Catálogo de publicaciones - libros

Stock market prediction has always been one of the hottest topics in research, as well as a great challenge due to its complex and volatile nature. However, most of the existing methods neglect the impact from mass media that will greatly affect the behavior of investors. In this paper we present a system that combines the information from both related news releases and technical indicators to enhance the predictability of the daily stock price trends. The performance shows that this system can achieve higher accuracy and return than a single source system.

We have designed a kind of practical artificial neural network development software for ordinary engineering technicians. This software, with graphic interface, not only supports multiple types and algorithms of artificial neural networks, but also supports the IEC 61131-3 International Standard. This article, through three application examples of artificial neural networks, shows the feasibility and the easy implementation of this development software, as well as the realization of artificial neural networks in IEC 61131-3 Standard-based software. It also shows the application value of artificial neural networks development tool and the realistic significance of applying artificial neural networks control in the projects.

The gamma ray dual modality densitometry was presented to measure salinity independent of water volume fraction in pipe flows. The simulation geometries of the dual modality densitometry were built using Monte Carlo software Geant4. Computer simulations were carried out with different types of salt and various salinity. The results show that type of salt and salinity have significant effects on the water volume fraction measured by dual modality densitometry. By means of measuring attenuation of transmitted and scattered radiation of dual modality densitometry, the information about the salinity changes can be obtained. But it is difficult to calculate WVF and salinity from dual modality densitometry models. The RBF neural networks were used to predict salinity and water volume fraction. The results show that the predicting values fit true values well. It was demonstrated that the water volume fraction measuring errors caused by salinity can be reduced by using RBF neural networks.

In this paper research and development are ongoing in the implementations of Brillouin-active fiber based, highly versatile active optical device for optical communication, sensing and computation. An active device in general requires the employment of nonlinearity, and possibly feedback for increased efficiency in device function. However, the presence of nonlinearity together with intrinsic delayed feedback has been repeatedly demonstrated to lead to instabilities and ultimate optical chaos. Our effort is then to exploit device function and suppress instabilities by simulation, and design for optimization based on neural networks in smart structures. Instabilities are unavoidable in optical fiber due to its intrinsic nonlinearity and feedback instabilities. The suppression of such instabilities is devoted to the exploitation of them for memory capacity. These memories can be estimated as an optical logic function used for all-optic in-line switching, channel selection, oscillation, optical logic elements in optical computation with neural network application.

This paper presents a novel approach to solve the hydro plant dispatch problem based on the artificial neural network (ANN) and genetic algorithm (GA). In this work, the difficult water balance constraints are embedded and satisfied throughout the proposed encoding and decoding algorithms. The ANN is used as a pre-dispatch tool to generate raw hydro output for each hour temporarily ignoring time-dependent constraints. Then, the proposed decoding algorithm decodes the raw schedule of each plant into a feasible one. Finally, a GA is used to find the optimal schedule. The proposed approach is applied to an actual utility system of four hydro plants and 22 thermal units with great success. Results show that the new approach obtains a more highly optimal solution than the conventional dynamic programming method.

The paper presents a wavelet packet neural network (WPNN) approach for solving the waveform distortion problem of protective relaying testing instrument. With its excellent time-frequency localization property and approximation ability, WPNN is used to establish an identification model of the non-linear amplifier of the protective relaying testing instrument. The fault data to be put into the instrument is compensated by an adjusting function getting from the identification model, which makes the whole instrumentation system show linear performance so that the distortion of the output waveform is constrained greatly. Simulation results indicate the feasibility and validity of the proposed approach, and a prototype has been put into practical operation.

The controlled system is an uncertain nonlinear differential-algebraic subsystem (DASs) in a large-scale system. The problem of robust stabilization for such class of uncertain nonlinear DASs is considered in this paper. The robust stabilization controller is proposed based on backstepping approach using two-layer Artificial Neural Networks (ANN) whose weights are updated on-line. The closed-loop error systems are uniformly ultimately bounded (UUB) and the error of convergence can be made arbitrarily small. Finally, using the design scheme proposed in this paper, a governor controller is designed for one synchronous generator in a multi-machine power systems. The simulation results demonstrate the effectiveness of the proposed scheme.

This paper presents a state-of-charge (SOC) estimation method based on extension neural network (ENN) theory for lead-acid batteries. First, a constant electric current discharging experiment with an electronic load for lead-acid batteries is made to measure and record the internal resistance and open-circuit-voltage by utilizing internal resistance tester. Then, the experimental data are adopted to construct an estimation method based on ENN for recognizing the residual capacity of the lead-acid battery. The simulated results indicate that the proposed estimation method can determine the residual capacity of lead-acid batteries rapidly and accurately with less time and memory consumption.

This paper describes the development of an artificial neural networks-based in-process flash monitoring system (ANN-IPFM) in the injection molding process. This proposed system integrates two sub-systems. One is the vibration monitoring sub-system that utilizes an accelerometer sensor to collect and process vibration signals during the injection molding process. The other, a threshold prediction sub-system, predicts a control threshold based on the process parameter settings, thus allowing the system to adapt to changes in these settings. The integrated system compares the monitored vibration signals with the control threshold to predict whether or not flash will occur. The performance of the ANN-IPFM system was determined by using varying ratios of polystyrene (PS) and low-density polyethylene (LDPE) in the injection molding process, and comparing the number of actual occurrences of flash with the number of occurrences predicted by the system. After a 180 trials, results demonstrated that the ANN-IPFM system could predict flash with 92.7% accuracy.

In this paper, the control approach to irregular structures excited by multi-dimensional ground motions is presented by using semi-active tuned liquid column damper (TLCD). A back propagation Artificial Neural Network (ANN) is used to predict the responses of structure due to two-dimensional seismic inputs. The semi-active control strategy is established and implemented based on ANN. The numerical examples have shown that it is an effective method presented for controlling the translational and rotational responses of irregular structures.