Catálogo de publicaciones - libros

– This paper describes a new method for monitoring electronic components in computer servers. Conventional monitoring techniques implemented in mid-range and high-end computer systems rely on low-speed low-resolution internal instrumentation that produces low quality measurements of critical system parameters. The poor quality measurements lead to degraded performance of diagnostic and prognostic algorithms. The developed method is based on the capability of an auto-associative memory to restore an associated response given an imprecise key vector. The method is performed in two stages. In the collection stage, high-quality signatures representing observable parameters of known good and faulty electronic components are collected using high-speed high-resolution instrumentation. The collected signatures are then stored in the auto-associative memory matrix. In the monitoring stage, low-quality measurements acquired by the low-speed, low-resolution internal instrumentation (an A/D converter accessible via a system bus) are fed into the auto-associative memory. The auto-associative memory restores the high quality signatures. The restored high-quality signatures are then used by diagnostic algorithms for fault detection and isolation and by prognostic algorithms for the remaining useful life estimation.

— This paper discusses the selection of tools in milling operations. An expert system hinged on numerical methods has been developed to carry out this research. The knowledge base is given by limitations in process variables, which let us define the allowable cutting parameter space. The mentioned process variables cause instabilities due to tool-work-piece interaction, knowing as chatter vibration, and the power available in the spindle motor. Then, a tool cost model is contrived. It is used to choose the suitable cutting tool, among a known set of candidate available cutters. Once the cutting tool is selected, the cutting parameters are calculated. For this purpose, other two cost functions are designed, which are depended on the time and frequency domains response properties. An example is presented to illustrate the method.

Canonical normal forms to represent the small signal models of dc-dc converter systems are defined in the paper and their dynamical properties and interconnections are further studied. Both open loop and closed-loop characteristics are covered by using a system theoretic approach. The structural properties of the resulting multi input-multi output models are investigated. Both internal and input-output stability of connected converters described by normal forms are investigated. Implications to controller synthesis are given.

Texture segmentation is an important component in texture analysis. Gabor wavelet shows potential capacity in describing the texture. Kernel-based methods have demonstrated excellent performances in a variety of pattern recognition problems. This paper extracts texture features in the Gabor wavelet transform domain and segments the computed feature image with kernel-based techniques, Spectral Clustering Algorithm (SCA) and Support Vector Machine (SVM). Due to the severe computational complexity of kernel-based techniques, we split the main segmentation process into two steps. SCA is firstly applied to the sampled feature image. Then the clustering result serves as training samples and is used to train SVM. The initial segmentation result is obtained by labeling the feature image with the trained SVM. A median filter is employed to improve the segmentation result. Four mosaic texture images and a natural scene are used to test our new algorithm.

In this paper, an algorithm has been developed to remove additive mixed noise in images with edge preservation. The noise characteristics may vary in the same application from one image to another. In these environments, nonlinear general filters will not perform well and adaptive non-linear filters are best suited. The algorithm based on local statistics such as signal variance and noise Variance is considered. It depends on minimum mean square estimation of the corrupted signal. The signal variance and noise variance are calculated by moving signal window and moving noise window respectively. This offers optimal adaptive filtering in the homogeneous regions as well as in the edges. The adaptive alpha trimmed mean filters with a threshold value have been proposed to preserve edges. The performance of the filter in the presence of different types of noise are evaluated and compared with general mean, median and alpha trimmed mean filters. The image enhancement factor has been calculated as the performance-measuring factor. The Lena image has been considered to carryout the subjective and objective analysis of the proposed filter.

Developing a signal model for approximation of a signal, involves reducing the error between the original signal and that reconstructed from the signal model. This topic has been a research area in signal processing and various mathematical techniques have been applied to solve this problem. This paper aims at presenting a solution using a uniform and redundant transmultiplexer structure, with Singular Value Decomposition (SVD) technique.

In this paper we present an example of a distributed control system implementing state observer controller. Distribution is based on middleware approach. As the middleware for the hardware and network abstraction we used the LabMap/LabNet. For the controller platform we took MATLAB/Simulink. The approach illustrated in this work offers an outlook for the cases where computational resources are limited and distributed among several hardware components, especially in view on embedded application area.

A very simple analog computer was designed and tested to solve any first order constant-coefficients and linear differential equation. The analog computer was built using operational amplifiers, resistors and capacitors. Using the Pspice simulator, various input types were tested across the input terminals of the analog computer.

To master the production it is necessary to be able to adapt the control of this one to the characteristics of a product and individually to ensure the traceability of this last. This way, the product can be considered as the pivot that ensures coherence between the levels of management and manufacture execution. Thus, it becomes necessary to control the production through the product, at the moment “t” and/or through time interval. We present in this article, the problems of production follow-up met within an existing system of production industrial; as well as the used methodology and tools which made it possible to rebuild, by means of systematic recordings, all the process history, to identify the useful elements with precision and to determine the element which acted, which was made or used, at which time, etc.

Water and Power Development Authority (WAPDA) Pakistan has lack of reactive power reserves. The station which needs reactive power to boost the voltage during under voltage problem also requires some source to absorb reactive power to overcome over voltage problem. In this research work, the behavior of static Voltage control through STATCON has been reviewed and real data of WAPDA 500/220 kV power system regarding extreme cases of voltage control has been simulated on power system simulator PSS/E. Different rating of simulated STATCON has been proposed at various locations to overcome voltage stability problem, increase steady state, dynamic state and transient stability. Due to practical nature of the research work, it would be effectively utilized for resolving issues and bottlenecks related to WAPDA primary transmission line.