Wang C.-C.,Chung Cheng Institute of Technology
Expert Systems with Applications | Year: 2011
This study compares the application of two forecasting methods on the amount of Taiwan export, the ARIMA time series method and the fuzzy time series method. Models discussed for the fuzzy time series method include the Factor models, the Heuristic models, and the Markov model. When the sample period is prolong in our models, the ARIMA model shows smaller than predicted error and closer predicted trajectory to the realistic trend than those of the fuzzy model, resulted in more accurate forecasts of the export amount in the ARIMA model. Especially, the coefficient of the error term for the previous period has increased to 79%, implying the influential effect of external factors. These external factors attribute to the export amount of Taiwan according to the economic viewpoints. However, this impact reduces as time progressing and the export amount of the lag period of 12 or 13 do not affect current export amount anymore. In conclusion, when the sample period is shorter with only a small set of data available, the fuzzy time series models can be utilized to predict export values accurately, outperforming the ARIMA model. © 2011 Published by Elsevier Ltd.
Ting T.-H.,Chinese Military Academy |
Wu K.-H.,Chung Cheng Institute of Technology
Journal of Magnetism and Magnetic Materials | Year: 2010
Polyaniline/BaFe12O19 (PANI/Ba ferrite) composites were synthesized by in situ polymerization at different aniline/Ba ferrite weight ratios (Ani/Ba ferrite=1/2, 1/1 and 2/1) and introduced into epoxy resin to be microwave absorber. The spectroscopic characterizations of the formation processes of PANI/Ba ferrite composites were studied using Fourier transform infrared, ultraviolet-visible spectrophotometer, X-ray diffraction, scanning electron microscopy, transmission electron microscopy and electron spin resonance. Microwave-absorbing properties were investigated by measuring complex permittivity, complex permeability and reflection loss in the 2-18 and 18-40 GHz microwave frequency range using the free space method. The results showed that a wider absorption frequency range could be obtained by adding different polyaniline contents in Ba ferrite. © 2010 Elsevier B.V. All rights reserved.
Tan K.-H.,Chung Cheng Institute of Technology
IEEE Transactions on Power Electronics | Year: 2016
A wavelet Petri fuzzy neural network (WPFNN) controller is proposed to control squirrel-cage induction generator (SCIG) system with an ac/dc power converter and a dc/ac power inverter for grid-connected wind power applications. First, the ac/dc power converter and the dc/ac power inverter are developed to deliver the electric power generated by a three-phase SCIG to power grid. Moreover, the ac/dc power converter and the dc/ac power inverter are mainly designed to control the mechanical rotor speed, dc-link voltage, and reactive power output of the SCIG system, respectively. Furthermore, since the varying active power outputs of the dc/ac power inverter seriously affect the tracking control of the dc-link voltage, a novel intelligent WPFNN controller is proposed to replace the traditional proportional-integral controller for the tracking control of the dc-link voltage in this study. In addition, the network structure and the online learning algorithm of the proposed WPFNN are described in detail. Finally, some experimental results are provided to show the effectiveness of the intelligent controlled-SCIG system using the proposed WPFNN controller for grid-connected wind power applications. © 1986-2012 IEEE.
Cheng S.-J.,Taiwan Power |
Miao J.-M.,National Pingtung University of Science and Technology |
Wu S.-J.,Chung Cheng Institute of Technology
Applied Energy | Year: 2013
The main purpose of this paper is to realize a metamodeling optimal approach that can be employed cost-efficiently and systematically to improve the performance of power density in PEMFC. First, an power density database is generated that corresponds to different levels of PEMFC unit operating parameters (factors) using the Design of Experiment (DoE) scheme, screening experiments, and Taguchi Orthogonal Array (OA). Then, metamodel is constructed by Radial Basis Function Neural Network (RBFNN) to represent the PEMFC system as a nonlinear complex model. The cross-validation procedure is implemented to prove the metamodel correctness and generalization. Moreover, Genetic Algorithm (GA) is applied to avoid local point and reduce time consumption to search the global optimum in promoting the performance of design factors. The proposed optimization methodology from experimental results provides an effective and economical approach to improve the performance of fuel cell unit and can be easy extended to the fuel cell stack system in energy applications. © 2013.
Lou D.-C.,Chang Gung University |
Hu C.-H.,Chung Cheng Institute of Technology
Information Sciences | Year: 2012
Statistical steganalysis schemes detect the existence of secret information embedded by steganography. The x 2-detection and Regular-Singular (RS)-attack methods are two well-known statistical steganalysis schemes used against LSB (least significant bit) steganography. The embedded message length can be estimated accurately by these two steganalysis schemes. For secret communication, the resistance of steganography against steganalysis is very important for information security. To avoid the enemy's attempts, the statistical features between stego-images and cover images should be as similar as possible for better resistance to steganalysis. In this manuscript, a reversible histogram transformation function-based LSB steganographic method is proposed to resist statistical steganalysis. The experimental results show that the proposed method resists not only RS-attack but also x 2-detection methods. © 2011 Elsevier Inc. All rights reserved.