Applied Computational Intelligence

FOREWORD.

CONTENTS.

No abstract received.

No abstract received.

No abstract received.

Hand gesture is a natural and efficient communication channel to humans. We first describe important basic notions of hand gesture, and then, show several real world application problems of hand gesture, in which it is effectively used for human-friendly interaction/interface and various soft computing techniques are adopted in order to handle the gesture signals with uncertainty and inconsistency.

Higher-order neural units have good computational properties in information processing and control applications. This paper presents higher-order neural units for visual image processing. We use the higher-order neural units for edge detection and employ the Hough transform to process the edge detection results. The higher-order neural unit based edge detection method has been applied to solve routing problems for mobile robots. Simulation results show that the proposed higher-order neural units are efficient for image processing and routing applications of mobile robots.

The main objective of this paper is to point out the key relevance of the classification issue in mathematical modeling. In particular, it is stressed that standard logical structures are quite simple classification structures, where the allowed degrees of truth or falsehood are connected according to a linear ordering. But when viewed as a classification problem, much more complex logical structures appear in a natural way, showing in fact that the standard linear assumption was kind of artificial. Moreover, once we focus our attention on the classification problem we realize that a relevant portion of classical research devotes to decision making, therefore imposing some misleading restrictions: the true scientific issue is much more related to classification rather than to decision making.

No abstract received.

No abstract received.

This paper summarizes the major research progress in the logical algebras – lattice implication algebras (LIAs), algebraic logics – lattice-valued logics based on LIAs, as well as uncertainty inference and automated reasoning based on these lattice-valued logics. The relevant further research topics and prospects are also outlined. The theory and method of these kinds of logic systems for characterizing the lin guistic truth-value-based uncertainty inference and automated reasoning are highlighted and will be specially investigated.

Lukasiewicz and others created, in a pure conceptual domain, the many valued logic that we can describe, in a symbolic way, by lattice structures. Zadeh, with the definition of the continuous membership function, created Fuzzy Sets for vagueness, the operations of which are different from the usual set theory as intersection, union and complement. L-fuzzy sets are built when the membership function assumes symbolic values L. Meta theory of uncertainty based upon modal logic is built to unify measures of vagueness, uncertainty and inconsistence in natural language.

This paper shows the connection between the Meta theory of uncertainty based upon modal logic and the implication algebra.

The paper mainly attempts to define the → so as to make (L, ∨, ∧, O, I) be a quasi-lattice implication algebra. The method is as follows: Firstly, we decide some values according to the basic attributes; while there are still some values we can't decide, we clarify them according to its conditions. Secondly, according to these conditions, we list all the formulas in a set to prove that the values satisfy the conditions. For those formulas without variables, it is easy to verify; while for those with variables, we have to decide the value for those variables. Hence, the main task of this paper is to find a method to decide those values. (here we use the theory of rough set to solve the problem). At Last, an specific example is given to further illustrate this method.

In this paper, the relationship between the lattice implication algebras and the Łukasiewicz logic systems is discussed. The main results are as follows: if (L(n), ≤) is a chain of n elements with a least element θ and a greatest element I, then there exists a unique lattice implication algebra (L(n), ∨, ∧, →, ¬, θ, I) with the derived order ≤, and it is isomorphic to the n-valued Łukasiewicz logic system. Moreover, the necessary and sufficient condition that a finite lattice implication algebra is a chain in relation to the derived order is given.

Brouwerian lattice and lattice implication algebra both are important generalizations of Boolean algebras. In this paper, we investigate the relations between them, and verify that we can construct a lattice H implication algebras by closed elements of any Brouwerian lattice with 0.

In this paper, we show that bilattices are robust mathematical structures that provide a natural accommodation to, and bridge between, intuitionistic fuzzy sets and interval-valued fuzzy sets. In this way, we resolve the controversy surrounding the formal equivalence of these two models, and open up the path for a new tradition for representing positive and negative information in fuzzy set theory.

The purpose of this paper is to introduce the notions of fuzzy prime LI-ideals and fuzzy maximal LI-ideals in lattice implication algebras. Some properties of fuzzy ILI-ideals are investigated. Nine characterizations of fuzzy prime LI-ideals in lattice implication algebras and eight characterizations of fuzzy prime LI-ideals in lattice H implication algebras are obtained. The extension properties of fuzzy prime LI-ideals is given. The relations among various fuzzy LI-ideals and LI-ideals are established, respectively. We introduce the notion of fuzzy zero LI-ideals of lattice implication algebras. Finally we describe some classes of lattice implication algebras via fuzzy ILI-ideals (resp. fuzzy prime LI-ideals, fuzzy maximal LI-ideals).

This paper shows that the set of all LI-filters of a lattice implication algebra does a Brouwerian lattice, a pseudo-complemented distributive lattice and a Dedikind lattice, respectively. And the set of all involutory LI-filters of a lattice implication algebra does a Boolean algebra. Some algebraic properties of fuzzy LI-filters in lattice implication algebras are investigated. The level LI-filter characterizations of fuzzy LI-filters in lattice implication algebras are obtained. These generalize corresponding results in Ref. 11.

The article discusses the properties of implication operator "→" in lattice implication algebra, and gives the method to set up lattice implication algebra with an order-reversing involution be defined implication operator, finally studies the union operation, intersection operation, inverse operation and composition operation of implication operator "→" in lattice implication algebra.

In this paper, we discuss the fuzzy reasoning with L-valued level reliability, which based on the L_vpl, transferred to modal fuzzy reasoning with truth level and get some results which could be used to simulate the fuzzy reasoning of human being.

We know that it is concerned with the language circumstances that a proposition is true or not. In this paper, a model about time is given. And under this model, the concepts of ever, ever always be, will and will always be are introduced, then some problems related to resolution principle on lattice-valued tense propositional logic are studied, including (α, t) — soundness of resolution principle and completeness of weak delete strategy. Finally, a resolution algorithm on lattice-valued tense propositional logic is proposed.

The aim of this paper is to discuss algebra structure of a kind of linguistic terms. These linguistic terms consist of the basic elements set (C) and hedges set (H). In some cases, by considering natural linguistic ordering and T = H × C, the set of linguistic terms (T) is a lattice. Employing operators "′" and " → ", T becomes a lattice implication algebra. Hence, linguistic terms can be handled based on lattice-valued propositional logic system LP(X).

We propose a pathway to deal with the logical problem in analogy, and indicate that the constraints on analogy mapping can be studied in special kind of first order logic. The logical problem in analogy is about the using of substitution rules, but a valid analogy inference can satisfy more flexible mapping than mappings that substitute rule restrict. The logical validity of semantic constraints will be explained with expressing the common part meaning of concepts with formulae sets, and in most cases semantic similarity will violate the constraints on logical valid analogy mapping.

In the present paper, automated reasoning based on lattice-valued propositional logic LP(X) where the clause only contains 0-IESFs or 1-IESFs as its generalized literals will be focused. A method of automated reasoning is given, and the soundness and completeness theorems of this method are also proved.

We introduce a novel Mamdani-type fuzzy model, referred to as the all-permutations fuzzy rule-base (APFRB). We show that any standard feedforward neural network is mathematically equivalent to an appropriate APFRB. This implies that we can easily extract the knowledge embedded in a trained neural network in the form of fuzzy rules.

Artificial Neural Networks (ANN), especially Multilayer Perceptrons (MLP), are successfully applied for the modelling of non-linear relations between input and output variables. Usually the ANN is interpreted as Black-Box-Model. Characteristic values of the training error or test error (error characteristics) serve as criterion for evaluation of ANN quality. Starting point for the investigation, described in this paper is the fact that the information about the global process behavior is stored in the trained weight topology of the MLP. It will be demonstrated, that a correlation exists between the process behavior to be reproduced and the adjusted weight characteristics of the MLP. To analyze the weight characteristics a correlation weight is defined and compared with a statistical value. The methodology will be demonstrated on an example.

Many important theories and mechanisms of large-scale biological systems can satisfy future requirements to Internet. In this paper, we design a novel bio-network architecture and its simulation platform with abilities of service emergence and evolution. The design and implements of its components are discussed in details. We can realize network emergent service and application in the bio-network platform.

This paper presents the concept of fuzzy structured element, and provides the theorem of structured element representation of fuzzy number and fuzzy-valued function. The operations of fuzzy numbers and the calculus of fuzzy-valued functions are easier than it was with the new representation theorem.

With three indices embodying the general characteristics of fuzzy numbers: the expectations of fuzzy numbers, the middle points and the spreads of α–cut sets, a new approach to rank fuzzy numbers through the weighted average method was proposed, and an absolute order relation on the basis of the real number set was generated simultaneously. The ranking result can not only reflect the preference of the decision-maker, but also be computed conveniently, especially in ranking triangular and trapezoidal fuzzy numbers. Numerical experiments indicated that this approach can overcome some pitfalls rooted in most existing ranking methods to some extent.

Inclusion is a quasi-order relation that plays an important role in rough set theory. We present a formula to compute the inclusion degree of two rough sets in an approximation space, and we study its properties. Relationships between this inclusion measure for rough sets and a refinement order relation are discussed.

The paper presents a semantic model of linguistic truth value. The model is built based on two hypotheses. Firstly, the number of linguistic truth variable (abbr. LTV) is finite; Secondly, the election of LTV is fuzzy and random, when a man wants to express something. The original model is the example in paper [3], as the membership function of the original LTV, and the derivative LTV is the solution of using the operations to original LTV. Based on the two hypotheses and the later works, we widen the curve of the membership function into string and use finite number of strings to cover the area [0,1]×[0,1].

A new approach of evaluating and decision based on analysis and learning of attribute coordinate in nuclear emergency is presented in this paper.

Causal reasoning plays an essential role in both informal and formal human decision-making. Causality itself as well as human understanding of causality is imprecise, sometimes necessarily so. A common sense understanding of the world tells us that we have to deal with imprecision, uncertainty and imperfect knowledge. A difficulty is striking a good balance between precise formalism and commonsense imprecise reality. An algorithmic method of accommodating imprecision in causality is needed. Today, data mining holds the promise of extracting unsuspected information from very large databases. However, the most common data mining rule forms do not express a causal relationship. Without understanding the underlying causality, a naïve use of data mining rules can lead to undesirable actions.

In this paper, we give a general treatment for some kind of sequences such as customer sequences, document sequences, and DNA sequences, etc. Large collections of transaction, document, and genomic information have been accumulated in recent years, and embedded latently in it there is potentially significant knowledge for exploitation in the retailing industry, in information retrieval, in medicine and in the pharmaceutical industry, respectively. The approach taken here to the distillation of such knowledge is to detect strings in sequences which appear frequently.

In recent years, extensive research has been carried out by using association rules to build more accurate classifiers. The idea behind these integrated approaches is to focus on a limited subset of association rules, i.e. those rules where the consequence of the rule is restricted to the classification class attribute. This paper aims to contribute to this integrated framework by adapting the CBA (Classification Based on Associations) algorithm. More specifically, CBA was modified by coupling it with a new measurement of the quality of association rules: i.e. intensity of implication. By means of this measurement, the sequence in which the class association rules are chosen, was changed when building the classifier. The new algorithm has been implemented and empirically tested on 16 popular datasets from the UCI Machine Learning Repository. Furthermore, the results were validated with original CBA, with C4.5 (both on original and on discretized datasets), and with Naïve Bayes. The adapted CBA algorithm presented in this paper, proved to generate a lowest average error rate and produced classifiers that are more compact than original CBA.

In this paper, a matrix-based approach for mining a special kind of boolean association rules is presented and discussed through an example from the viewpoint of uncertainty reasoning.

The Entity-Relationship model (ER) along with its extensions (EER) is a powerful means for business and data modeling at a conceptual level. However, the models need to keep extended or enriched with new semantics as the real world changes and the understanding of the real world improves. This paper focuses on enriching the ER model with new knowledge discovered from databases. A particular type of knowledge of concern is association rule (AR). In bridging the paradigms of AR and ER, a framework of semantic correspondence is proposed. In doing so, the notion of specialization is extended so as to reflect sub-groupings of entities and relationships in terms of E-specialization and R-specialization, respectively.

Knowledge discovery in database (KDD) is the process of discovering previously unknown and potentially interesting patterns in large databases. This process consists of many interacting steps performing specific data manipulation and transformation operations. Although several process models are already available, they cannot fully satisfy the need of real applications such as information security. In this paper, experiences of the knowledge discovery process are characterized and a new model by means of an extension of the model by Fayyad et al. (1996) is formalized. Two case studies are illustrated why the process model is proposed and in what situation it can be used in practice.

This paper describes the rules for translation of an intuitionistic fuzzy SQL (IFSQL) statement to a standard SQL statement. The IFSQL is used with an intuitionistic fuzzy relational database (IFRDB) model, which covers the need for storing and querying imprecise information. The model uses the terms and operations defined in the intuitionistic fuzzy sets theory in order to describe the relational algebra model with intuitionistic fuzzy terms.

In the past decade, various clustering algorithms based on the behaviour of real ants were proposed. The main advantage of these algorithms lies in the fact that no additional information, such as an initial partitioning of the data or the number of clusters, is needed. In this paper we show how the combination of the ant-based approach with fuzzy rules leads to an algorithm which is conceptually simpler, more efficient and more robust than previous approaches.

This paper proposes an enhanced RBF network that enhances learning algorithms between input layer and middle layer and between middle layer and output layer individually for improving the efficiency of learning. The proposed network applies ART2 network as the learning structure between input layer and middle layer. And the auto-tuning method of learning rate and momentum is proposed and applied to learning between middle layer and output layer, which arbitrates learning rate and momentum dynamically by using the fuzzy control system for the arbitration of the connected weight between middle layer and output layer. The experiment for the classification of number patterns extracted from the citizen registration card shows that compared with conventional neural networks and the ART2-based RBF network, the proposed method achieves the improvement of performance in terms of learning speed and convergence.

Spiking neural networks represent a more plausible model of real biological neurons where time is considered as an important feature for information representation and processing in the human brain. In this paper, we apply spiking neural networks with dynamic synapses for pattern recognition in multidimensional data. The neurons are based on the integrate and-fire model, and are connected using a biologically plausible model of dynamic synapses. Unlike the conventional synapse employed in artificial neural networks, which is considered as a static entity with a fixed weight, the dynamic synapse (weightless synapse) efficacy changes upon the arrival of input spikes, and depends on the temporal structure of the impinging spike train. The training of the free parameters of the spiking network is performed using an evolutionary strategy (ES) where real values are used to encode the dynamic synapse parameters, which underlie the learning process. The results show that spiking neurons with dynamic synapses are capable of pattern recognition by means of spatio-temporal encoding.

This paper addresses the problem of modeling heterogeneous time series through a community of neural net-based autoregressive models which evolves based on the dynamicity detects from system's observed behavior. The system assigns a NN-based autoregressive model to each time window which are defined based upon the estimated values of noise variance. So, the system can learn more whenever new data point is present. In other words the VC-dimension of the learning machine is usually enough to learn the whole series.

In this paper, we propose and evaluate a novel recognition algorithm for container identifiers that effectively deals with various distortions and recognizes identifiers from container images captured in various environments. The proposed algorithm, first, extracts the area containing only the identifiers from container images using CANNY masking and bi-directional histogram method. The extracted identifier area is binarized by the proposed fuzzy binarization method. Then a contour tracking method is applied to the binarized area in order to extract the container identifiers, which are the target for recognition. In this paper we also propose and apply an enhanced ART1-based RBF network for recognition of container identifiers. The results of experiment for performance evaluation on the real container images showed that the proposed algorithm performs better for extraction and recognition of container identifiers compared to conventional algorithms.

This paper presents an algorithm for edge detection in images based on fuzzy sets theory. In order to perform edge detection we exploit the fuzziness present at edge locations in an image. The proposed parametric scheme is based on the notion of the a-order entropy of a fuzzy set. The parameter a controls the sensitivity of the algorithm to detect various types of edges. A detailed comparison with the algorithm proposed in [1] is carried out.

This paper demonstrates the use of some methods of signal analysis, performed on ECG and in some cases blood pressure signals, for the identification of the heart health status. Spectral analysis, continuous wavelet transform and fractal analysis were tested. The analysis was made on data from mice and rats. A correlation was found between the health status of the mice and the rats and some of the statistical descriptors, most notably the phase of the cross-spectra between ECG and blood pressure, and the fractal properties and dimensions of the interbeat series (RR-interval fluctuations).

A non-intrusive method of two-phase flow identification is investigated in this paper. It is based on image processing of data obtained from dynamic neutron radiography recordings. Classification of the flow regime types is performed by an artificial neural network (ANN) algorithm. The input data to the ANN are some statistical functions (mean and variance) of the wavelet transform coefficients of the pixel intensity data. The investigations show that bubbly and annular flows can be identified with a high confidence, but slug and churn-turbulent flows are more often mixed up in between themselves.

To evaluate the sperm quality of fertile men, the sperm morphology was examined based on a fuzzy data fusion. All sperms were observed in infertile patients. The abnormal spermatozoa of the fertile men were classified into different types: head, neck, and tail. Detailed study of each sperm track was possible as the image sequence with the individual sperm movement traces for easy inspection. The criteria defining the types of the classification could be set in order to track the details of fuzzy fusion analysis. The morphology was performed as a complete in the evaluation. The numbers of the normal and abnormal sperm were identified, and the type of defect was recorded for each abnormal spermatozoon. We showed the results in this paper that fuzzy fusion morphology provided a unified and the consistent framework to express different shapes of spermatozoa in fertile human.

In this paper, an approach that utilises fuzzy logic is presented in order to develop linguistic summaries of some patterns in remote-sensed images. Techniques such as clustering and genetic algorithms are used to mine images and optimise the linguistic summary to the most suitable one for each pattern/object in the database.

This paper proposes a software for image retrieval using linguistic expressions of colors. The retrieval requires a classification of images according to dominant colors. Pixels of the images are considered in HLS space and membership to classes are then computed. The classes correspond to the linguistic expressions proposed in the software: classes for hues and classes for colors qualifiers. These data are stored in a database to facilitate the user queries.

In this paper we present a pixel coloring algorithm, to be considered as a tool in fuzzy classification. Such an algorithm is based upon a sequential application of a divisive binary procedure on a fuzzy graph associated to the image to be classified, taking into account surrounding pixels. Each color will suggest a possible class, if homogeneous, and the hierarchical structure of colors will allow gradation between classes.

Before implementing an engineering system in the design process are studied different proposals to evaluate and rank them. In this evaluation process several experts assess different aspects and criteria according to their knowledge and preference on them. These criteria may have different nature (quantitative, qualitative) and the experts could belong to different areas and have different knowledge on each criterium, so the assessments used to express the value on each criterium could be assessed with different types of information (numerical, linguistic, interval-valued). In such a case, to select the best proposal we must deal with this heterogeneous information to evaluate and rank the different proposals. In this contribution we show different fuzzy approaches for dealing with heterogeneous information.

A framework for modelling the safety of an engineering system using a fuzzy rule-based evidential reasoning (FURBER) approach has been proposed recently, where a fuzzy rule-base designed on the basis of a belief structure (called a belief rule expression matrix) forms a basis in the inference mechanism of FURBER. In this paper, a learning method for optimally training the elements of the belief rule expression matrix and other knowledge representation parameters in FURBER is proposed. This process is formulated as a nonlinear objective function to minimize the differences between the output of a belief rule base and given data. The optimization problem is solved using the optimization tool provided in MATLAB. A numerical example is provided to demonstrate how the method can be implemented.

Under the circumstance that opinions of experts of group decision making are assumed as L – R fuzzy numbers, the membership functions are moved toward the left according to the sizes of their expectations. The similarity degree of two experts' opinions is calculated for an alternative where their expectations equal zero and the mean distance of two members' opinions is represented as their opinion differences. Using simple additive weighting (SAW) method, a new aggregation for fuzzy opinions under group decision making is proposed in this paper. For each expert, the degree of similarity between his and the others is averaged. The weight of specific expert's opinion in aggregation is proportional to the degree of average agreement of that expert.

A conception of the left and right projection which one fuzzy vector project on another is introduced in this paper. The purpose is to propose a new method to select an optimal alternative in fuzzy multi-attribute decision making environment. Using the conception of the left and right projection, the difference of each alternative with the fuzzy ideal solution or negative ideal solution are projected on the fuzzy weight vector. The size of the combination projection coefficient that combines the left and right projection coefficients is used as a judgement standard to measure each alternative. The decision making criterion is that the more the combination projection coefficient, the more superior the alternative.

A conception of the left and right projection which one fuzzy vector projects on another is introduced in this paper. Using the conception of the left and right projection, each alternative is projected on the fuzzy weight vector. The size of the combination projection coefficient that combines the left projection coefficient with the right one is used as a judgement standard to measure each alternative.

The appraisement of enterprise human resource is an important aspect in the modern human resource management. It is very important to meet the requirement of information society and to strengthen an organization cohesive affinity by improving the rationality of human resource appraisement. The appraisement system of human resource should have certain level rationality, which has important meaning for each organization. This paper puts forward the model of working process of the appraisement system of human resource, and more, analyzes the infection of appraisement object, appraisement quantity table, appraisement corpus, appraisement result synthetic method, appraisement result application and feedback, appraisement policymaker and environment factors etc.

It is necessary that the appraisement system of human resource is provided with the enough reasonability. An organization should strengthen its key competitive ability in information society, therefore, need to strengthen its cohesive affinity by various effective means, the reasonable appraisement system of human resource is a kind of important means to strengthen organization cohesive affinity. This paper puts forward a model of evaluation of appraisement system of human resource.

For linear bilevel programming problems, the branch and bound algorithm is the most successful algorithm to deal with the complementary constraints arising from Kuhn-Tucker conditions. This paper proposes a new branch and bound algorithm for linear bilevel programming problems. Based on this result, a web-based bilevel decision support system is developed.

Many business decisions can be modeled as multiple objective linear programming (MOLP) problems. When formulating a MOLP problem, objective functions and constraints involve many parameters which possible values are assigned by the experts who are often imprecisely or ambiguously known. So, it would be more appropriate for these parameters to be represented by fuzzy numbers. In this paper, a new fuzzy goal approximate algorithm is developed for solving fuzzy multiple objective linear programming (FMOLP) problems with fuzzy parameters when the fuzzy goals for the objective functions need to be achieved. And an illuminative example is also given to illustrate the algorithm developed.

Support Vector Machine (SVM) is a new machine learning method, and Least Squares Support Vector Machine is an SVM version that involves equality instead of inequality constraints, and works with a least squares cost function. Both of these two methods learn the decision surface from two distinct classes of input points, and each point is assigned to one of these two classes. But in many applications, a few points are not sure assigned to one of these two classes. In this paper, we apply a fuzzy degree based on membership function to each input point and reformulate the LS-SVM such that different input points can have different contributions to the learning of the decision function.

A new approach to quality of training online evaluation in virtual reality worlds is proposed. This approach uses evolving fuzzy neural networks (EFuNN) for modeling and classification of trainee in pre-defined classes of training.

The PageRank algorithm takes advantage of the link structure of the Web to improve the ranking of query results, independent of any particular users. Obviously, the PageRank score reflects a "democratic" importance that has no preference for any particular pages. In fact, users always browse those Web pages in which they have more interest. So, we propose a new personalized ranking vector–UserRank (UR) based on user behaviors of browsing pages to estimate the importance of Web pages. In order to yield more accurate query results, we take the UR vector combined with PageRank vector to generate a new ranking vector–Personalized PageRank (P²R). Experimental results demonstrate that our P²R algorithm can more closely reflect user preference and more largely improve the precision of ranking results.

Recent research on the crawler algorithm becomes a hot spot in the field of search engine. By analyzing the crawler algorithm and its programs, an intelligent spider's algorithm has three characteristics: (1) intelligently guarded by the user interests; (2) a much smaller user search space than the space of Internet; (3) online with the keywords. An implementation of an intelligent spider's search engine is presented in this paper.

In this paper, a model for semantic of linguistic information on lattice implication algebra is presented and discussed by taking a decision making problem for example. In this model, semantic of linguistic information is treated as a triple and the processing of it is combined by formal inference.

Fuzzy control algorithms have been extensively used in business and industry. In this paper, a fuzzy adaptive control technique based on L-R fuzzy number is presented and its application in the pension systems is discussed.

Based on the multiplex information-flow, Intelligentized Split-Spread and Intelligentized Renew-Compound of Multiplex Information-Flow in Working-Flow Management are put forward, the essential idea is that the multiplex information-flow is disassembled, spread, reverted, and replicated. The following problems are solved: repeatedly spreading overlapping data, spreading unamended data.

Validation is one of the most important steps in developing a reliable simulation model. It evaluates whether or not the model forms an accurate enough representation of the system of interest for the goals of the modelling study. Most of the methods that are currently available for validation are statistical techniques, and are as such binary in nature, in the sense that a model will turn out to be either fully valid or fully invalid. In this paper, we develop a new, fuzzy set theoretic method for validation that is continuous in nature, in the sense that it allows to express degrees of model validity. We experiment with our method on a number of simulation models for a real-life airline network.

In this paper, we introduce the method of information diffusion to evaluate the coefficient of loess collapsibility. Based on 68 sets of collapsing loess data from geo-technical tests in the Loess Plateau, China, in which moisture content, pore ratio and compression coefficient are chosen as major effecting factors. Using the fuzzy information diffusion method, first and second fuzzy approximation reasoning and information integration methods we establish the fuzzy relationship matrix between every evaluating elements and the collapse coefficient. The model is verified by comparing the calculation with the experimental result of another ten data sets of collapsing loess: the method described as simple, fast, practical and able to incorporate more factors to assess loess collapsibility objectively. It can be applied in the disaster prevention planning for inhabited areas, organization of urban geological maps in loess areas, and the regional analysis of loess collapsibility.

There is no believable risk map because of the tremendous imprecision of natural disaster risk assessment. To improve the representation of risk information spatially in cartographic form, an alternative mode of risk maps is required. A fuzzy risk represented by a possibility-probability-distribution, which is calculated by employing the interior-outer-set model as an information diffusion technique, can represent the imprecision. In this paper, we introduce fuzzy risk into the domain of risk maps and research into the appropriate way of transforming numeric fuzzy risks into risk maps. A mode of soft risk map is defined which is more representative for risk information with an example of soft risk map with respect to strong earthquakes in Yunnan province. A comparison between traditional risk map and soft risk map is presented.

In this paper, we discuss the benefit of the soft risk map calculated by the interior-outer-set model. We suppose that a company will invest three projects in a district consisting of four zones with different flood risks. The company chooses zones based on the risks beside water resource and cost. The result shows that the soft risk map is better than the traditional risk map.

In this paper, we suggest a risk assessment model of water shortage based on information diffusion theory and use it to analyze the bearing capacity of water resources in Jinhua City. From the application we know that the carrying capacity of water resources is the maximum support capability on human activities under a certain stage of social development and sound circulation in ecosystem. For Jinhua City, we discover that the water shortage is not the type of lacking of water sources, but the water quality and water conservancy projects. Therefore, only the middle scheme of giving attentions to both economic development and environmental protection is the first choice.

This paper presents an outline of a proposed risk assessment system for evaluating expected damage of structures and consequent financial losses and casualties due to a likely earthquake. Uncertainties caused by insufficient knowledge about the correlation of parameters has been considered by fuzzy relations, while uncertainties in eliciting the likelihood of earthquake magnitude from scarce past events, has been expressed by fuzzy probability instead of conventional probability. The approach represents fuzzy probability by constructing a possibility distribution over probability out of a small data set and a sample case calculation is presented.

Conditional event algebra put forward by Goodman has been advocated as a mathematical framework that ensures coherence for defining logical implication in probabilistic setting and allowing logical deductions. On the other hand, Bamber proposed a nice logical setting for dealing with assertion of high conditional probability. This paper attempts to review the basic reliability models of serial and parallel system in the light of Goodman's conditional event algebra. Especially, sufficient rules together with the underlying assumptions are investigated in order to provide a full characterization of serial / parallel systems.

This paper describes the function of a noise detection and suppression filter that uses fuzzy logic techniques and aims both at improved performance and reduced computational complexity. It is applicable to color images and outperforms other similar reported techniques in terms of quality indexes and image detail preservation. The filter specializes in the detection and elimination of random and fixed valued impulse noise. The improved performance of the algorithm is confirmed by experimental results.

This paper presents a new fuzzy motion adaptive video deinterlacer that is adaptive at pixel and frame level. It is mainly based on a decomposition of the corresponding fuzzy motion detector into three main modules: a linear spatio-temporal low-pass filter described by two separable 1D FIR filters and two fuzzy modules described by linear saturation functions. Moreover, the involved saturation parameters are on-line adjusted taking into account the motion quantity of each frame. Experimental results with several video benchmarks demonstrate the robustness and high-quality reconstruction of the presented algorithm with relatively low computational load.

The paper presents a color enhancement method based on a transform defined in the framework of a bounded logarithmic model. The enhancement transform uses three fuzzy surfaces associated with the following image parameters: brightness, contrast and saturation.

The availability of satellite imagery and GIS databases enhances the evaluation of human impact on changing ecosystems. This area was not selected by accident; it has been cultivated by people for very long time starting from the Palaeolithic age. Furthermore it is a trans-border zone between settled and nomadic cultures. The impact of extensive human activities to changes in land use are described by landscape features in a GIS layer based environment using several physical characteristics (vegetation, soils, agricultural use). MapModels will be used to develop scenarios of anthropogenic impact on the process of increasing steppe development. The use of fuzzy set based logical generalizations in modeling geographical phenomena is argued to improve human impact assessment as well as to use GIS as an advanced decision support tool to promote regional development. A knowledge - driven approach is suggested as a working environment to further enhance environmental risk evaluation. Future steps necessary are listed.

We describe how the multi-resolution analysis using the Haar wavelets and the corresponding discrete wavelet transformation change when the approximate functions involved are represented by fuzzy systems. Using the scaling function of the Haar wavelet or the step function as the input fuzzy sets, we prove that the fuzzy system representation of the intermediate approximate functions differ from the corresponding functions in the MRA of the Haar wavelet by O(h) where h is the length of the subintervals. Thus, the fuzzy system representation of the functions are identical with those of the Haar wavelet when it is applied to the gray scale image or when the manipulations of the function values are performed in integers with a fixed number of bits. We prove that the approximation based on the Haar wavelet picks up the trend curve when it is applied to a band pass filtered sinusoidal signal. An example to determine the trend of the temperature change curve from the power of the Johnson noise signal is included.

In this paper, the stability-plasticity behaviors of Minimal Resource Allocation Network (MRAN) and Probabilistic Neural Network (PNN) in differentiating targets by means of sonar signals have been investigated. MRAN has shown to have lower network complexity but experiences higher plasticity compared to PNN in the experimental study. In terms of on-line learning performance, MRAN has also proven to be more superior to PNN.

Our aim in this paper is to develop a systematic method for design of stable fuzzy controllers using describing function technique. A given multi-input single-output fuzzy system is first reduced to a set of single-input single-output fuzzy systems using the additivity property. Then, the analytical expressions for the describing functions for each of the single-input single-output system are evaluated. The describing function of a fuzzy system is an interval-valued function; therefore, robust control results for the stability analysis are used. We have applied the theoretical results to the single-axis control of the Phantom haptic interface robotic manipulator.

This paper proposed an intelligent robot for In Vitro Fertilization (IVF) clinic research. The surgeon consistent is important issues for the research. The end effector of the robot was obtained through right, left, up, and down motions at the location. The robot was attachable to the bedside using a conventional scope holder with multiple joints. The control system was expected to enhance the overall performance of the system during eggs retrieval procedure. The auto-tracking control algorithms were demonstrated and computed for the potential assistant robot system.

A world model adapted to changes in the robot's environment as well as to changes in its operations permits a situated agent, i.e. a mobile robot, to properly plan and execute its tasks within its workspace. Such adaptation of the symbolic world model involves two problems: maintaining the world model coherent with the environment information, and optimizing the model to perform efficiently the robot tasks. This paper presents a first step towards the self-adaptation of the symbolic world model of a mobile robot with respect to both aspects. Then, the paper focuses on the optimization subproblem, proposing an evolution-based approach to keep the symbolic model of the robot optimized over time with respect to the current agent's operational needs (tasks).

This paper describes the use of a Genetic Algorithm (GA) for the problem of Offline Point-to-Point Autonomous Mobile Robot Path Planning". The problem consist of generating "valid" paths or trajectories, for the robot to use to move from a starting position to a destination across a flat map of a terrain, represented by a 2 dimensional grid, with obstacles and dangerous ground that the Robot must evade. This means that the GA optimizes possible paths based on two criteria: length and difficulty. First, we decided to use a Conventional GA to evaluate its ability to solve this problem (using only one criteria for optimization), and due to the fact that we want to optimize paths under the two criteria or objectives, then we extended the Conventional GA to implement the ideas of Pareto optimality, making it a Multiple Objective Genetic Algorithm (MOGA). We present useful performance measures and simulation results of the Conventional GA and of the MOGA that show that both Genetic Algorithms are effective tools for solving the point-to-point robot path planning problem.

The application of fuzzy logic based control is a way for modeling human behavior successfully proven. These techniques are very useful, particularly in the automated vehicle field, where the elements to be controlled are mostly very complex and cannot be described by a linear model. This paper presents a control system for vehicle steering. We have developed a set of controllers, which have been fine tuned according to human experience and a study of the human driving. The comprehension and interaction of input control variables are very important factors during control adjustment. These controllers have been used and tested on human-like automatic driving routes, generating a robust and repetitive system.

This paper presents a new sonar based landmark to represent significant places in an environment for localization purposes. This landmark is based on extracting the contour free of obstacles around the robot from a local evidence grid. This contour is represented by its curvature, calculated by a noise-resistant function which adapts to the natural scale of the contour at each point. Then, curvature is reduced to a short feature vector by using Principal Component Analysis. The landmark calculation method has been successfully tested in a medium scale real environment using a Pioneer robot with Polaroid sonar sensors.

Automatic parking of a car-like robot in the presence of a priori unknown obstacles is the topic of the paper. This problem is solved by means of a bio-inspired approach in which the robot controller does not need to know the car kinematics and dynamics, neither does it call for a priori knowledge of the environment map. The key point in the proposed approach is the definition of performance indexes that for automatic parking happen to be functions of the strategic orientations to be injected, in real time, to the robot controller. This solution leads to a dynamical multicriteria optimization problem, which is extremely hard to be dealt with analytically. A genetic algorithm is therefore applied. The results of computer simulations are finally discussed.

This paper deals with the modeling of the relations between the functional properties (outputs) and the structural parameters (inputs) of nonwoven products using soft computing techniques. To reduce the complexity of such model, a selection of the most relevant input parameters is required. A new selection method based on a ranking criterion is first presented. Several models, taking into account specificities of the nonwoven families, are secondly defined using multilayer feed forward neural network. The interest of the selection method is also tested and discussed.

In this paper, a new fuzzy multi-attribute decision making framework for group decision making is proposed. The proposed framework combines both the concept of fuzzy outranking and fuzzy attributes to provide a more flexible way of comparing the alternatives. It consists of three steps. In the first step, the preferences of the decision makers are gathered and then aggregated. In the second step, the attributes are normalized by considering the threshold and the target values of each attribute. At the last step, the results of the previous steps are used as input to a fuzzy multi-attribute decision making approach in order to rank and/or select alternatives. A fuzzy logic based computation is applied in each step for the imprecise or ambiguous data.

In this paper, an efficient fault diagnosis method for Air-Handling Unit (AHU) using dynamic fuzzy neural networks (DFNNs) is presented. The proposed fault diagnosis method has the following salient features: (1) Structure identification and parameters estimation are performed automatically and simultaneously without partitioning the input space and selecting initial parameters a priori; (2) Fuzzy rules can be recruited or deleted dynamically; (3) Fuzzy rules can be generated quickly without resorting to the backpropagation (BP) iteration learning, a common approach adopted by many existing methods. Simulation results demonstrate that fast training and diagnosis speed and high diagnosis rate can be achieved.

To increase customer satisfaction, Quality Function Deployment (QFD) is used to translate customer needs (CNs) into technical design requirements (DRs). Determination of DRs for product development is very important since these requirements are the vital keys of successful products. The methods used to evaluate DRs in the literature can be categorized into multi-criteria evaluation methods such as scoring methods, analytic hierarchy method (AHP), analytic network process (ANP), etc. There is a limited number of papers using multi-attribute outranking methods to evaluate DRs. This paper aims at comparing the results of three different fuzzy outranking methods to evaluate the DRs. A numerical example is presented to illustrate the use of these methods. A sensitivity analysis by changing thresholds is also made by using a software.

This paper proposes an analytic strategic planning framework that can assist managers to put their e-business projects into practice in an effective and efficient way. The suggested framework combines two widely used strategic management tools, the SWOT (Strengths, Weaknesses, Opportunities and Threats) matrix and the balanced scorecard, identifying the four critical successful perspectives for strategy development in e-business application. A decision analysis method, fuzzy analytic hierarchy process, is applied to the framework for improving the quantitative information basis of the strategic planning process and for providing with a more effective and comprehensive decision support under uncertain conditions. This paper also includes a case study in Turkey in which the proposed framework is used for the e-health service strategy development.

In this paper, a decision problem whether the landowner should begin to construct the building at present for sale in the future is considered. The uncertainty of building price in the future is characterized by possibility distribution to reflect the potential how much the building price will be. Two focus points, called active focus point and passive focus point are introduced as sorts of equilibriums to balance utility and possibility to show which values should be considered for making decision under possibilistic information. Based on such two focus points, a general model is developed to analyze the investment behavior of landowner. The optimal building size based on active focus point is larger than the one based on passive focus point. Increasing the uncertainty of price can make active investor increase investment scale and careful investor decrease investment scale. It can be seen that the proposed method seems more reasonable for such one-shot decision problem than maximizing expected utility in probability framework.

An effective facility layout evaluation procedure necessitates the consideration of qualitative criteria, e.g. flexibility in volume and variety and quality related to the product and production, as well as quantitative criteria such as material handling cost, adjacency score, shape ratio, material handling vehicle utilization, etc. in the decision process. This paper presents a decision-making methodology based on data envelopment analysis (DEA), which uses both quantitative and qualitative criteria, for evaluating Facility Layout Design (FLD). A computer-aided layout-planning tool, VisFactory, is adopted to facilitate the layout alternative design process as well as to collect quantitative data. Fuzzy AHP is then applied to collect qualitative data related to quality and flexibility. DEA methodology is used to solve the layout design problem by simultaneously considering both the quantitative and qualitative data. The purposed integrated procedure is applied to a real data set consisting of twelve FLDs provided of the plastic profile production system.

We describe in this paper a new hybrid intelligent approach for industrial quality control combining neural networks, fuzzy logic and fractal theory. We also describe the application of the neuro-fuzzy-fractal approach to the problem of quality control in the manufacturing of sound speakers in a real plant. The quality control of the speakers was done before by manually checking the quality of sound achieved after production [4]. A human expert evaluates the quality of sound of the speakers to decide if production quality was achieved. Of course, this manual checking of the speakers is time consuming and occasionally was the cause of error in quality evaluation [8]. For this reason, it was necessary to consider automating the quality control of the sound speakers.

In recognition of the importance of the ERP software selection and of the sizable risk that organizations take when they decide to buy this type of technology, this study proposed a systematic decision process for selecting a suitable ERP package. We are interested in two stages of decision-making: the identification of serious software candidates and the choice of the most appropriate ERP software package. To achieve this, we propose an integrated approach based on fuzzy logic, heuristic multi attribute utility and conjunctive methods. Finally, a case study is given to demonstrate the potential of the methodology.

The ease allowance is an important criterion in garment sales. It is often taken into account in the process of construction of garment patterns. However, the existing pattern generation methods can not provide a suitable estimation of ease allowance, which is strongly related to wearer's body shapes and movements and used fabrics. They can only produce 2D patterns for a fixed standard value of ease allowance. In this paper, we propose a new method of estimating ease allowance of a garment using fuzzy logic and sensory evaluation. Based on these values of ease allowance, we develop a new method of automatic pattern generation, permitting to improve the wearer's fitting perception of a garment. The effectiveness of our method has been validated in the design of trousers of jean type. It can also be applied for designing other types of garment.

Quality consultants are used in ISO 9000 implementation projects by especially small and medium-sized enterprises. Clients do not always appreciate differences between quality consultants. The aim of this paper is to provide an analytical tool to select the best quality consultant providing the most customer satisfaction. The clients of three Turkish quality consultancy firms were interviewed and the most important criteria taken into account by the clients while they were selecting their consultancy firms were determined by a designed questionnaire. The fuzzy analytic hierarchy process was used to compare these consultancy firms. The means of the triangular fuzzy numbers produced by the customers and experts for each comparison were successfully used in the pair wise comparison matrices.

The nonlinear autoregressive moving average with exogenous inputs (NARMAX) model provides a powerful representation for time series analysis, modeling and prediction due to its capability of accommodating the dynamic, complex and non-linear nature of real-world time series prediction problems. This paper focuses on the modeling and prediction of NARMAX-model-based time series using the fuzzy neural network (FNN) methodology. Both feedforward and recurrent FNNs approaches are proposed. Experiments and comparative studies demonstrate that the proposed FNN approaches can effectively learn complex temporal sequences in an adaptive way.

The investigation and application of on-line monitoring programs has been ongoing for over two decades by the U.S. nuclear industry and researchers. To this date, only limited pilot installations have been demonstrated and the original objectives have changed significantly. Much of the early work centered on safety critical sensor calibration monitoring and reduction. The current focus is on both sensor and equipment monitoring. This paper presents the major lessons learned that contributed to the lengthy development process including model development and implementation issues, and the results of a recently completed cost benefit analysis.

Empirical modeling techniques have been applied to on-line process monitoring to detect equipment and instrumentation degradations. However, few applications provide prediction uncertainty estimates, which can provide a measure of confidence in your decisions. This paper presents the development of analytical prediction interval estimate methods for three common non-linear empirical modeling strategies: artificial neural networks (ANN), neural network partial least squares (NNPLS), and local polynomial regression (LPR). The techniques are applied to nuclear power plant operational data for sensor calibration monitoring and verified via bootstrap simulation studies.

The monitoring system with MLP and RBF network has been developed for NPP. MLP is used to detect the symptom of anomalies and RBF network is used to identify the abnormal events. From the off-line test results using PWR simulator, it is clear that the monitoring system can successfully detect and diagnose small anomalies earlier than the conventional alarm system.

This paper describes the technical aspects of the implementation of a Fuzzy Decision Aid System developed by the Mathematics Department of the Polytechnic Faculty of Mons for ONDRAF/NIRAS/NERAS, the Belgian Radioactive Waste Management Agency. The proposed system is intended to deal with the economic uncertainties inherent to radioactive waste management, and especially to High-Level-Waste repository projects.

SDU is a software system that detects leak of the reactor coolant at quasistationary processes. The SDU system has been installed at an Nuclear Power Plan Dukovany. The reactor coolant leak is calculated from instant data in the plant information system by means of a balance model for defined RCS volumes. The system performs an automatic classification of the operation conditions. SDU Functional Tests have been carried out at Unit 2 of Nuclear Power Plan Dukovany in 2003. Tree types of leaks have been used and SDU output information has been evaluated with good result.

In this paper, we look into the issue of clustering of signal transients for the reliable monitoring and timely diagnosing of nuclear components and systems and show that the choice of the metrics upon which the clustering is based is critical for obtaining geometric clusters as close as possible to the real physical classes. The a priori known information regarding the true classes will be exploited to select, by means of an evolutionary algorithm, an optimal metrics. In case the classification thereby obtained were still unsatisfactory, an iterative procedure is used to split the less compact physical classes in further subclasses.

In accordance with the conditions in terms of "RODOS" (Real-time Online Decision Support System) which was developed by the European Community operations so far, this research work has combined itself with actual conditions existing in china and made exploratory discussion upon the construction of COSY and the application of fuzzy decision-making theory/methods in early nuclear emergency decision.

In this work, a Constrained Genetic Programming (CGP) based approach is used to automatically generate optimum surveillance tests policies for a PWR Auxiliary Feedwater System (AFWS). The main idea is to produce surveillance test schedules that maximize the system reliability. The AFWS reliability model is the same used in previous works, where the use of a Genetic Algorithm (GA) was proposed to solve the same problem. The results demonstrated several advantages of the CGP-based approach over the GA-based model.

A new method for estimating reactivity parameters, such as moderator temperature coefficient (MTC) and void reactivity coefficient (VRC), is proposed using steady-state noise data. In order to solve the ill-posed problem of reactivity parameter estimation, a concept of a gray box model is newly introduced. The gray box model includes a first principle based model and a black-box fitting model. The former model acts as a priori knowledge based constraints in a parameter estimation problem. After establishing the gray box and noise source models, the maximum likelihood estimation method based on Kalman filter is applied. Furthermore, it is shown that the frequency domain approach of the gray box model is useful in the case of VRC estimation. The effectiveness of the proposed algorithms is shown through numerical simulation and actual plant data analysis.

Helium turbine is the key component of the power conversion unit for High Temperature Gas-cooled Reactor (HTGR) with direct gas cycle. Gas turbine design currently is a multidisciplinary process in which the relationships between constraints, objective functions and variables are very complicated. Due to the ever-increasing complexity of the process, it is very difficult for the designer to foresee the consequences of changing certain parts. Furthermore, helium is very different in thermal properties from air, it is uncertain whether the operation experiences of an air turbine may be used in the conceptual design of a helium turbine. Therefore, it is difficult to produce an optimal design with the general method of adaptation to baseline. In this paper, the genetic/gradient method is applied in the conceptual design of helium turbine. The design space is reduced by genetic algorithm and then the optimal solution is located by the gradient method. Results show that the convergent speed is high and the optimal design can be obtained by this method. The experiences from air turbine conceptual design could be used in helium turbine conceptual design without evident error.

This paper presents the main characteristics of two different methods to realize the aggregation and the center of gravity stages of a fuzzy controller that is under development for its integration, as an alternative power control algorithm, in the control console of the TRIGA Mark III reactor of the Mexican Nuclear Center. In one case, an innovative method determines, in every control cycle, the group of lines that define the fuzzy aggregated set of the rule base in the continuous domain of the output variable. Likewise, the center of gravity of this aggregated set is analytically obtained. In the other case, the method used to determine both the aggregated set and its center of gravity is based on the discretization of the universe of discourse. These methods were simulated in the ascent and regulation of neutron power and comparison results of their performances are presented.

The nuclear core design optimization problem consists in adjusting several reactor parameters in order to minimize the average peak factor in a three-enrichment zone reactor, considering restrictions on the average thermal flux, criticality and sub-moderation. We solve this problem with a Niching Genetic Algorithm (NGA) with fuzzy clustering and compare its performance to the Standard Genetic Algorithm (SGA). After several experiments, we observed that the former technique performs better than the latter due to a greater exploration of the search space.

In this work, a probabilistic neural network (PNN) that has been applied well to the classification problems is used in order to identify the break locations of loss of coolant accidents (LOCA) such as hot-leg, cold-leg and steam generator tubes. Also, a fuzzy neural network (FNN) is designed to estimate the break size. The inputs to PNN and FNN are time-integrated values obtained by integrating measurement signals during a short time interval after reactor scram. An automatic structure constructor for the fuzzy neural network automatically selects the input variables from the time-integrated values of many measured signals, and optimizes the number of rules and its related parameters. It is verified that the proposed algorithm identifies very well the break locations of LOCAs and also, estimate their break size accurately.

The investigations for insulation particle genesis and transport gain in importance regarding the reactor safety research for Pressurised Water Reactors (PWR) and Boiling Water Reactors (BWR). All types of loss-of-coolant accidents (LOCA) as well as short and long term behaviour of emergency core coolant systems were considered for analysis. A gist of these investigations is the development of 3-D-models simulating two-phase flow consisting of water and insulation particles in large geometries. The background of experimental investigations consists of the following parts:

• generation of a wide data base,

• development and validation of CFD-models (computational fluid dynamics) for the description of insulation particle transport phenomena in flows.

These analysis will be carried out for various geometric and fluidic boundary conditions, as well as sedimentation, resuspension, agglomeration, clogging and increasing of differential pressure at hold-up devices. Three plexiglass test facilities were built for exploration of the mentioned single phenomena. Especially modern flow measurement and digital image processing technologies were applied.

In this paper we use a combination of fuzzy clustering estimation and sliding mode method to control a chaotic system, which its mathematical model is unknown. At first the model of chaotic system is identified without using any input noise signal. In this case the recurrent property of chaotic behavior is used for estimating its model. After estimating the fuzzy model of chaos, control and on-line identification of the input-related section is applied. In this step, we estimate the system model in normal form, such that the dynamic equations can be used in sliding mode control. Finally the proposed technique is applied to the Lorenz system as an example of chaotic system. The simulation results verify the effectiveness of this approach in controlling an unknown chaotic system.

Chaos-based secure communication has been an attractive topic recently. In this paper, the techniques of secure communication based on chaotic synchronization are reviewed, including the schemes of modulating the message signal and the chaotic signal, the approaches of synchronization between the chaotic transmitter and the chaotic receiver, and the security analysis of the communication. Discussion on future studies is carried out.

This paper presents an overview of recent studies on the interaction between fuzzy logic and chaos theory. On the one hand, it is shown that chaotic systems can be transformed to either model-free fuzzy models or model-based fuzzy models, which means that fuzzy systems can also be chaotic. On the other hand, it is further shown that fuzzy systems can be made chaotic (or chaotified) with some simple and implementable controllers, and the existing chaotification approaches are mathematically rigorous in the sense of some commonly used mathematical criteria for chaos such as those defined by Deveney and by Li-Yorke.

The proposed chaos-based encryption algorithm first employs a sawtooth-like chaotic map to generate a pseudo-random bit sequence, then uses it to determine the types of encrypting operations performed on video coding streams. Four kinds of selective encrypting operations are introduced to efficiently scramble and shuffle all of the DC coefficients, part of the AC coefficients of I blocks as well as Motion Vectors (MVs) through XOR, XNOR and replacement operations. With slight computational overhead and tiny data dilation, the encryption algorithm is amenable to the H.263 video-conference coding standard. Finally, the feasibility and the security of the proposed algorithm are demonstrated by experiments carried out on several segments of H.263 coded video streams.

We describe in this paper a new theory of chaos using fuzzy logic techniques. Chaotic behavior in non-linear dynamical systems is very difficult to detect and control. Part of the problem is that mathematical results for chaos are difficult to use in many cases, and even if one could use them there is an underlying uncertainty in the accuracy of the numerical simulations of the dynamical systems. For this reason, we can model the uncertainty of detecting the range of values where chaos occurs, using fuzzy set theory. Using fuzzy sets, we can build a theory of Fuzzy Chaos, where we can use fuzzy sets to describe the behaviors of a system. We illustrate our approach with two cases: Chua's circuit and Duffing's oscillator.

This paper presents a hardware implementation for an improved symmetry features point extraction algorithm used in a range estimation task, for an autonomous mobile robot application. The algorithm that computes the symmetry points is an improved version of the Symmetry Feature Point Extraction algorithm (SFP) presented in [1]. With this algorithm the symmetry feature points are obtained directly from the cycle image acquired by the camera. All experiments are realised with the resources presented in [1] and a hardware development board with a Xilinx2E FPGA circuit.

This paper addresses the design of a fuzzy control system for a drum boiler-turbine-generator unit. In the design procedure, a dynamic fuzzy augmented system is suggested for the nonlinear boiler-turbine plant to deal with its non-minimum phase behavior. The fuzzy control system is synthesized from a local concept viewpoint on the issue of the optimal control theory. The good performance of the designed fuzzy control system is shown by simulations in various conditions.

Fuzzy logic controllers are nowadays mostly implemented as software code and run on conventional microprocessors. If there is a need for high speed processing, the controller must be implemented in hardware. One of the solutions is implementing a fuzzy logic controller in a programmable logic device. Taking into consideration some initial limitations, a universal fuzzy controller core can be constructed. Such controller can be quickly adapted to various system transfer functions, even while the controller is operating. This paper outlines some important design issues that we came across while constructing such fuzzy controller cores.

This paper presents a hardware implementation for a Parallel Pipeline Floating-Point Fuzzy Processor (PPFF) which uses Gaussian membership functions in order to represent the membership functions. The processor inputs are the parameters from a Sugeno fuzzy system. These parameters are developed in an off-line phase and are supposed to be known. The processor inputs are represented in single-precision floating-point and all operations are realised in pipeline. The hardware implementation is realised using a H.O.T II Development System.

The paper presents the issues of dealing with dynamic aspects of operators' performance. They are evaluated on the base of macroscopic "second-by second" context model of cognition and communication processes. The human action context is represented as a field of interaction between human, machine, technology and organization. They form context together but the field influences dynamically their configuration. Some useful achievements of mathematical psychology are applied to context quantification and dynamic reconfiguration of cognition and communication processes modeling. A simplified probabilistic approach to dynamic quantification of operators' performance is demonstrated by the Performance Evaluation of Teamwork method.

AUTHOR INDEX.