Catálogo de publicaciones - libros

The consequence of noise on nonlinear dynamical systems is an extremely significant area of research. All systems, physiological and other, grow in the existence of noise pouring forces. It is a notion that noise has only a blurring effect on the growth of dynamical systems. In nonlinear systems, noise can significantly alter the deterministic dynamics. In cortical cell synaptic bombardment can be contemplation as a source of current noise. The firing behavior of the cells may be effectively described by assuming that the model differential equations governing the excitability of the cells are coupled to a random variable describing the properties of this current noise. Inspired from these facts, we have studied the effect of noise on FitzHugh-Nagumo model and Cortical Neuron model. This study of noise in neuron models is helpful in understanding the brain dynamics.

In this paper we present some real time results of the implementation on a mobile robot of visual self-localization algorithms based on Morphological Heteroassociative Memories (MHM). We propose a dual set of min/max MHM storing the views that serve as landmarks for self localization. The binarized input images are subject to erosion in order to increase the robustness of the recall process. We present some empirical results on basic navigation experiments in an indoor environment. We use as the measure of performance of our approach the rate of false recognition, conditioned to some landmark being recognized.

Analysis of < > models is considered as part of a general approach to solving optimization problems with fuzzy coefficients. This approach consists in formulating and solving one and the same problem within the framework of mutually interrelated models with constructing equivalent analogs with fuzzy coefficients in objective functions alone. The use of the approach allows one to maximally cut off dominated alternatives. The subsequent contraction of the decision uncertainty regions is based on reduction of the problem to models of multiobjective decision-making in a fuzzy environment with the use of fuzzy preference relation techniques for analyzing these models. Three techniques are considered in the paper. The first technique is of a lexicographic character and consists in step-by-step introducing criteria (fuzzy preference relations). The second technique is based on building of a membership function of a subset of nondominated alternatives with simultaneous considering all preference relations. The third technique is related to aggregating membership functions of subsets of nondominated alternatives corresponding to each preference relation. The results of the paper are of a universal character and are already being used to solve problems of power engineering and management.

In order to realize autonomous running for a vehicle in the agricultural works, this paper proposes an automatic acquisition method of fuzzy control knowledge for orbit tracking using genetic algorithms (GA). We decide the steering angle of a vehicle by fuzzy inference according to the relative distance and angles between the orbit and the vehicle to track an orbit for agricultural work. The table of fuzzy control rules and the membership functions of input and output for fuzzy inference can be automatically and quickly decided by the method proposed in this paper. The efficiency of the method has been verified by simulations and tests of real vehicle model.

A fuzzy controller requires a control engineer to tune its fuzzy rules for a given problem to be solved. To reduce the burden, we develop a gradient-based tuning method for a fuzzy controller. The developed method is closely related to reinforcement learning, but takes advantages of a practical assumption made for faster learning. In reinforcement learning, values of problem states need to be learned through lots of trial-and-error interactions between the controller and the plant. And the plant dynamics should also be learned by the controller. In this research, we assume that an approximated value function of the problem states can be represented as a function of a Euclidean distance from a goal state and an action executed at the state. And, using it as an evaluation function, the fuzzy controller is tuned to have an optimal policy for reaching the goal state despite an unknown plant dynamics. Our experimental results on a pole-balancing problem show that the proposed method is efficient and effective in solving not only a set-point problem but also a tracking problem.

Evolutionary strategies are heuristic algorithms mimicking natural evolution processes. They have been developed to solve non-linear optimization problems. We employ an evolutionary strategy to solve identification problem of a fuzzy measure on a finite set. We formulate the problem as an non-linear minimizing problem and report some results of numerical experiments.

Development and the design of the product divide the product into the unit or the module. And, work is divided into an individual designer. Therefore, it is difficult to confirm the influence that the design parameter of an individual unit gives to the product. It turns out that the common cognitive process by design review is needed for such a system design of the whole apparatus and needs a CSCW (Computer-Supported Cooperative Work) tool.

In this paper, the print process by electrophotography was analyzed using CSCW tool based FISM(Flexible Interpretive Structural Modeling), consequently the strong complicated parameter of nonlinear nature is extracted easily, and a designer can be supported.

A spatial-temporal segmentation algorithm based on comparative histogram for video object segmentation is proposed in this paper. First, a comparative histogram algorithm based on hierarchical distributed genetic algorithm is used to color segmentation. Next, moving regions are identified by a motion detection method, which is developed based on the several consecutive frame differences to circumvent the motion estimation complexity for the whole frame. At the third step, color segmentation and temporal segmentation results are integrated to obtain video object initial mask. Moreover, post-processing is used to eliminate these noise regions and to filter out the ragged boundary. The proposed algorithm is evaluated for several typical MPEG-4 test sequences. Experimental results show that this algorithm can give accurate object masks and object boundaries throughout the entire test sequences.

GA has an advantage that it can treat target functions without depending on their forms. Recently, many studies have been conducted on applying GA into the policy search. Especially, approaches using a rule based policy with Gaussian Mixture are promising. There is not, however, any genetic operator to create a new normal distribution from plural ones. We propose an effective policy representation and a new genetic operator INDX for it. The performance of the proposed method is shown, applying it to two benchmark problems, the Mountain Car and the Cart-Pole Swing up task.

The new PCNN algorithm introduced here is an autonomous image segmentation algorithm. This algorithm outweighs conventional PCNN algorithms in the following aspects: PCNN’s iterative computation can be terminated automatically; noise can be reduced effectively by the algorithm of pulsing-pattern-based noise detection and adaptive synaptic modification; all information contained in PCNN’s output sequence can be utilized. Furthermore, our algorithm inherits the advantages of PCNN, such as parallel processing and parameter robustness. Experiments show that this algorithm performs well in different kinds of images. As for future research, it would be very attractive to apply this algorithm to the field of pattern recognition.