In public spaces, there are a number of different contents such as visitors, physical information resources, and virtual information resources via their embedded output devices. Therefore, we might unexpectedly enter the public spaces that have our unauthorized contents and/or unwanted characteristics. The previous papers have introduced the novel concept of Secure Spaces, physical environments in which any visitor is protected from being pushed her unwanted information resources on and also any information resource is always protected from being accessed by its unauthorized visitors, and the model and architecture for space entry control and information access control based on their dynamically changing contents. Aiming to build more flexible Secure Spaces, this paper proposes an extended model for context-aware search control based on how preferentially a virtual information resource should be outputted in a Secure Space as well as spatial entry control based on whether a virtual information resource should be granted or denied to be outputted in a Secure Space.

抄録全体を表示

We have researched an interactive manipulation model of virtual liquid. Representing a vortex, a spinning flow of liquid, simulating a stirring action of liquid, and coloring of the water object through changing the surface color has been developed. In this paper we propose a VR chemical laboratory system using these techniques. This system can be used for different kind of purposes, for example e-learning. E-learning offers potentially cheaper education, increased accessibility and convenience for users. Most existing systems rely on transmitting information through text material, but by adding virtual reality content to the system the learning experience can surely be improved.

抄録全体を表示

This paper presents a preliminary study on the possibility of using ultrasound in driver assistance systems. Subjects' lap time in a driving video game was measured as an index of their performance of driving operations under acoustic conditions with and without an ultrasound signal. The results show that the performance characteristics of the subjects changed when the ultrasound signal was presented.

抄録全体を表示

In these days, the global warming has been becoming a major concern. According to some news resources, the amount of CO2 emissions in the home in a year is much larger than that by a car. With this background, the objective of our study is to let home residents recognized how much they use electricity in their everyday lives and make them aware of home ecology. As a first step for this, we propose a method to verbalize the amount and the trends of electricity consumption so that they can recognize them easily.

抄録全体を表示

This paper introduces an integral control approach to high-gain observer design. and its stability is analyzed based on Lyapunov theory. It is assumed that their states are unmeasurable. The proposed high-gain observer has the integrator of the estimation error in dynamics. It improves the stability of high-gain observers, which is robust to noisy measurements, uncertainties and peaking phenomenon as well. Its stability is analyzed by the Lyapunov approach. In order to verify the effectiveness of the proposed scheme, some graphical analysis is given

抄録全体を表示

In the pattern classification problem, fuzzy classifier is useful when linguistic interpretability is required. The performance of a fuzzy classifier depends on its rules and membership functions. In this paper, we have proposed a method that constructs customized fuzzy sets by Bayesian probability. Based on the probability, we build probability ratio map for feature space and transformed it into membership functions. The proposed method can be used in a variety of applications such as a recommender system for web contents of IPTV services. Simulation and experiments results are given to show the better performance of the proposed method than other methods.

抄録全体を表示

In this paper, for human identification, the silhouette image is segmented into four components to evaluate the partial information about gait. A number of combinations of the components are considered to demonstrate the effectiveness of each component to human gait recognition. Besides, we use multiple distances for different templates and fuse them at a score level. The experiments on the well-known CASIA gait dataset A are carried out to demonstrate the validity of the proposed methods.

抄録全体を表示

This paper proposed an algorithm and 3D graphical simulation approaches for autonomous navigation of UGV(Unmanned Ground Vehicle). The autonomous navigation algorithm is proposed for the environment where the vehicle is navigating on a multi-lane road with other multiple vehicles. The algorithm for autonomous navigation is mainly focused on seeking the optimal trajectory to avoid collisions with adjacent multiple vehicles navigating on multi-lane road environment. The 3D graphical simulation model is constructed to visualize the autonomous navigation control and verify the proposed algorithm and the control system.

抄録全体を表示

This paper is concerned with gain-scheduling-based track-seek control for single stage actuator of hard disk drives. The proposed controller schedules controller gains to enable improving the transient response and minimizing overshoot during the functions of the read/write head positioning servomechanism for the seek control. Finally, simulation example is utilized to illustrate the effectiveness of the controller proposed in this paper.

抄録全体を表示

Voxel based morphometry (VBM) is an approach which extracts voxels whose brain tissue density is statistically different from control groups using magnetic resonance (MR) images. It has been applied to diagnose cerebral disorders, cognitive dementias, etc. However, VBM is not applied to neonatal brains because of large differences between the adult brain and the neonatal brain. This article proposes a modified VBM for neonatal brain. The modified VBM segments the brain region by using skull stripping method for neonates, and normalizes the segmented brain into the standardized neonatal brain. The standardized neonatal brain is constructed by using non rigid image registration. The proposed method was applied to 12 neonates whose age was between 0 weeks 30 weeks. For any neonates, the method normalized the brain shape successfully.

抄録全体を表示

In case of pediatric, about 80% intractable epilepsy patients are accompanied with cortical dysplasia. It is very effective to detect cortical dysplasia lesions in order to we plan treatment of intractable epilepsy. Although MR image analysis attracts the considerable attentions because it is non-invasive for the human body, automated methods for detecting cortical dysplasia with MR images of pediatric brain are not established yet. In addition, image features on the pediatric brain MR images of cortical dysplasia are not clarified. This paper proposes an estimation method for detecting cortical dysplasia in pediatric brain MR images. And, fractal dimension is evaluated as a new image feature of cortical dysplasia. The experimental results in four patients (6, 7, 2 and 3 years old) with cortical dysplasia showed that a mean sensitivity of 94.2 %, a mean specificity of 93.8%, and a mean efficiency of 93.8 %.

抄録全体を表示

Synergetic Computers (SCs) are a type of self-organizing ones and have widely been used for pattern recognition or classification problems. In conventional SCs, the order parameters are usually encoded by real values, which determines internal states. In encoding the information of images, however, the complex or hyper-complex values are used to achieve the outperformance. In this paper, we propose Complex-Valued Synergetic Computers (CVSCs) where the order parameters are encoded by complex values. We further investigate the performances of our proposed CVSCs through the recognition of gray-scale images, by comparing it with conventional Real-Valued Synergetic Computers (RVSCs).

抄録全体を表示

Quantum-Inspired Evolutionary Algorithm (QEA) is an extension of evolutionary algorithm in which quantum mechanics and its representations are introduced. A chromosome in QEA is represented as a series of quantum-bit (qubit), and phase-rotation gates are embedded into the selection process over generations. This algorithm has been shown to have better performances than the classical ones in small benchmark problems, but this has not yet been applied to larger scale problems. We show the performances of this QEA by solving the Knapsack Problem and Maximum Search Problem. We also investigate the diversity of individuals in a population in order to estimate the robustness against environmental changes.

抄録全体を表示

It is important to record daily activity for well-maintained human health care. For this purpose, a monitoring system based on multiple microelectromechanical systems (MEMS) has been developed. Using the MEMS based monitoring system, several kinds of numerical data of subject's activity can be stored. When subject's activity on a single day is recorded, a huge volume of data is obtained. To estimate the subject's behavior from such a huge volume data, we propose a fuzzy rule based approach. Our proposed method consists of two steps of abstraction. First, action primitives are defined. In the first-step abstraction, a fuzzy rule which maps a part of numerical data onto an action primitive are generated from sample data or human knowledge. Therefore, numerical data are expressed by a sequence of action primitives. Next, a fuzzy rule which maps a sequence of actions onto a behavior is defined by human user for each behavior. In the second-step abstraction, each action sequence is expressed as a behavior by using the defined fuzzy rules. From the results of the abstraction, we can estimate the subject's state.

抄録全体を表示

Understanding of soft computing methodology often requires grasping abstract concepts or imagining complex interactions of large models over long computing cycles. But this can be difficult for students with weak background in mathematics, especially in the early stages of soft computing education. This paper introduces applying a visual programming paradigm as a tool for educational introduction to soft computing methods. For the visual programming paradigm, IntelligentPad proposed by Y.Tanaka is used. IntelligentPad defines a visual appearance to objects or classes, and allows users to operate and link different objects together using the mouse. This paper reports on using IntelligentPad to teach the basic mechanism of cellular automata. The proposed method was applied to 3rd year college students to verify the validity of the proposed teaching method.

抄録全体を表示

In light of the development of information-processing technology and its globalization because of the spread of the Internet, the promotion of IT talent is strongly required, and the creation of not only programming techniques, but also an entire information system and the education of managerial techniques are required. However, in the system-management practice domain, it takes time to prepare a practice environment and to evaluate the possible problem solutions. In this paper, we propose a practice support system for system management, and report on the results of its actual use in real practice

抄録全体を表示

Self-Organizing Feature Map (SOM) is a layered neural network consisting of an input layer and a competitive layer for the data visualization and vector quantization. The accuracy of SOM vector quantization depends on the number of competitive layer's neurons because the codebook vectors correspond to the competitive layer's neurons. Therefore, when an unknown data set is given, it is difficult to decide the sufficient competitive layer size. In this paper, we propose a Tree-Structured SOM (TS-SOM) based method in order to adaptively change the competitive layer size and structure. TS-SOM is a faster SOM method applying tree search algorithm. We applied the pruning of neurons and layer creation to the tree structure of TS-SOM by using the means error among neighboring neurons.

抄録全体を表示

Many factors are required to demonstrate the urban activity for evaluation of city planning policy in the local city. The multi-agent model is proposed to be established concerning with many urban activities and interaction between agents. In addition to the fundamental purpose of regional development in city development policy evaluation, the evaluation method of city development is discussed connecting with the problems in local city. The multi-agent model is proposed as an artificial society involving measurement mechanism of environmental load. The citizen agents and commercial store agents are assumed in the complex system model. The fundamental formulation of the agent activities is discussed. Furthermore, the evolutional process for urban activities is additionally mentioned. As the artificial society as summary of urban citizens is established, comprehensive evaluation of city planning policy becomes available and the sustainable direction of city planning can be shown with observation of future change of society.

抄録全体を表示

The global warming problem by greenhouse gas emission has become more serious. Also in the transportation section, construction of the sustainable transportation system in environment has been an important subject. Not only improvement of service level of modes except private vehicle, but also consciousness for environmental problem of individual trip maker is important for eco-commuting promotion. Eco consciousness would be changed by influence of other people. The multi agent simulation system for planning of eco-commuting promotion is developed with social interaction model for eco consciousness in the study. Modal shift is modeled by considering the result of a transport social experiment for eco-commuting promotion as reference. Influence of promotion plan such as improvement of service level of public transport, promotion of individual eco consciousness and strengthening interaction of individuals can be analyzed with the proposed model. Finally, it can be concluded that the proposed multi agent simulation with social interaction for eco consciousness is useful for planning of eco-commuting promotion.

抄録全体を表示

Global warming and its mitigation measures such as GHG (Greenhouse Gas) reduction have been extensively discussed all over the world. The transport sector occupied about 20% within the GHG of Japan. Countermeasures in the road traffic are urgent, since about 90% of the GHG emission of the transport sector is emitted from automobiles. For the effective reduction of the GHG such as CO2, quantitative analysis is necessary. Though the emission from the individual vehicle depends on various factors such as engine rotational frequency, speed, and acceleration, that relation is complicated. Therefore, there is a limit on the modeling by the regression analysis. Then, the model which estimates the CO2 emission from the individual vehicle using fuzzy neural network is developed in this study. By including this model in the traffic simulation, the estimation of CO2 emission becomes possible. Therefore, it is possible to examine traffic countermeasures considering the environmental effect.

抄録全体を表示

The authors propose an augmented reality application system for Mahjong. The idea of this application is that visual augmentation by means of the AR method will help a Mahjong beginner play games, learn rules from his/her game experiences under the support of the system, and thus become an experienced player who can play without any support. The system captures Mahjong tiles via a USB camera, recognize the current state, and visually augment tiles to guide the user perform desirable actions. Our implementation and evaluation of the system are reported in this paper.

抄録全体を表示

Providing attractive and interesting e-learning courses material as well as normal lecture in the class is one of important issues in e-learning society. Knowing about this issue, first, it is essential to investigate student interest about topic and types of e-learning courses delivered by measuring their responses. Thus, this study engages in student's responses initiated by two distinction e-learning materials; one is characterized by an interactive material and the other one is non-interactive material, based on physiological signals or biosignals, i.e., hemodynamic responses. Use of multilinear principal component analysis (MPCA) was then proposed in this study to extract the representative features of hemodynamic responses. Finally, the classification was performed by support vector machine (SVM) to discriminate hemodynamic response of student.

抄録全体を表示

The authors have developed a programming training system based on the ARCS learning model, and have applied the system in an actual programming course. The ARCS model is known as a learning model aiming at stimulating and sustaining the learner's motivation to learn. In this training system, learners create programs similarly to solving a puzzle game. In such a learning support system, it is important that the learner can view one's learning progress objectively. Furthermore, if the learning progress could be compared with those of other learners, competition will occur to stimulate the learner's motivation. In this paper, the features to realize the real-time monitoring of each the learner's progress are proposed, and implemented in the puzzle-based programming training system CAPTAIN currently under development. The proposed features allows the learner to view one's own progress ranking compared to the whole class, and also enables the teacher to monitor each individual learner, in order to quickly spot and support learners having difficulty.

抄録全体を表示

In this paper, for improving classification performance of a term selection based on GA, we modify its evaluation function and mutation operation. In the term selection based on GA, two objectives which are maximizing correctly classified texts and minimizing selected terms are optimized. The weighted sum of these two objectives was used as the evaluation function. Therefore, GA-based term selection is performed aiming at the improvement in classification performance on testing sets. This causes the performance deterioration over completely unseen texts. This is because terms are deleted excessively even when the terms have important role for the classification. First, we use NSGA-II for finding non-dominated solutions. As the result, we can have a set of pareto-optimal solutions. Each individual is evaluated by using SVM with -fold cross validation. In this paper, we also modify the mutation operation. The modified mutation operation uses the statistic information of each term as the mutation probability. From numerical simulation results, we show effectiveness of our modification.

抄録全体を表示

In this study, we developed a waterproof cardiography/accelerogram device for human health and activity monitoring. The sensor system consists of a cardiograhy sensor, a tri-axial accelerometer, sensor interface, microcontroller and a RF module. It was coated with Poly dimethyl siloxane ( PDMS ) to achieve a waterproof and patch-type device. The RF module with low power consumption was utilized to transfer the cardiography/accelerogram data from the device to the base station. Static postures and long-term human activity were estimated from the data of cardiography sensor, triaxial accelerometer and other integrated sensors. The details of the system configuration and some measurement examples for human health monitoring are described in this paper.

抄録全体を表示

Patients falling in hospital and care house is very important issue. Each patient at risk of falling should receive multifaceted clinical and environmental interventions that could reduce the risk. However, it is difficult to predict falling from bed using sensor and video monitoring. Also, most of them are not able to use practically, because system is so large and high costs. We propose new falling analysis system using Intelligent Vision Sensor (IVS). The IVS is a CMOS sensor like as human retina. This sensor is able to protect the privacy of patients using image processing on the CMOS hardware. Also, the algorithm has developed to analyze patient's falling pattern. The test bed system is estimated using our proposed motion analysis algorithm for non-handicapped person. It is verified to predict falling in real time and to be able to predict before the accident.

抄録全体を表示

This paper studies rule induction based on rough sets. When the decision attribute is ordinal, inducing rules about upward and downward unions of decision classes is advantageous in the simplicity of the obtained rules. However, because of independent applications of the rule induction method, inclusion relations among upward/downward unions in conclusion parts are not inherited to the condition parts of the obtained rules. This non-inheritance may debase the quality of obtained rules. To ensure the inheritance of inclusion relations among conclusions to conditions, we have proposed rule induction approaches in the previous study. In this paper, performance of the proposed approaches are further examined by numerical experiments using more general data sets.

抄録全体を表示

In this paper, we make some basic consideration for reformulating roles of sequences in uncertain information processing. For the purpose, we reformulate naive subsets and multisets as some quotient sets induced from free monoids as the set of finite sequences, i.e, strings based on equivalence relations. Then we explain the process of computing images of mappings in the framework of the above reformulation. Finally we remark that fuzzy multisets would be better defined as multisets on the set of truth values excluding 0, which should behave as the identity element because multisets must not count non-membership.

抄録全体を表示

The classical rough set model has been generalized in many ways. The equivalence relation underlying the classical rough set model is generalized to a tolerance relation, a similarity relation, a dominance relation and to general relations. Moreover, the partition equivalent to the equivalence relation underlying the classical rough set model is generalized to a cover and families. Those relation-based generalization and family-based generalization depend on the interpretations of rough set models: classification and approximation. On the other hand, lower and upper approximations are generalized to positive and nonnegative regions by relaxing their conditions to be a member so as to accept some errors. The models generalized in this way are called variable-precision rough set models. In this paper, we combine the variable-precision type generalization with the relation-based and family-based generalizations. We investigate the variable-precision generalized rough set models from classification-oriented and approximation-oriented views. Several possible definitions are explored and their properties are shown. I

抄録全体を表示

In this paper, a descriptive decision-making model under uncertainty has been introduced. At first, Evidential Decision Making Problem (EDMP) has been defined where Dempster-Shafer Theory has been used to represent uncertainty. Then we showed probability approximation approach of solving EDMP. For deciding the decision weights in different attitudes of decision maker, we have used Ordered Weighted Averaging (OWA) operator. Later on, we applied Prospect Theory to accomplish a descriptive decision-making model. To show the effectiveness of our approach, we have also provided a real life decision problem of choosing job from a set of alternatives.

抄録全体を表示

After classifying a dataset by k-means algorithm, whether or not each cluster should be split or be merged is evaluated by introducing a distortion-ratio map. The distortion-ratio map, which has two axes of the distortion-ratio on splitting and that on merging, is composed of partitional regions having the evaluation on splitting and merging, and each region provides a score for a cluster in the region. Measuring the goodness of clustering results through the map, we can estimate the appropriate number of clusters in the dataset, and acquire its corresponding clustering result.

抄録全体を表示

This article is dealing with the fuzzy clustering method which combines the deterministic annealing (DA) approach with entropy, especially Shannon entropy and Tsallis entropy. By maximizing Shannon entropy, fuzzy entropy or Tsallis entropy within the framework of the fuzzy c-means (FCM) method, membership functions similar to the statistical mechanical distribution functions are obtained. We examine the characteristics of these entropy and membership functions from the statistical mechanical point of view. After that, both Shannon and Tsallis entropy based FCMs are formulated as DA clustering using the very fast annealing (VFA) method as a cooling schedule. Numerical experiments are performed and the obtained results indicates that Tsallis entropy based FCM clustering is suitable for very fast DA clustering.

抄録全体を表示

This paper decomposes every gain gamble into a risky gamble and a certain gain, and develops an axiom system for a weighted additive model with certainty equivalent. A concatenation operation is introduced not only to give a formal expression (segregation expression) to the decomposition but also to enable an extension of the expression (extended segregation expression) in the sense that the risky part includes the concatenations of risky gambles. The certainty equivalent of a gamble can be decomposed into the sum of a conditional certainty equivalent of the risky gamble and the certain gain. Two conditions---segregation monotonicity and segregation equivalence---are provided for such limited extended segregation expressions as the risky part consists of the concatenations of gambles having the same chance event to make the conditional certainty equivalent in the multiplicative form; that is, multiplication of the usual certainty equivalent by a weight that depends on an event partition and the amount of certain gain.

抄録全体を表示

Active noise control has become an attractive area of research due to its several advantages compared to passive methods. This paper introduces two model-based techniques for engine noise canceling, that is, using a moving average (MA) model and a feedforward neural network (NN). A least mean square (LMS) algorithm is used to minimize the error in the MA model while backpropagation algorithm is employed to optimize the NN. Due to its advantages of good robustness and nonlinear processing, the NN is considered to be suitable for a nonlinear signal like engine noise. In order to reduce computational cost, the backpropagation algorithm is applied once at each time step with only one iteration. A comparison of the two methods is carried out. It appears from the study that both methods are effective in reducing the engine noise and that the NN based technique is superior over the MA model, especially for low frequency band.

抄録全体を表示

As a solution of the binding problem with separated extraction of information from the several objects, there is a hypothesis that the property binding of information seems to be represented by phase-locking among neuronal oscillatory firing. We introduce this synchronous firing hypothesis into the self-organizing maps, or SOMs algorithm. Here we propose oscillatory self-organizing neural models. By utilizing the computer simulation results, we report that our models could separate and extract information of the shapes and the colors from two simultaneous inputs without the binding problem.

抄録全体を表示

This study presents a nonlinear control method where a genetic algorithm (GA) is employed in the design of a state feedback controller and estimation of latent states. By assuming that there exist unobserved states and variations, the real plant may not be regulated to the desired point. It is thus required to estimate the latent states and then adjust the controller adaptively in order to maintain the desired performance. By considering a rotary crane system as the control object and employing a feedforward neural network (NN) as a controller, this work shows that the two above tasks can be handled by using GA, which first determines the unobserved states and then redesigns the controller based on the obtained estimates with on-the-fly computations. Simulations show that the GA-based state estimation is effective and the controller is able to regulate the system to the desired state.

抄録全体を表示

In this paper we propose a neural network as a controller (NC) for suppression of load swing with disturbance like a gust in a rotary crane system. The rotary crane system has structure which rotates around the vertical axis. Then, the rotary crane system is a nonholonomic system. In general, it is difficult for nonholonomic system to design the classical control method using a static continuous state feedback law. So, it is necessary for nonholonomic system to design a time-varying feedback controller or a discontinuous feedback controller. Previous research had been successful in constructing the suppression controller of the load swing with a single initial rotation angle and three initial rotation angles when a gust occurred. The NC optimized by genetic algorithm (GA) had good control performance, but could not stabilize the load swing in the untrained condition. The control purpose of this paper is to improve the performance of the NC optimized by GA when the gust occurred at random.

抄録全体を表示

A method to estimate hands' motion of car-driver has been investigated. A camera to capture image of hands around the steering has been used for the sensor, which is set at top of the side pillar of the car driver. After perspective transformation of the captured image to be looked straight at the steering, positions of the hands over the steering have been estimated. Particle filter with elaborated state space model has been used for estimating the hands' motion. Video images in driving simulator have been examined for the hands' motion estimation for evaluation purpose. Also, real-time system of the estimation has been developed and demonstrated in some events in public space.

抄録全体を表示

This paper proposes a particle filter for estimating the hybrid latent states of dynamic systems in an online manner. The hybrid states generally consist of both continuous and discrete valued elements and naturally appear in a variety of tracking applications. For the hybrid state estimation, particle filters have been widely used because of its nonlinearity and non-Gaussianity. The contribution of this paper is to introduce an optimal probability function for discrete elements, and to combine the function and Monte Carlo approximated density for continuous elements. Experimental results show that this combination outperforms the conventional particle filter.

抄録全体を表示

A technique for an estimation of a ship's position is introduced. In general, since an observed position signal with GPS has a large noise component, the probability density function of the observation noise does not satisfy the assumption of Gaussian sequence. In order to solve this problem, sequential data assimilation by using Monte Carlo Filter (MCF) which is a kind of a Particle Filter is introduced. Moreover, in this study, we assume Cauchy distribution as the probability density function of the observation noise. The verification of the proposed technique is carried out by numerical experiments and onboard experiments. From the comparison of the proposed procedure and the procedure by using Kalman Filter, it is confirmed that the proposed procedure can remove the influence of the large noise component.

抄録全体を表示

In this paper, we give a note on chord progression as sequences in Kripke models for non-classical logics. We make some basic considerations on (1) chords as possible worlds and (2) chord progressions as transitions among chords as possible worlds.

抄録全体を表示

In this paper, we propose a heuristic algorithm to extract decision rules based on variable precision rough set models (VPRS models). The VPRS models provides a theoretical basis of regarding probabilistic / inconsistent information in the framework of rough set theory. The main idea of our algorithm is based on construction of suitable $\beta$-lower approximations by giving up to discern some discernible objects that belong to different decision classes each other. All decision rules extracted by our algorithm are guaranteed that the certainty of all extracted decision rules are equal to or higher than the predefined threshold of certainty.

抄録全体を表示

We apply a combined method of heuristic attribute reduction and evaluation of relative reducts in rough set theory to gene expression data analysis. Our method extracts as many relative reducts as possible from gene expression data and select the best relative reduct from a viewpoint of constructing useful decision rules. As an experimental result, our method extracted decision rules about a gene that has been identified as a novel biomarker of human breast cancer in recent studies. This result indicates a possibility of our method as a useful tool for gene expression data analysis.

抄録全体を表示

Game strategy planning is a difficult task in Real Time Strategy (RTS) game AI development. Tree searching technique has been one of the common approaches. However, the increasing use of complicated game rules leads to tree models that are huge and complex, and sometimes even unmanageable. Traversing and modification of the tree structure becomes a time consuming and inflexible task. Our research tries to avoid this top-down strategy planning method and propose a bottom-up approach by apply Fuzzy Integral in extracting useful game strategies from data. We also developed a new fuzzy integral for game decision-making. Compared with the traditional Choquet Integral, we achieved a better result using real Warcraft III battle data and the detail is reported in this paper.

抄録全体を表示

This paper discusses sampled-data control design of fuzzy systems with nonlinear perturbation by the first-order hold input. When we use digital devices to control systems, we need a sampled-data control input. Here, we employ the first-order hold sampled-data control input, and allow a time-varying sampling. Since the first-order hold input can be regarded as a delayed signal, the closed-loop system with the first-order hold state feedback control becomes a system with time-varying delay in state. Hence, we introduce a time-varying delay system approach in this paper. We first give a sufficient condition for the stability of the closed-loop system in terms of linear matrix inequalities(LMIs). We employ a generalized Lyapunov-Krasovskii functional and Leibniz-Newton formula to obtain such a stability condition. These lead to a less conservative stability condition. Then, based on such a stability condition, we propose a design method of the first-order hold sampled-data state feedback controller. The result is extended to robust sampled-data stabilization for a class of uncertain fuzzy systems. A numerical example is given to illustrate our sampled-data state feedback control design.

抄録全体を表示

This paper proposes stabilization of nonlinear control systems approximated by fuzzy if-then rules with singleton consequents. The approximated model is found to be piecewise bilinear. The method of a full-state feedback linearization is applied to stabilize piecewise bilinear control systems. Illustrative examples are given to show the validity of the proposed methods.

抄録全体を表示

This paper proposes a numerical method for determining the control parameter value of nonlinear dynamical systems such that they get chaotic. It can reduce the burden of a person who needs to generate chaos from them. The proposed method combines (a) an efficient numerical method for approximating the maximal Lyapunov exponent with sufficient accuracy, and (b) Particle Swarm Optimization (PSO) method for maximizing the maximum Lyapunov exponent numerically. In addition, (c) the boundedness of trajectory is also introduced as a constraint. Empirical results will also be presented, which verify the usefulness of the proposed method.

抄録全体を表示

In this study, we propose a new multi-point type quasi-chaotic optimization method that can be implemented without a vital parameter tuning. The multi-point type quasi-chaotic optimization method (M-Q-COM), which has been proposed recently, is a global optimization method to solve unconstrained optimization problems in which the simultaneous perturbation gradient approximation is introduced into a multi-point type chaotic optimization method. The M-Q-COM can be applied to a class of problems whose objective function values only can be computed. It is very effective to solve unconstrained multipeaked optimization problems with 100 variables. However, in the M-Q-COM, a parameter called initial sampling parameter must be tuned to generate quasi-chaotic search trajectory that is used to implement global search. In this study, we introduce two approaches to deal the issue. In the first approach, a descent sign vector extracted from an approximated gradient vector is used as a search direction of a search point. Then, its moving distance is determined so that its search process conforms to the search process of the quasi-chaotic optimization method obligatorily. The other approach is an initial sampling parameter estimation. We confirm effectiveness of the proposed methods through applications to the aforementioned benchmark problems.

抄録全体を表示

Vortex phenomena in nature are targeted for creating new metahuristics. The focused dynamics of vortex phenomena are approximated to logarithmic spirals, such as a whirling current, a low pressure, a nautilus shell, arms of spiral galaxies and so on. In this paper, we show that a logarithmic spiral can realize an effective strategy regarding the metaheuristics and propose a new 2-dimensional metaheuristics called spiral optimization utilizing a feature of the logarithmic spiral. The simulation results for 2-dimensional benchmark problems show the effectiveness and the potential of the proposed method.

抄録全体を表示

Conventional eye gaze tracking are based on special contact, optical equipment, Electrooculogram, etc. Although they have been produced valuable results in many areas, they cannot obtain the absolute eye gaze but obtain relative one based on calibration. This study proposes a method for detecting absolute eye gaze using MR images. Because MR images can obtain the sectional images of the eye ball, the method needs no calibration. The proposed method detects the eye gaze by segmenting the Vitreous, and lens of the eye ball. The method was applied to healthy subjects successfully. The estimated results were validated by comparing with the conventional method based on optical equipment.

抄録全体を表示