzzy membership values as the fuzzy properties &ldquo / low&rdquo / , &ldquo / medium&rdquo / and &ldquo / high&rdquo / . In the rough fuzzy MLP, initial weights and near optimal number of hidden nodes are estimated using rough dependency rules. A rough fuzzy RBF structure similar to the rough fuzzy MLP is proposed. The rough fuzzy RBF was inspected whether dependencies like the ones in rough fuzzy MLP can be concluded.

Numerical optimization of multimodal or otherwise nontrivial functions has stayed around the peak of the interest of many researchers for a long time. One of the promising methods that appeared is the hybrid approach of the Learnable Evolution Model that combines the well-established ways of artificial intelligence and machine learning with recently popular and efective methods of evolutionary programming. In this work, the method itself was reviewed with respect to what has been already implemented and tested and several possible new implementations of the method were proposed and some of them consequently implemented. The resulting program was then tested against a set of chosen nontrivial real-valued functions and its results were compared to those achieved with EDA algorithms.

Master theses deals with Constructive Neural newtorks. First part describes neural networks and coresponding mathematical models. Furher, it shows basic algorithms for learning neural networks and desribes basic constructive algotithms and their modifications. The second part deals with implementation details of selected algorithms and provides their comparision. Further comparision with backpropagation algorithm is provided.

This dissertation thesis deals with the application of expert systems and soft computing methods in field of power oil transformers. The main work is divided into theoretical and practical part. First, the theoretical part presents the basic elements of the transformer, and approaches to its diagnosis. The work focused mainly on the diagnostics of the insulation system, and diagnostic methods and approaches in this specific area. Next part describes the basics of expert systems and other soft computing methods such as: fuzzy logic, neural networks, genetic algorithms and their combinations and extensions. At the end of the theoretical part, the possibility of optimization approaches by means of artificial intelligence and its application in fuzzy model optimization are described. The practical part begins with description of the used data file that runs through the entire work. The work is then divided into four parts, namely in parts which deal with the expert system for transformer diagnostics, DGA module, prediction module, and optimization using artificial intelligence. The section describing the expert system gives specific information about the particular expert system. The means and techniques used for constructing given system are described, and then the complete system design and description of all subsystems and modules are presented. The next section describes the developed DGA module and all selected approaches to its implementation and expansion. At the end of the chapter, the results of comparison between all implemented methods are evaluated. The third part deals with the prediction module and describes its design and construction, including description of the main parts which are based on the selected predictive approaches. Also, the predictions of selected quantities from the data file are included. There are two predictive approaches being used: the one step prediction, and the multiple step prediction. The comparison of prediction accuracy and computational cost of given methods is presented at the end of this chapter. The last part deals with the possibilities of optimization using artificial intelligence methods, namely differential evolution, PSO, and genetic algorithms. Both the single-objective and the multi-objective optimization are considered. The methods are compared in a series of synthetic tests and then applied to optimize the fuzzy models of DGA tests from an earlier part of this work. The dissertation also includes chapters: "The Aims", "The Contribution of the Work", and a list of publications, products, and projects of the author.

The military recently recognized the importance of taking sociocultural factors into consideration. Therefore, Human Social Culture Behavior (HSCB) modeling has been getting much attention in current and future operational requirements to successfully understand the effects of social and cultural factors on human behavior. There are different kinds of modeling approaches to the data that are being used in this field and so far none of them has been widely accepted. HSCB modeling needs the capability to represent complex, ill-defined, and imprecise concepts, and soft computing modeling can deal with these concepts. There is currently no study on the use of any computational methodology for representing the relationship between adverse events and infrastructure development investments in an active war theater. This study investigates the relationship between adverse events and infrastructure development projects in an active war theater using soft computing techniques including fuzzy inference systems (FIS), artificial neural networks (ANNs), and adaptive neuro-fuzzy inference systems (ANFIS) that directly benefits from their accuracy in prediction applications. Fourteen developmental and economic improvement project types were selected based on allocated budget values and a number of projects at different time periods, urban and rural population density, and total adverse event numbers at previous month selected as independent variables. A total of four outputs reflecting the adverse events in terms of the number of people killed, wounded, hijacked, and total number of adverse events has been estimated. For each model, the data was grouped for training and testing as follows: years between 2004 and 2009 (for training purpose) and year 2010 (for testing). Ninety-six different models were developed and investigated for Afghanistan iv and the country was divided into seven regions for analysis purposes. Performance of each model was investigated and compared to all other models with the calculated mean absolute error (MAE) values and the prediction accuracy within ±1 error range (difference between actual and predicted value). Furthermore, sensitivity analysis was performed to determine the effects of input values on dependent variables and to rank the top ten input parameters in order of importance. According to the the results obtained, it was concluded that the ANNs, FIS, and ANFIS are useful modeling techniques for predicting the number of adverse events based on historical development or economic projects’ data. When the model accuracy was calculated based on the MAE for each of the models, the ANN had better predictive accuracy than FIS and ANFIS models in general as demonstrated by experimental results. The percentages of prediction accuracy with values found within ±1 error range around 90%. The sensitivity analysis results show that the importance of economic development projects varies based on the regions, population density, and occurrence of adverse events in Afghanistan. For the purpose of allocating resources and development of regions, the results can be summarized by examining the relationship between adverse events and infrastructure development in an active war theater; emphasis was on predicting the occurrence of events and assessing the potential impact of regional infrastructure development efforts on reducing number of such events.

The developments of advanced non-linear control strategies have attracted a considerable research interests over the past decades especially in process control. Rather than an absolute reliance on mathematical models of process plants which often brings discrepancies especially owing to design errors and equipment degradation, non-linear models are however required because they provide improved prediction capabilities but they are very difficult to derive. In addition, the derivation of the global optimal solution gets more difficult especially when multivariable and non-linear systems are involved. Hence, this research investigates soft computing techniques for the implementation of a novel real time constrained non-linear model predictive controller (NMPC). The time-frequency localisation characteristics of wavelet neural network (WNN) were utilised for the non-linear models design using system identification approach from experimental data and improve upon the conventional artificial neural network (ANN) which is prone to low convergence rate and the difficulties in locating the global minimum point during training process. Salient features of particle swarm optimisation and a genetic algorithm (GA) were combined to optimise the network weights. Real time optimisation occurring at every sampling instant is achieved using a GA to deliver results both in simulations and real time implementation on coupled tank systems with further extension to a complex quadruple tank process in simulations. The results show the superiority of the novel WNN-NMPC approach in terms of the average controller energy and mean squared error over the conventional ANN-NMPC strategies and PID control strategy for both SISO and MIMO systems.

Layout Decision Analysis and Design is a ubiquitous problem in a variety of work domains that is important from both strategic and operational perspectives. It is largely a complex, vague, difficult, and ill-structured problem that requires intelligent and sophisticated decision analysis and design support. <br /><br /> Inadequate information availability, combinatorial complexity, subjective and uncertain preferences, and cognitive biases of decision makers often hamper the procurement of a superior layout configuration. Consequently, it is desirable to develop an intelligent decision support system for layout design that could deal with such challenging issues by providing efficient and effective means of generating, analyzing, enumerating, ranking, and manipulating superior alternative layouts. <br ><br /> We present a research framework and a functional prototype for an interactive Intelligent System for Decision Support and Expert Analysis in Multi-Attribute Layout Optimization (IDEAL) based on soft computing tools. A fundamental issue in layout design is efficient production of superior alternatives through the incorporation of subjective and uncertain design preferences. Consequently, we have developed an efficient and Intelligent Layout Design Generator (ILG) using a generic two-dimensional bin-packing formulation that utilizes multiple preference weights furnished by a fuzzy Preference Inferencing Agent (PIA). The sub-cognitive, intuitive, multi-facet, and dynamic nature of design preferences indicates that an automated Preference Discovery Agent (PDA) could be an important component of such a system. A user-friendly, interactive, and effective User Interface is deemed critical for the success of the system. The effectiveness of the proposed solution paradigm and the implemented prototype is demonstrated through examples and cases. <br /><br /> This research framework and prototype contribute to the field of layout decision analysis and design by enabling explicit representation of experts? knowledge, formal modeling of fuzzy user preferences, and swift generation and manipulation of superior layout alternatives. Such efforts are expected to afford efficient procurement of superior outcomes and to facilitate cognitive, ergonomic, and economic efficiency of layout designers as well as future research in related areas. <br /><br /> Applications of this research are broad ranging including facilities layout design, VLSI circuit layout design, newspaper layout design, cutting and packing, adaptive user interfaces, dynamic memory allocation, multi-processor scheduling, metacomputing, etc.