Parameter optimization of continuous sputtering process based on Taguchi methods,neural networks,desirability function,and genetic algorithms

To combat climate change, many industries have participated in the research on alternative energies. Industrial Technology Research Institute in Taiwan has developed techniques for the solar energy selective absorption film continuous sputtering process. For this extremely complicated process, plenty of parameters would influence the output quality. If parameters settings simply rely on the experience of engineers, the defect rate may increase due to instability. A more reliable approach is desirable to optimize the condition of manufacturing process parameters, thus improving the quality.The present study applies a systematic procedure for the parameter optimization of the absorption film continuous sputtering process. First, possible variables are determined based on collected data and engineering knowledge. Second, Taguchi methods are utilized to search for the significant factors and the optimal level combination of parameters. Finally, the integration of back-propagation neural network, desirability function, and genetic algorithms is used to obtain the optimal parameters setting. According to the experiment results, the performance of the integrated procedure is better than that of Taguchi methods and traditional approach. Furthermore, if applying the integrated method, the saving energy would achieve 9770.53 kiloliter of oil equivalent (kLOE) per year, which is 11.2 times the saving kLOE of the traditional paint process.     

An adaptive neural fuzzy filter and its applications

A new kind of nonlinear adaptive filter, the adaptive neural fuzzy filter (ANFF), based upon a neural network's learning ability and fuzzy if-then rule structure, is proposed in this paper. The ANFF is inherently a feedforward multilayered connectionist network which can learn by itself according to numerical training data or expert knowledge represented by fuzzy if-then rules. The adaptation here includes the construction of fuzzy if-then rules (structure learning), and the tuning of the free parameters of membership functions (parameter learning). In the structure learning phase, fuzzy rules are found based on the matching of input-output clusters. In the parameter learning phase, a backpropagation-like adaptation algorithm is developed to minimize the output error. There are no hidden nodes (i.e., no membership functions and fuzzy rules) initially, and both the structure learning and parameter learning are performed concurrently as the adaptation proceeds. However, if some linguistic information about the design of the filter is available, such knowledge can be put into the ANFF to form an initial structure with hidden nodes. Two major advantages of the ANFF can thus be seen: 1) a priori knowledge can be incorporated into the ANFF which makes the fusion of numerical data and linguistic information in the filter possible; and 2) no predetermination, like the number of hidden nodes, must be given, since the ANFF can find its optimal structure and parameters automatically     

The role of coping in the relationship between perceived racism and racism-related stress for Asian Americans: Gender differences.

On the basis of stress and coping theory, the authors examined coping as a mediator of the relationship between perceptions of racism and racism-related stress with a sample of Asian American college students (N = 336). Results indicated that coping mediated the relationship between racism and racism-related stress differentially by gender. The more that men perceived racism, the more likely they were to use support-seeking coping strategies that were associated with higher levels of racism-related stress. The more that women perceived racism, the more they used active coping strategies that were associated with higher levels of racism-related stress. The findings demonstrate how coping with racism differs for Asian Americans on the basis of gender. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Lower-error and area-efficient antilogarithmic converters with bit-correction schemes

In this article, we have proposed fast lower-error and area-efficient antilogarithmic converters. By employing new approximation schemes with two-region shift-and-add operations, our proposed converters can perform high-speed conversions from logarithmic numbers to binary numbers. Synthesis results show that our proposed converters can achieve time savings of over 32.5% and can save 42.3% of the area used compared with previously proposed methods. In addition, the percent error ranges for various logarithmic number system (LNS)-based operations used by our proposed logarithmic processor are lower than those of previous methods. Our proposed converters can be applied to LNS-based processors to ease the tremendous computation overhead and boost the performance.     

A hybrid of genetic algorithm and particle swarm optimization for recurrent network design

An evolutionary recurrent network which automates the design of recurrent neural/fuzzy networks using a new evolutionary learning algorithm is proposed in this paper. This new evolutionary learning algorithm is based on a hybrid of genetic algorithm (GA) and particle swarm optimization (PSO), and is thus called HGAPSO. In HGAPSO, individuals in a new generation are created, not only by crossover and mutation operation as in GA, but also by PSO. The concept of elite strategy is adopted in HGAPSO, where the upper-half of the best-performing individuals in a population are regarded as elites. However, instead of being reproduced directly to the next generation, these elites are first enhanced. The group constituted by the elites is regarded as a swarm, and each elite corresponds to a particle within it. In this regard, the elites are enhanced by PSO, an operation which mimics the maturing phenomenon in nature. These enhanced elites constitute half of the population in the new generation, whereas the other half is generated by performing crossover and mutation operation on these enhanced elites. HGAPSO is applied to recurrent neural/fuzzy network design as follows. For recurrent neural network, a fully connected recurrent neural network is designed and applied to a temporal sequence production problem. For recurrent fuzzy network design, a Takagi-Sugeno-Kang-type recurrent fuzzy network is designed and applied to dynamic plant control. The performance of HGAPSO is compared to both GA and PSO in these recurrent networks design problems, demonstrating its superiority.     

Pattern recognition using a family of design algorithms based uponthe generalized probabilistic descent method

This paper provides a comprehensive introduction to a novel approach to pattern recognition which is based on the generalized probabilistic descent method (GPD) and its related design algorithms. The paper contains a survey of recent recognizer design techniques, the formulation of GPD, the concept of minimum classification error learning that is closely related to the GPD formalization, a relational analysis between GPD and other important design methods, and various embodiments of GPD-based design, including segmental-GPD, minimum spotting error training, discriminative utterance verification, and discriminative feature extraction. GPD development has its origins in basic pattern recognition and Bayes decision theory. It represents a simple but careful re-investigation of the classical theory and successfully leads to an innovative framework. For clarity of presentation, detailed discussions about its embodiments are provided for examples of speech pattern recognition tasks that use a distance-based classifier. Experimental results in speech pattern recognition tasks clearly demonstrate the remarkable utility of the family of GPD-based design algorithms     

Broadband Microwave Measurements of Overdoped Y0.9Ca0.1Ba2Cu3O7−δ Films Using Corbino Geometry

The surface impedance of Y_0.9Ca_0.1Ba₂Cu₃O_7?δ thin films was measured using the Corbino spectroscopy method. This special geometry, in which the sample dimensions are well defined by a ring pattern, is ideal for broadband high frequency reflection measurements. Using the complex reflected signal, S ₁₁, measured by a vector network analyzer, one can find the surface impedance of the thin film, from which the complex conductivity can be deduced. In the current work we present the three-standard approach for calibration of the Corbino method and demonstrate the benefits of this approach in measuring superconducting Y_0.9Ca_0.1Ba₂Cu₃O_7?δ thin films up to 20 GHz and down to 6 K. For the data analysis the well-known generalized two-fluid model was implemented, taking into account a film thickness which is much smaller than the normal state skin depth and superconducting penetration depth.     

Tribological impact of touchdown detection on bit patterned media robustness

Touchdown detection by thermal-flying height control (TFC) has been implemented to calibrate flying height (FH) of magnetic sliders for all hard disk drives. For bit patterned media (BPM) to be successful as a revolutionary technology to further increase recording areal density, it must experience the touchdown process with similar robustness as conventional continuous media. Here we numerically study the tribological impact of TFC touchdown detection on the continuous media and unplanarized BPM by three-dimensional (3D) transient finite-element models with the frictional heating and thermal-elastic-plastic materials included. Our results demonstrate that the continuous media exhibits no plastic deformation due to the TFC touchdown with an over-push as large as 2 nm, whereas the plastic strain of the BPM may reach 3 % at higher sliding velocities, and it exists over a wide range of bulge radius and disk velocity. Such plastic deformation can lead to permanent media damage and data loss. Besides, the temperature rise of the BPM (~27 K) is approximately 1.3 times of that of the continuous media (~21 K), and may have to be considered when designing a robust head-disk interface for BPM. Although planarization may improve slider’s flyability performance, our analysis shows that planarizing BPM with SiO₂ deteriorates the tribological robustness of the media in particular at a high disk velocity probably due to the inhomogeneous composition and mismatch of material properties between the filling material and recording material. Hence extreme caution must be exercised when choosing a filling material.     

A computational algorithm for catalysts experiencing concentration‐dependent deactivation

Abstract

A computational algorithm is proposed for catalyst pellets or reactors experiencing concentration‐dependent deactivation. In the integration of the deactivation equation in each time interval, the concentration of poison, reactant or/and product is considered to be a constant. The value of concentration is recalculated from the mass balance equation before integrating the deactivation equation. By such an approach, the number of equations is reduced; thus a two‐dimensional problem can be converted to a single‐dimensional one.